In this app note, we will show you how to tune a multisub setup (i.e. multiple subwoofers) using a miniDSP.

• Why use multiple subwoofers?
• Choice of DSP
• Getting connected
• The procedure
• Single-subwoofer baseline
• Method A
• Method B
• Method C
• Wrapping up

Why use multiple subwoofers? [Top]

The primary reason for using multiple subwoofers is to improve evenness of bass response across the whole listening area. While EQ can correct the response to be flat in one location in the room, it cannot correct for spatial variation. For example, if the level of 40 Hz in one seat is 10 dB different to the level in the next seat, the difference between the two seats will always be 10 dB, no matter how much EQ is applied. The only solutions are to find a better position for the (single) subwoofer, or to use multiple subwoofers.

Multiple subwoofers can also help with other problems. For example, a persistent null in the response can be filled in by using multiple subwoofers. (A null is not just an area of low output but an area of zero output. No amount of EQ can properly correct for a null.) Because real listening rooms are limited to where subwoofers can be placed (due to furniture, doors, and aesthetics), it may not be possible to locate a single subwoofer in the optimum position. Use of more than one subwoofer provides a greater degree of freedom with subwoofer location.

Even two subwoofers can deliver an improvement over a single (see graphs below), but three or four is generally considered optimal.

Choice of DSP [Top]

We will assume in this app note that the subwoofers are being used for home theater, driving two to four subs from the AVR's subwoofer output. If purchasing a dedicated DSP for subwoofer tuning, the four-output miniDSP units are ideal. They are summarized in Table 1.

Table 1. miniDSP four-output unit summary

While this app note is written for the case where a four-output unit is dedicated to subwoofer management, any miniDSP product with enough outputs can be used in this application. Units with a higher number of outputs may be suited for applications with more subwoofers, combined active crossover/multisub systems, and multiroom installations. See the full range. If in doubt or if you have any questions, please contact our tech support team.

Getting connected[Top]

An example setup is shown in Figure 1 below. The subwoofer output from the AVR (or AVP) is connected to the miniDSP input, and up to four powered subwoofers are connected to the miniDSP outputs.

Figure 1. Typical multisub setup using powered subwoofers

Figure 2 shows another option, where "pro" amps are used to obtain higher power levels (i.e. the subwoofers themselves do not have amplification built in).

Figure 2. Multisub setup using passive subwoofers and external amplification

Either of the above or a combination is a fine choice. If you have a simple system and don't need a full-featured AVR, you can use a miniDSP nanoAVR HDA instead of the AVR.

For a miniDSP 2x4 or miniDSP 2x4 Balanced, the best choice of plugin is the 4Way Advanced. For a miniDSP 2x4 HD, use the 2x4 HD1 plugin and set the routing matrix like this:

The procedure[Top]

The response of subwoofers in rooms is unpredictable and there are various approaches to obtaining the best response when using multiple subwoofers. Acoustic measurements are an essential part of this task. In this app note we will describe three methods in order of increasing difficulty and provide examples of measurements in our test room.

The overall procedure is:

1. Take multiple measurements of the sub or subs around the listening area.

2. Change parameters like delay and gain to reduce spatial variation. (You can also try changing subwoofer location.)

3. Run a global EQ to improve smoothness.

Note the general approach: multiple subs to reduce spatial variation, global EQ to improve smoothness. We will use three measurement locations – one in the center of the listening area and two in different locations. You can use more measurement locations but at some point it will be easier to switch to using Multi-Sub Optimizer (Method C).

Single-subwoofer baseline[Top]

We recommend that you establish a baseline using a single subwoofer. This will ensure that you have your miniDSP operating correctly with the rest of your equipment and are able to take measurements and run REW Auto-EQ. It will also provide a useful basis for comparison to assess the improvement obtained with additional subwoofers.

When taking measurements for optimization, ensure that all filtering is turned off. Since AVRs often have a low pass filter on the LFE channel, it's best to send the test signal directly into the miniDSP. In Figure 3, the computer running REW uses the USB input of a 2x4 HD for the test signal. If using a miniDSP 2x4 or 2x4 Balanced, disconnect the AVR from the miniDSP input and connect the output of your soundcard instead.

Figure 3. Example measurement setup (2x4 HD)

Choose one of the subwoofers to use as the baseline. In the plugin, mute the other subs, then take a number of measurements around the listening area. Graph 1 shows the measurements we obtained for a subwoofer located in the center of the front wall. This is the best location we have found for a single subwoofer in this room.

Graph 1. Uncorrected response of single subwoofer, measured at three locations

We now need to apply global EQ using Room EQ Wizard:

1. Use REW to take an average of the three measurements. (Use the All SPL tab to select the measurements, then click on the Average the Responses button.)

2. Use REW Auto-EQ to calculate a correction filter for the average response. Refer to the app note Auto EQ with Room EQ Wizard and be sure to use the correct sample rate setting and number of EQ bands for your hardware/plugin.

3. Load the correction filter into the PEQ block of input channel 1.

4. Measure again at the three locations to get the final response.

(For the sake of easier comparison, we use a completely flat target curve in all examples. You can set up a target curve with low bass boost if you wish, or use additional PEQ after applying global EQ.)

Graph 2 shows our measurements at the three locations after applying global EQ. So that's our baseline! Let's see what improvements we can get with more subs.

Graph 2. Corrected response of single subwoofer, measured at three locations

Method A [Top]

Method A uses two or four identical subwoofers positioned symmetrically in the room. It is based on research at Harman International (PDF download). The best subwoofer locations were as shown in Figure 4.

Figure 4. Symmetrical layouts for multiple subwoofers

To test Method A, we placed two identical sealed subs in our test room as per Layout 1. Graph 3 shows the result after measuring in three locations, applying global EQ to the average, and then remeasuring in the same three locations. The spatial variation is less than for a single subwoofer (Graph 2), except perhaps in the 90 to 105 Hz area, and there are fewer notches.

Graph 3. Corrected response of two subwoofers (Layout 1), measured at three locations

You may find that you can "tweak" the subs to improve the response. In our case, we found that adding 2 ms of delay to the front subwoofer eliminated a couple of notches in the response and reduced spatial variation above 80 Hz. Overall, compared to Graph 2, this is definitely an improvement!

Method B [Top]

Method A is not always feasible: real rooms are often not rectangular and the needed locations may not be available for practical reasons (furniture, equipment, doors/walkways, etc). Method B is based heavily on (but is not the same as) the technique described by Dr. Earl Geddes in two short papers here. It uses three or more subwoofers with asymmetrical or "random" placement. Figure 5 shows some example layouts with three subs (Geddes also recommends lifting one sub closer to the ceiling if possible).

Note: as part of the optimization process, the delay of individual subs will be changed. If you prefer to initialize the delays to correspond to the distance of each sub from the main listening position, there is no harm in doing so.

Figure 5. Sample asymmetrical layouts for multiple subwoofers

To test Method B, we used Layout 4 in our test room with three sealed subwoofers (similar but not all identical). Prior to starting, we measured all three subwoofers at three locations with no adjustment (delay, gain, EQ) applied to any of them. This is shown in Graph 4. While spatial variation is very low up to 50 Hz, it is quite high above 50 Hz. Furthermore, all measurement locations are quite "notchy" and this will be difficult to EQ.

Graph 4. Response of all three subwoofers (Layout 4) with no optimization, measured at three locations

So let's see what improvement we can get! Method B proceeds as follows:

1. Mute all subs except S1 and run a set of measurements.

2. Unmute S2 (leave S1 unmuted) and run the measurements again. If the result is better than step 1, proceed to the next step. Otherwise, use the plugin to change the gain and delay (*) of S2 and repeat.

3. Unmute S3 (leave S1 and S2 also unmuted) and run the measurements. If the result is better than step 2, proceed to the next step. Otherwise, use the plugin to change the gain and delay (*) of S3 and repeat.

4. If you are using four subs, do the same for S4.

5. Apply global EQ to the average of the last set of measurements. Then remeasure. Graph 5 shows our final result.

Graph 5. Optimized and corrected response of all three subwoofers (Layout 4), measured at three locations

This is obviously a big improvement in spatial variation compared to just using multiple subs without this method (Graph 4). There are also no significant notches up to nearly 100 Hz. Compared to Method A (Graph 3), there is more variation in the 20 to 40 Hz range and less above 50 Hz. We re-ran the measurements for each additional subwoofer only a small number of times, so it's possible that more time spent on varying sub parameters and measuring would yield even better results.

Method C [Top]

Method C uses the Multi-Sub Optimizer (MSO) program. This awesome piece of software runs an optimization algorithm on multiple measurements of your subs to generate the best possible response across the listening area.

In a nutshell, the procedure for using MSO is as follows:

1. Take a measurement of each subwoofer at each measurement location. In our example, with three subwoofers and three measurement locations, that means nine measurements. These measurements will need to use a timing reference.

2. Load the measurements into MSO. This requires some setup and configuration – for example, telling MSO that you have three subwoofers, assigning the correct measurements to each subwoofer, and so on.

3. Run the optimization algorithm to minimize spatial variation. MSO will generate gain, delay, and PEQ settings for each subwoofer.

4. Load the settings generated by MSO into the respective output channels of the miniDSP plugin.

5. Re-run the measurements to verify your results.

6. Optionally, apply global EQ to get the flattest response or to apply a target curve.

We have provided a detailed step-by-step guide for the DDRC-88A in the application note Optimizing multiple subwoofers with the DDRC-88BM and Multi-Sub Optimizer. With the exception of plugin configuration in steps 2 and 3, the instructions also apply to the miniDSP 2x4 / HD / Balanced. Take your time and follow all steps carefully. Be sure to set the correct sample rate and number of PEQs available for your hardware/plugin.

To test Method C, we used Layout 4 with the same three subwoofers that we used for Method B. Graph 6 shows our final result. Spatial variation is very low indeed to well above 100 Hz.

Graph 6. Response of all three subwoofers (Layout 4) optimized with MSO, measured in three locations

Wrapping up[Top]

It's important to remember that the graphs shown here are not necessarily representative of the results that you will get in your room. As mentioned before, subwoofers in rooms are unpredictable, and the only way to know which method is best for you is to try them.

Finally, if you are using bass management, you can now set it up. To do this, pass your measurement signal through the AVR instead of directly through the miniDSP. With a multisub system, the concept of "distance to subwoofer" can be a bit meaningless so you may find that you need to experiment with the delay on the subwoofer channel to get the smoothest response through the crossover region.

That's it for this app note! Have fun, and please let us know of your multisub and miniDSP results in our forum.

This app note will explain how to use the acoustic timing reference feature of Room EQ Wizard (REW) with a miniDSP UMIK-1 to accurately set delay between speakers in your home theater.

It assumes that you already have a working knowledge of measurements with REW. If not, first see the app note UMIK-1 setup with REW. REW version 5.15 or later is required.

• Introduction
• Setting things up
• Measure front left and right speakers
• Measure the remaining speakers
• Tabulate your results
• Update the configuration
• Wrapping Up
• Digging deeper

Introduction[Top]

Since version 5.15, REW can be set to use an acoustic timing reference. Typically, you will use one of your existing speakers for this purpose. You can then measure the frequency response of any speaker (including the one being used for the timing reference) with time delay embedded in the measurement. This has two key uses:

1. If you export measurements to use in another program, that program can use the timing information to understand time (or phase) delays between the measurements. See the app note Multi-Sub Optimizer with UMIK-1 for an example.

2. If you need to set time delay in an output channel of a miniDSP processor in order to time-align different speakers, REW will show you the delay of the measured speaker relative to the timing reference. This is a common requirement in home theater. This app note will show you how to measure and set these delays accurately.

A typical setup is shown in this diagram:

In this example, we are using a nanoAVR HDA. You can also use nanoAVR HD with an A/V Receiver, or a 10x10 HD with analog connections. The measurement technique will still apply even if you are not using a miniDSP multi-channel processor but the measurement procedure is a bit less convenient.

As shown in the diagram above, the left front speaker is used as the acoustic timing reference – this is indicated by the red "wave" coming from it. Each speaker (including the left front) will be measured for its time delay relative to this timing reference.

Setting things up[Top]

First of all, disable bass management in the nanoAVR-BM plugin. The easiest way to do this may be to switch to an unused configuration i.e. one that is at the default settings. Or, if you are connecting your computer directly to an AVR for speaker timing measurement, disable bass management in your AVR (often accomplished by setting all speakers to "large".)

Start REW. If you haven't already, set up the UMIK-1 calibration. For multichannel sound system measurement, it it best to point the microphone at the ceiling and provide your 90 degree calibration file when REW asks for it. (You can download this file based on your microphone's serial number from the UMIK-1 page.)

Open the REW Preferences window and on the Analysis tab, under Impulse Response Calculation, select "Use acoustic timing reference" from the drop-down menu.

Measure front left and right speakers[Top]

Click on the "Measure" button (top left of the main window). Along with the other key settings (frequency sweep range and level), set both the output channel and the Timing Reference channel to "left", like this:

Now run a levels check and then the measurement sweep. You will hear a short "peep" from the left speaker, shortly followed by a full frequency range sweep. Rename the measurement to "Front Left".

Click on the "Measure" button again. This time, set the output channel to "Right" but leave the timing reference at "Left." You will use this setting for all remaining speakers. It looks like this:

Now run the measurement sweep and label it "Front Right."

Measure the remaining speakers[Top]

To measure the remaining speakers, the measurement signal needs to be routed to the other channels. The simplest way to do this is to use the Routing tab in the nanoAVR-BM plugin, which will be described below. Alternatively, you can use the method of routing to different HDMI channels described in our app notes Using the UMIK-1 and REW with HDMI output - Windows and Using the UMIK-1 and REW with HDMI output - Mac.

To measure the subwoofer, route input channel FR In to Sub Out, like this:

(The other rows don't matter, just FL In and FR In.) Now run a measurement sweep, leaving the output channel set to "Right" and the reference channel set to "Left," as set previously. Rename the measurement to "Subwoofer."

Repeat for the rest of the speakers. Here for example is the routing to measure the center speaker:

Tabulate your results[Top]

On the REW main window, click on the Front Left speaker measurement, and then click on the Info button:

Locate the parameter "System Delay" on the window that pops up, and write it down. For the front left speaker, it should be at or close to zero.

Other speakers will have a different delay. For example, this is our subwoofer:

Click on each of the other measurements and write down their value of System Delay. Here are the results from our test system:

  Left front:      0.00
  Right front:     0.02
  Subwoofer:       2.69
  Center:         -0.27
  Left Surround:  -2.33
  Right surround: -3.04

If the delay value is positive (right front and subwoofer in our example), it means that the speaker being measured is further away than the timing reference. If the value is negative (center and surrounds in our example), it means that the speaker being measured is closer than the timing reference. In our example, the microphone was very slightly closer to the left speaker than the right.

Update the configuration[Top]

Now we need to calculate the delay to apply to each speaker. Identify the speaker with the highest positive delay value. In our example, that is the subwoofer with 2.69 ms delay. Subtract each speaker's delay from that value and write it down. In our example:

  Left front:      2.69 -  0.00 = 2.69
  Right front:     2.69 -  0.02 = 2.69 (*)
  Subwoofer:       2.69 -  2.69 = 0.00
  Center:          2.69 - -0.27 = 2.96
  Left Surround:   2.69 - -2.33 = 5.29
  Right surround:  2.69 - -3.04 = 5.73

(*) We are setting the same delay on front left and front right, as the slight difference is a measurement error.

Return to your regular configuration in the nanoAVR-BM plugin and set these delays in the output channels. Like this:

Wrapping up[Top][Top]

And that's it! You can now set up your bass management, based on accurate delays between all speakers. Don't forget to save your configuration to a file regularly.

Have fun, and please let us know how you go in our forum. If you want to learn a bit more about how this feature works, read on to the "Digging Deeper" section below.

Digging deeper[Top]

What's going on here? We can gain a better understanding by viewing the impulse response of the speakers. Click on the Overlay button and then on the Impulse button above the plot:

Select the plots that you want to display and adjust the graph limits. You should see plots that look as in the following example. (If not, move the cursor onto the plot and select %FS from the drop-down menu. Adjust the time scale in seconds. Below, for example, we have set the scale from –0.003 to 0.003 seconds, or –3 to 3 ms.)

In the above screenshot, the left speaker's impulse response is shown in red. You can see that its largest peak is at 0 ms. This is where REW has determined the delay of the speaker relative to the timing reference to be.

The green plot shows the impulse response of the left surround speaker. We have placed the cursor at the location of its largest peak, which is –2.333 ms. That is, the left surround speaker is 2.333 ms "closer" to the microphone than the timing reference. This is the same (within margin of error and cursor placement) as the value that REW determined as the delay for this speaker.

Typically, the speaker being used as the timing reference has zero delay, but this is not always the case. The timing reference signal (the initial "peep") runs from 5 kHz to 20 kHz, whereas the measured response of the whole speaker runs over the full range. The impulse response of the full range measurement may be sufficiently different to the reference impulse response that the calculated delay is not zero, although it will be small. This is taken care of in the calculation method explained above.

The miniDSP DDRC-24 includes not only Dirac Live^® but also a full set of crossover functions on its four output channels. One use of this is as a flexible tool for integrating subwoofers into your system along with Dirac Live. This app note shows you how.

This app note assumes that you have already installed the DDRC-24 utility and Dirac Live Calibration tool as per the User Manual.

• Getting connected
• Configuring the plugin
• Initial Measurement
• Run Your Dirac Live Calibration
• Wrapping Up

Getting connected[Top]

Connect the system as shown in this diagram:

The diagram assumes that you are using one subwoofer. If you are using two subwoofers, you can connect the second to output 4.

Configuring the plugin[Top]

Start the DDRC-24 plugin and click the Connect button. (See the User Manual for more information.) Set the Routing matrix like this (you can use the Outputs tab to change the displayed labels if you wish):

On the Outputs tab, click on the Xover button for channel 1. Here you will set a high pass filter, to remove low frequencies from the left speaker. Here is a typical example:

Do the same for the right speaker (channel 2).

For the subwoofer (channel 3), set a low pass filter to remove high frequencies. Here is a typical example:

Initial Measurement[Top]

Although not absolutely essential, it is recommended that you perform a measurement to check for initial integration between the subwoofer and the main speakers around the crossover frequency. While Dirac Live will correct for frequency response anomalies, it can't fully correct if you have a large "hole" in the frequency response at the crossover.

If you wish, you use Room EQ Wizard (REW) or a similar program to do this measurement. To use REW with the UMIK-1, please refer to our application section on acoustic measurements.

If you don't wish to learn how to use REW, you can simply use the Dirac Live Calibration Tool to perform this measurement. Save your configuration to a file and quit the DDRC-24 plugin first. Then follow the normal calibration procedure described in the User Manual, but only perform a single measurement (instead of the full set of nine) before proceeding to the Filter Design tab. The "before" measurement shows the combined responses of the subwoofer and speakers:

If there is a large "hole" in the response around the crossover, you will need to make some adjustments and remeasure. Quit DLCT before starting the DDRC-24 plugin again. Adding a delay to either the speakers or the subwoofer will change the response, as will inverting the subwoofer. You can also adjust the low pass and high pass crossover frequencies and change the slope. If the subwoofer is much higher or lower in level than the speakers, adjust the gain on the subwoofer.

See the User Manual for information on how to make these adjustments. After making adjustments, repeat the measurement.

Run Your Dirac Live Calibration[Top]

You can now proceed to run the full set of nine measurements for the Dirac Live calibration. Adjust the target curve to suit your preference. For example, this screenshot shows a target curve with an elevated bass level, which many people prefer, and the "after" response:

Wrapping up[Top]

If you haven't already, use the learning remote feature to set up remote control for volume, mute, preset selection, and Dirac Live filtering. Then sit back and enjoy some tunes!

That's it for this app note! Have fun, and please let us know how you go in our forum.

Apple TV 4th generation ("ATV4") is the latest generation of audio and video streamer from Apple Inc. New to generation 4 is inbuilt Dolby® decoding for multichannel audio. This new feature means that it is now possible to connect it directly to the nanoAVR DL, miniDSP's tiny audio processor with Dirac Live®, the world's premiere room correction system.

The diagram below illustrates how to set up the Apple TV together with the nanoAVR DL. Just connect the HDMI output of the ATV4 to one of the HDMI inputs of the nanoAVR DL. (You can connect another source like a Blu-ray player to the second input of the nanoAVR DL, provided that the source generates multichannel linear PCM.) The output from the nanoAVR DL connects to an HDMI input on your receiver.

This photograph shows the HDMI connection from the ATV4 to the nanoAVR DL (other connections are left out for clarity):

To set up the Apple TV to output multichannel linear PCM to the nanoAVR DL, set its "Surround Sound" setting to "Auto", as described in this article from Dolby Laboratories:

• http://developer.dolby.com/News/Dolby_Audio_Support_on_Apple_TV.aspx

Now you can set up the nanoAVR DL as described in our comprehensive User Manual. When you are done, all audio from the Apple TV will have Dirac Live® room correction applied to it!

That's it for this app note! Have fun, and please let us know about your Apple TV and miniDSP experience in our forum.

JRiver Media Center is a popular playback program for Windows and Mac. JRiver recently made a version of Media Center available for the Raspberry Pi, a small embedded computing platform. Combined with a U-DAC8, it makes a very compact and effective networked multichannel (and stereo) audio playback solution!

• 1. Getting Started
• 2. Setting up the U-DAC8
• 3. Setting up an audio library
• 4. Going headless
• 5. Limitations
• Wrapping Up

1. Getting Started [Top]

You will need a Raspberry Pi 3 and a miniDSP U-DAC8. Simply connect the USB cable from one of the Raspberry Pi's USB outputs to the U-DAC8 and connect the outputs of the U-DAC8 to your amplification and subwoofers, as shown in this diagram:

Install the full distribution of the Raspbian operating system on an SD card. You can find downloads and instructions here:

• https://www.raspberrypi.org/documentation/installation/installing-images/README.md

Connect a monitor or TV via HDMI and a keyboard and mouse to the USB ports, then insert the SD card into the Raspberry Pi's SD card slot and power it on. You can then install JRiver Media Center by following these instructions:

• https://yabb.jriver.com/interact/index.php?topic=95578.0

2. Setting up the U-DAC8[Top]

Start up JRiver Media Center (JRMC) on the Raspberry Pi - drop down the Raspberry menu, go to "Sound and Media," and select "JRiver Media Center."

In JRMC, drop down the Player menu and select Playback Options. Click on the Audio Device, scroll down and select "surround71:CARD=UDAC8,DEV=0 [ALSA]".

To use bass management or to have JRiver "upmix" stereo content to 5.1 outputs:

1. Drop down the Player menu and select DSP Studio.
2. Click on the Output Format item in the list on the left. (Make sure you turn on the checkbox to enable it.)
3. For Channels over on the right, drop down the menu and select "5.1 Channels."

To set up bass management, click on the Room Correction item in the left (turn on the checkbox), and proceed through each speaker to set up the crossover to the subwoofer. Here is a typical example:

(If necessary, you can adjust levels and delays on this screen as well.) And that's it! Now set up an audio library.

3. Setting up an audio library[Top]

You can attach a hard drive containing music files directly to the Raspberry Pi. (It's best if it's a powered drive, as the Raspberry Pi won't supply a lot of power over USB.) You can even just plug in a USB stick containing music files. Use the JRiver documentation for setting up a library:

• https://wiki.jriver.com/index.php/Library

Once you've done that, browse to the album view, select an album, and click "Play"! Here's how it looks when we set up a small library of multichannel test tracks (from 2L of Norway):

Alternatively, if you have your library on a network drive, you can simply mount that drive on the Raspberry Pi. Follow these instructions to mount a network drive (assuming that your NAS supports SMB):

• http://geeks.noeit.com/mount-an-smb-network-drive-on-raspberry-pi/

Then you can simply do the same as above for an attached drive, but browse to the location that you mounted the network drive. (Note: if you have difficulty with the network drive not mounting on boot, add the following to the options in fstab (right after "iocharset=utf8"): ,_netdev,x-systemd.automount).

4. Going headless[Top]

If you don't want to leave a screen, keyboard and mouse attached to the Raspberry Pi, you can make it "headless." Follow these instructions:

• http://yabb.jriver.com/interact/index.php?topic=95578.msg674431#msg674431

With that done, you can connect to the Raspberry Pi 3 with a remote desktop. However, you will most likely also want to install JRemote on your iPad or your Android tablet. Here's how it looks with our multichannel test library:

5. Limitations[Top]

Here are some limitations to be aware of:

• This app note is relevant to audio source files (e.g. FLAC format) only. The Raspberry Pi may not have the CPU power to decode video in JRiver Media Center.
• The Raspberry Pi does not have enough CPU power to transcode multichannel DSD into PCM. So: this setup is only for multichannel PCM audio files (e.g. FLAC).

Please note also that miniDSP cannot provide support for third-party hardware or software. While this app note showed you how to set up the miniDSP U-DAC8 with JRiver Media Center running on a Raspberry Pi 3, the features and functions of this hardware and software are beyond the scope of miniDSP support.

Wrapping up[Top]

That's it for this app note! If you want to try some multichannel audio files, you can download samples from 2L of Norway.

Have fun, and please let us know about your Raspberry Pi experience in our forum.

How to time-align speaker drivers

In this application note, we show you how to "time align" the drivers in your active DSP loudspeaker using the delay function of the miniDSP platforms.

Why time-align?

In a multi-way loudspeaker, the sound from different drivers will take slightly different amounts of time to reach your ears. The figure below illustrates a typical scenario: since the woofer cone is "deeper" than the tweeter, the acoustic center of the woofer is further away from the listener's ear than the tweeter. The sound from the woofer will therefore arrive at the listener's ear slightly later than the sound from the tweeter. This can have a detrimental effect on the speaker's response around the crossover.

One approach to this issue is to build a slanted or stepped baffle. In passive crossover design, the designer may use specific techniques to shift the phase of one driver or another to compensate. In a miniDSP crossover, we use the Delay parameter in the output channels to delay the signal from the tweeter by a small amount of time. The result is that the acoustic waveform from both drivers will arrive at the listener's ear at the same time.

How much time delay?

Sounds travels at 343 meters per second, or 1126 feet per second. If you can measure approximately the distance between the tweeter and the woofer cone as indicated above, you can calculate the time delay with one of these formulae:

• Delay in milliseconds = distance in cm / 34.3
• Delay in milliseconds = distance in inches / 13.5

Alternatively, if you have an acoustic measurement program, you can calculate the delay by measuring the arrival time of the signal from each driver, and calculating the difference. The result is the time delay needed. The arrival time measurement can be done in Room EQ Wizard (REW) with the use of a two-channel soundcard - see the REW documentation for Use Loopback as Timing Reference.

Another method is to run a measurement sweep on both drivers at the same time. If one driver is significantly delayed, you will be able to measure the time difference between the impulse response peaks.

How to set time delay

Almost all miniDSP DSP platforms will support time alignment, but for the sake of clarity, we'll use the miniDSP 2x4. In the interface to any of the miniDSP 2x4 plugins (Stereo 2-way PEQ, 4-way PEQ, and so on), click on the Delay/Gain/RMS block. Set the time delay of the closest driver as indicated in the screenshot below.

In the plugins with eight output channels (4x10, 10x10, miniSharc), all delays appear on the Output tab. With these plugins, you can implement three-way or four-way speakers and use time delays to time-align all drivers. Just remember to calculate delays relative to the furthest-away driver. This screenshot shows four channels of one of these plugins:

Wrapping up

Once you have delays set correctly for all drivers, your crossovers will be easier to design. Don't forget to ask on the miniDSP forum if you have further questions!

Looking for an Audio processor doing time alignment?
Check out the miniDSP webstore.

In this application note, we will explain what gain structure is, and how to use this concept to optimize your DSP-based system.

Some basics

Any component of a hifi system has some fundamental properties, like gain, noise, and maximum signal levels. When multiple components are combined into a system, mismatches in these factors can sometimes lead to unsatisfactory performance, such as excessive noise or signal overload in one or more components.

Noise

Noise is present in all electronic systems. You cannot get away from it, but it must be minimized to an acceptable level. In hifi components, the noise level is typically specified as a signal-to-noise ratio (SNR), or how far down the noise is relative to 1 V RMS (typically for line-level components) or full power output (for power amplifiers).

Maximum signal levels

A component has a maximum signal level that it can accept on its input, or generate on its output. In a typical analog line-level component, the output signal is the one to worry about, and is usually determined by the power supply voltage. DSP-based components, however, have very specific maximum input and output levels. In the miniDSP 2x4 kit, for example, the maximum input level can be set by a jumper to either 0.9 V RMS or 2.0 V RMS, and the maximum output level is always 0.9 V RMS.

Gain

The gain of a component is the ratio of its output signal level to its input signal level. A preamp with 12 dB of gain, for example, would generate an output signal four times higher than the input signal (with the volume control turned all the way up). A buffer is a special type of component that has 0 dB or unity gain - the output signal is the same as the input signal. A DSP component typically has 0 dB of gain, although some units may vary from this or have switchable input or output levels.

With loudspeakers, we can consider the sensitivity of the speaker (or the drive units, in the case of an active speaker) to be analogous to gain. That is, a speaker with high sensitivity will need less power to produce a desired acoustic output level, which will mean lower signal levels in the electronics components earlier in the chain.

Gain structure

The concept of gain structure is that, at each connection between components in the system, the signal level is as high as it can be (to minimize noise), but no higher than the maximum level that either component allows (so there is no distortion due to overload).

In the diagram below, the dynamic range of a music signal is indicated by a colored rectangle. The signal in red is too high, and will cause distortion. The signal in yellow is too low and the lower signals in the music are below the noise floor. The signal in green is "just right" and represents the ideal music signal level at that connection point. In an optimized system, each interconnection point would be "green" when the volume control is set to the maximum volume that you ever listen to.

It's not generally necessary to compute exactly all of the gains and signal levels throughout a system - understanding the principles of gain structure is usually enough to achieve a good result. The following tips cover the key points for a DSP-based system.

DSP input level

The input signal level to the DSP must never exceed the maximum allowed. If you are connecting a source component such as a DAC directly to a DSP, check the DAC's specifications for maximum output level and ensure that the maximum input level of the DSP is equal to or greater than this. For example, if the DAC max output is 2 V RMS, then you cannot use a 2x4 miniDSP set to 0.9 V maximum input. You must use the 2.0 V setting.

If the DSP is connected to a preamp output, then the 0.9 V input setting is usually preferred, as the signal is already attenuated in the preamp. In most domestic systems, 0.9 V RMS (the maximum DSP output level) will drive the system to high levels, but if higher analog signal levels are needed into the power amps, a unit such as the miniDSP Balanced 2x4 or miniDSP 2x8 kit can be used.

DSP output level

As a general rule, the maximum DSP output level should be equal to or greater than the signal level required to drive your amps to full output power.

However, a problem that occurs in many systems is that the amplifier/speaker combination has too much gain. That is, the maximum output power is never ever used. In this case, the signal level earlier in the chain is pushed down into the "yellow" zone, thus raising the noise present at the speakers. One solution is to reduce the gain of the power amps; another is to place a passive attenuator at the input of the power amps.

Matching levels in an active speaker

In an active speaker, each power amplifier/driver combination needs to be considered. One of the advantages of an active speaker system is that drivers with different sensitivity can be easily used in the same speaker. However, if not done with some care, this can lead to some of the problems noted above. If a loudspeaker driver is much more sensitive than another (for example, you have a 100 dB/W/m horn with an 88 dB/W/m woofer), then you have a mismatch that should be dealt with by reducing the gain of the amplifier on the more sensitive driver. Alternatively, use a passive attenuator at the amplifier input.

This concept of gain structure, taken to its logical conclusion, means that not only are the amplifier gains and corresponding driver sensitivities matched to each other, but the gain of all amplifier/driver combinations is fine-tuned so that the system is capable of producing the maximum output level that you desire, but no more. This ensures minimum noise in the system.

Summary

Optimizing the gain structure of your DSP-based system will result in as much SPL as you need, while minimizing noise. A complete analysis of every component is generally not necessary, but keeping in mind some simple guidelines will ensure success.

One of the confusing things for many people new to DSP equalization (EQ) is the choice between graphic and parametric EQ. In this application note, we explain the difference so you can choose which suits your application best.

Parametric EQ

In a parametric equalizer, each filter cuts or boosts a range of frequencies. Each filter has three controls:

• frequency: the center of the frequency range to be cut or boosted
• gain: the amount of boost or cut
• Q: the "sharpness" of the boost or cut, with higher Q meaning a narrower filter

Parametric EQ thus allows a single filter to be very narrow or quite wide, and it is therefore very useful for correcting frequency response errors in a loudspeaker or reducing peaks caused by room modes. When implemented digitally, parametric filters can also take the shape of a "shelving" filter, which boost or cuts frequencies above or below the filter frequency.

The screenshots below shows four filters in use: a low-shelf filter (boost), a narrow notch, a broad boost, and a high-shelf filter (cut). First the frequency curve as shown in the miniDSP plugin:

Here is the frequency response measured with Room EQ Wizard:

Many of the miniDSP plugins support parametric EQ on each input channel and each output channel, with five or six filters in each. Check the miniDSP plugins page for e.g. "6 band PEQ".

Graphic EQ

A graphic equalizer has a number of filters spread evenly across the audio bandwidth. Each filter is the same shape, and has just one control: the amount of boost or cut. The filters overlap, so the combined response forms the shape given by the positions of the sliders. Most common is the third-octave, or 31-band, graphic EQ, with center frequencies from 20 Hz to 20 kHz.

Graphic EQ can be implemented with analog circuits or with digital signal processing. In the miniDSP lineup, graphic EQ is a 31-band equalizer with up to 12 dB of boost or cut in each band. Here is an example of a miniDSP graphic EQ setting, set to approximately the same response as the PEQ above:

Here is the frequency response measured with Room EQ Wizard:

Check the miniDSP plugins page for "31 band EQ" to determine which plugins have a graphic EQ.

Which to use?

The type of EQ that you use depends on your application. Provided that you have a measurement microphone and software set up, the parametric EQ is ideal for correcting speaker and room response. If you want to be able to adjust the overall response by ear, then the graphic EQ may be better suited to your needs. You can of course try both and pick the one that suits you best.

A point to note is that the parametric EQ gives more flexibility for the same processing power. Each band of a graphic EQ requires the same processing power as a parametric EQ filter. So a 31-band graphic EQ uses the same processing time as 31 parametric EQ filters. That is why the parametric EQ plugins have EQ on every input and output channel, while the graphic EQ plugins have EQ only on the input channels.

Looking for an Audio processor doing PEQ or Graphic EQ?
Check out the miniDSP's innovative DSP platforms.

In this application note, we will explain the difference between FIR ("finite impulse response") and IIR ("infinite impulse response") filtering.

Infinite impulse response (IIR) filters

IIR filters are the most efficient type of filter to implement in DSP (digital signal processing). They are usually provided as "biquad" filters. For example, in the parametric EQ block of a miniDSP plugin, each peak/notch or shelving filter is a single biquad. In the crossover blocks, each crossover uses up to 4 biquads. Each band of a graphic EQ is a single biquad, so a full 31-band graphic EQ uses 31 biquads per channel.

The amount of processing that is required to compute a biquad is relatively small. This is what enables the low-cost miniDSP products to implement a full active crossover with parametric EQ on all input and output channels. The DSP (digital signal processor) on each board can compute a certain number of biquads, and this is the primary thing that determines how many filters are available in each plugin.

The miniDSP biquads can be programmed using the crossover parameters (slope and frequency), the parametric filter parameters (center frequency, gain, and Q), and so on. They can also be programmed with custom filter shapes by directly entering the biquad coefficients - five numbers that are used to compute the biquad output from its input. You can generate these coefficients by using the community-contributed custom biquad programming spreadsheet.

Finite impulse response (FIR) filters

An FIR filter requires more computation time on the DSP and more memory. The DSP chip therefore needs to be more powerful. miniDSP products that support FIR filtering include the OpenDRC and the miniSHARC kit.

FIR filters are specified using a large array of numbers. In the case of the OpenDRC, there are 6144 coefficients (or "taps") per channel. In the case of the miniSHARC, there are a total of 10240 taps assignable to all input and output channels. Generation of this large array of numbers must be done in a separate program, such as rephase, Acourate, and others.

FIR filtering has these advantages over IIR filtering:

1. It can implement linear-phase filtering. This means that the filter has no phase shift across the frequency band. Alternately, the phase can be corrected independently of the amplitude. See examples below.
2. It can be used to correct frequency-response errors in a loudspeaker to a finer degree of precision than using IIRs.

However, FIRs can be limited in resolution at low frequencies, and the success of applying FIR filters depends greatly on the program that is used to generate the filter coefficients. Usage is generally more complicated and time-consuming than IIR filters.

Examples of FIR and IIR

Here we will provide some simple examples to illustrate the difference between FIR and IIR.

Crossover filter

In a two-way crossover filter, the low pass and high pass outputs are sent to the woofer and tweeter respectively and are summed acoustically. We can simulate this behavior electrically - the figure below shows the measured phase response of a summed fourth-order Linkwitz-Riley crossover (24 dB/octave) at 300 Hz in blue. The phase of this crossover shifts by 360 degrees from low frequencies to high frequencies.

Shown in red is the measured output from a crossover with the same amplitude response curves, but implemented with a linear-phase FIR filter. The phase shift is very close to zero across the audio band.

Parametric filter

Parametric filters also have a phase shift. Consider a parametric filter with this response:

Below is the measured phase shift of this filter, in blue as implemented by an IIR filter, and in red as implemented by a linear-phase FIR filter. Again, the linear-phase filter has minimal phase shift across the audio band.

Note that sometimes the phase shift shown above is desirable, as it acts to correct phase as well as amplitude errors in (for example) the speaker driver being corrected. FIR filters can implement this curve either way: with the phase shift (minimum phase) or without it (linear phase).

Summary

FIR filters are more powerful than IIR filters, but also require more processing power and more work to set up the filters. They are also less easy to change "on the fly" as you can by tweaking (say) the frequency setting of a parametric (IIR) filter. However, their greater power means more flexibility and ability to finely adjust the response of your active loudspeaker.

Looking for an Audio processor with FIR or IIR support?
Check out the OpenDRC series (FIR) or miniDSP series (IIR) @  miniDSP.com

Welcome to the world of system tuning and DSP!

There isn't much to argue on the fact that speaker building is an art requiring time and understanding of acoustic theories. While some simulation software may help for the design of your enclosure, audio measurements are the basis for any evaluation of your system. No matter how pretty your box is or the choice of high specs drivers, your "system" (i.e. enclosure + speakers + crossover) may very well have poor specifications if it isn't properly tuned (equalization/time alignment).

Thanks to the partnership between the miniDSP platforms and the well known freeware Room Eq Wizard (REW), miniDSP is much more than a digital audio processor. By providing a complete measurement + system tuning solution fitting in the palm of your hand, the miniDSP product line innovates once again, yet keeping our price very competitive.

What's needed?

You will need:

• A miniDSP plugin that supports the "REW integration" feature. (Suitable examples includes the Advanced 2-way, Advanced 2.1, Advanced 4-way, and the 2x8, 4x8, 4x10 and 10x10 Crossovers.)
• The miniDSP hardware that runs the plugin - see the "Supported platforms" row on the plugin page linked above.
• Room EQ Wizard (REW), which is a free download from Home Theater Shack.
• A measurement microphone, such as the UMIK-1.

1. Make the initial measurement

In the miniDSP interface, disable all equalization and crossover filters on your subwoofer. Turn off all amps except the subwoofer amp(s), or disable the other outputs in the miniDSP interface. This is so you can measure the "raw" subwoofer response.

Then, in REW, run a measurement sweep from 10 Hz to 300 Hz. Ensure that your SPL is adequate to get a clean measurement.

2. Calculate the correction filters in REW

From the REW main screen, open the "EQ" tab.

Set the parameters as shown in the following screenshot. Note that in the Equaliser section you will need to select "MiniDSP" for plugins running at 48 kHz, and "MiniDSP-96k" for plugins running at 96 kHz.

Next, click on "Set Target Level" to set the target level to REW's recommendation. (You can manually change it later.)

Then click on "Match Response to Target." REW will generate filters to optimize the response and display the predicted response (response after correction). Here is an example with the measured response in dark blue and the predicted response in light blue:

You can see the filter settings calculated by REW by clicking on the "EQ Filters" button near the top of the screen:

In this example, only three filters were needed. If REW uses a lot of filters or uses filters with very high Q (> 20), experiment with the target level setting to try and get a smoother correction curve. You can also experiment with the flatness target and the max boost settings to get the best predicted response.

3. Load the correction filters into the miniDSP

Once the predicted response is satisfactory, click on "Send Filter Settings to Equaliser" and save the settings as a file.

Then, in the miniDSP plugin, navigate to the Parametric EQ block that will be equalizing the subwoofer. Set it to "Advanced" mode and click on "Import REW File." Select the file that you just saved, and the plugin will then display the correction filter:

Confirm your result by running your measurement sweep again - it should be much smoother than before! Here in light blue is the new measurement, compared to the original measurement in dark blue - as you can see, the measured result is almost identical to REW's prediction:

4. Integrate the subwoofer

In the crossover block in the miniDSP interface, re-enable the subwoofer's low pass filter. If you have a ported sub, you may wish to also add a high pass filter to protect the sub from over-excursion at high levels and low frequencies.

Then enable the other outputs (main speakers or drivers) and run full-range measurement sweeps. If there is a bump or dip at the crossover point, try adjusting the crossover settings (frequency, slope, phase) and the time delay of the subwoofer to flatten it out.

Next steps

Once you have mastered the above procedure, you can improve your system even further:

• Try different subwoofer locations to find the one with the smoothest response. The smoother the response without equalization, the better the result will be with equalization!
• Make several measurements around the listening area, and in the REW All SPL tab, click on "Average the Responses." Then use that average for the EQ calculation.
• If you have a sealed sub, apply a Linkwitz Transform.

You will now have the smoothest bass response you could imagine! Don't forget to ask on the miniDSP forum if you have questions.

The miniDSP's concept is to provide easy to use Digital Signal Processors(DSP) for audio filtering applications. From system tuning to multi-way crossover, our software programmable platforms can fulfil your dream audio project in few clicks of a mouse. There are few key points to remember while reading this website:

• The Hardware platform handles the signal processing. It is the brain of your system and as you'll see from this website, we do have many "variants" to fit your needs.
• The plug-in is the software user interface(UI) used for control/configuration of the DSP from a PC/Mac environment. At least one plug-in must be purchased to operate your system.
• Once configured, the hardware platform does not require a computer anymore as all settings are stored in memory. They are automatically reloaded at each boot-up.
• Finally, miniDSP kits are DIY/OEM friendly with upgrade abilities to build your custom system.

Now that you understand the basic operation of our software controllable DSP, let's decide which type of hardware platform you will need step by step: 

STEP1: Before thinking of the type of filter (FIR/IIR) you need, you first need to define the "I/O" (Input/Output) configuration of your system. Each system is unique so if you're unsure, start simple! Get a basic hand written diagram to confirm your system configuration. (e.g. CD player = 2 x IN, 2 way stereo speaker = 4 x OUT)

STEP2: Try to match your I/O configuration above to a "Hardware DSP platform". You can see that some boards are sold as Kits (i.e. no box) and others in a box.

STEP3: Next step is to figure out which "plug-in" will work for your application and your platform. Check out the plug-in page for a good visual summary of plug-ins and supported hardware platform.

STEP4: Hardware + Plug-in figured out, you're ready to order!

The Problem

Mother nature gave us a lot of gifts, but a perfect acoustic space for our Hifi systems wasn't one of them!

Put it simply, acoustic theories are very complex to measure, analyze, model and unless your room was designed by a world famous acoustician, it will always be an uphill battle against the laws of physics. While a certain amount of audio filtering; also called room correction can help to perfect the audio performance, achieving flat frequency response by boosting the signal at certain frequencies will only solve the issue if problems are not time dependant (see more below). Of course, quality of your loudspeakers, amplifiers and any other devices on the audio chain matters, but in the end acoustic theories trump. Following this little disclaimer, let's have a look at what our miniDSP board can do.

Let’s take as an example the Hifi system of a living room illustrated in the below picture.

Many elements will affect quality of the audio system, such as:

•    The window that won’t absorb acoustic energy and will most likely reflect lots of high frequency energy from the left speaker
•    The imperfect free field frequency response of your loudspeaker system
•    The asymmetrical room layout which will affect the stereo field
•    Or simply the listener’s position by the sofa. So close to the back wall, he will most likely experience more low frequency than he should from the so-called standing waves. All together, the end result is an imperfect acoustic space where even the best Hifi audio system will hardly produce a flat frequency response.

How MiniDSP kits can help?

So besides taking down walls, moving your sofa around the room 5 times or considering changing the color of your speaker cable, there are few tricks which can help you improve the overall system performance. Digital audio processors like our MiniDSP Kit are one of them. Thanks to their configuration flexibility, they provide a low cost option to configure audio filters, equalizers and signal monitoring from a simple software interface. However note that there is a limit to how much digital signal processing can solve acoustic issues. In particular, equalization will only solve amplitude related issues.While room modes and other comb filtering (cancelations of waves) issues will require either time alignement or acoustic treatment.

What miniDSP can do, is provide you with a set of filters to easily improve the performance of your speaker system:
•    Graphic equalizer: This bank of filters will  increase or decrease the signal at a specific frequency. Depending of the MiniDSP firmware configuration, you may have 31 bands (also called 1/3 octave) or 15 bands (2/3 octave) available. By boosting/dampening the signal at certain frequencies, speaker equalization is the most common tool for system tuning
•    Parametric equalizer: Similar to the graphic equalizer, but this time with only a limited number of equalization bands that can easily be configured to better fit the needs of the application.
•    Delay: In order for loudspeakers to produce constructive wave front  and prevent as much comb-filtering effect as possible, time alignement of different loudspeaker are typically required. The delay blocks will also come handy in digital crossover applications.

So to summarize this application note, flat frequency response of a system will not happen by the magic of a DSP (no matter which brand it is). However, our boards will help you improve the speaker equalization by boosting/dampening few specific frequencies your enclosure + driver combo may not be able to reproduce.

For more information about digital crossovers, see this section of our website.

Two popular methods of room correction are Room EQ Wizard (REW) and Dirac Live™. In this application note, we will compare and contrast these two methods.

• What is Room EQ Wizard?
• What is Dirac Live?
• The key differences
• A third option
• Wrapping up

What is Room EQ Wizard?[Top]

Room EQ Wizard (http://www.roomeqwizard.com) is a freeware acoustic measurement and analysis program. It runs on Windows, Mac, and Linux computers. In addition to its comprehensive acoustic analysis capabilities, it has the ability to generate parametric equalizer filters to correct a measured acoustic response. You can either correct towards a flat response or to a specified "target". These filters can be written to a file and then loaded into the PEQ block of a compatible miniDSP plugin.

Here, for example, is a measurement of a woofer (top plot in darker blue), the auto-generated correction filter (lowest plot in green), and the corrected response (middle plot in lighter blue):

In this example, the target response is flat (from 80 Hz to 1000 Hz). For additional information on how to use the Auto EQ feature, see the app note Auto-EQ tuning with REW. (The example above is a bit atypical - we have chosen it to make it easier to see the three plots mentioned.)

What is Dirac Live?[Top]

Dirac Live™ is a room correction algorithm developed by Dirac Research AB of Sweden. It uses a very powerful algorithm to analyze a set of nine measurements taken around the listening area and to generate a correction filter. The correction is implemented using what Dirac Live call mixed-phase filters. (More on that below.)

Dirac Live Calibration Tool corrects the measured in-room amplitude response to a specified "target curve." Usually, it has a rise in the bass and a slight fall in the treble. This image shows the uncorrected response of left and right speaker in blue, and the corrected response in green (the plots are an average of nine measurements):

Dirac Live Calibration Tool also corrects the time domain response of the loudspeaker. The next diagram illustrates a "perfect" impulse response of a loudspeaker, the measured impulse response, and the impulse response after correction with Dirac Live. This impulse response correction enhances stereo imaging and realism.

miniDSP has several products that incorporate the Dirac Live algorithm, including the stereo Dirac Series, the multichannel DDRC-88A and nanoAVR DL, and the forthcoming compact miniDSP 2x4 DL.

The key differences[Top]

• REW uses IIR filters, while Dirac Live uses mixed-phase filtering - in effect, a combination of IIR and FIR filters. FIR filters are more powerful than IIR filters, but more expensive to implement. See the app note FIR vs IIR filtering for information on the differences between these two different types of digital filter.

• Dirac Live Calibration Tool uses nine different measurements around the listening area and searches for optimizations that it can perform across the whole listening area. With REW, you can take multiple measurements and average them, but it is a more manual process and much less sophisticated.

• Dirac Live Calibration Tool does not require that you learn how to use any additional software. Everything is built into the tool. It provides a single solution for measuring and correcting both your speakers and your room. REW is a more hands-on tool, for those who like to learn and experiment.

• Dirac Live is a proprietary algorithm that miniDSP licenses from Dirac Research. While the cost of the licensing is built into the price of each Dirac Live-capable hardware unit, it is a significant part of the cost. REW is completely free. We are very happy to be able to collaborate with both groups to offer different solutions to our customers.

A third option[Top]

While we are often asked the question about the differences between REW and Dirac Live, there is also a third option: our OpenDRC series of processors. These processors contains onboard IIR (crossover and parametric EQ) filters as well as banks of FIR filters. They are "open" because they simply accept the coefficients for the FIR filters, which must be produced by other (non-miniDSP) design software.

With this option, you have the full power of combined IIR and FIR filtering available - but you need to learn how to use it! Designing FIR filters is not straightforward, but there are a number of software programs available that you can use, both free and commercial. We have put together a list to help you get started at FIR Filter Tools.

Wrapping up[Top]

That's it for this app note! Have fun, and regardless of which room correction solution you choose, please share your experience with other community members in our forum.

Filter Hose is a unique FIR (finite impulse response) filter creator software written and maintained by 3rd party developer HXAudio Lab. It enables the user to edit time domain data to construct an FIR filter very quickly, and allows them to experiment with different filter properties (tap length, windowing, latency, etc) very easily. To speed up the FIR creation, Filter Hose provides several frequency domain target responses.

This tutorial will provide a walk-through of filter creation using Filter Hose to linearize a loudspeaker frequency and phase response. Data will be exported to MiniDSP OpenDRC.

NOTE: Being a 3rd party software, all support questions should be asked directly to HXAudio lab or in the subsection of the miniDSP forum.

Also note a great app note is available here on the topic of "A meaninful loudspeaker phase response".

1.Loudspeaker Measurement

Although a loudspeaker can be measured in-situ, the impulse response will contain room reflections. This measurement can be used to create room correction filter. For venue/room tuning, average frequency responses at different audience areas are commonly used (no valid phase information), hence we can use Filter Hose to provide a room correction filter using averaged frequency response or power response curve. Filter Hose also provides advance manual input to accommodate different kinds of frequency domain data (i.e frequency, magnitude, with/without phase), including accepting RTA data.

1.1 Ground plane measurement

Although on a lot of occasions FIR is generated to do room correction, it is recommended to create FIR for loudspeaker corrections only and perform proper acoustical treatments. The user may also be interested in experimenting with FIR creation for loudspeaker systems measured in the room/venue vs FIR that corrects the loudspeaker only. Let us discuss several measurement techniques to create loudspeaker corrections (with minimum room influence).

To do a proper loudspeaker correction, a common measurement condition that is simple but effective is an outdoor (or a large room) ground-plane measurement on a reflective surface (such as smooth concrete/asphalt). Please see the pictures below.

Left - Outdoor Ground Plane; Right - Indoor (Warehouse) Ground Plane

Ground plane measurement is one of the most effective ways to minimize room reflections or ground reflection in the measured impulse response.

1.2 Free-space measurement

Of course a regular way of measuring loudspeakers is to simply use stands as shown below (taken outdoor).

This measurement condition will include a ground reflection. The frequency response usefulness will be limited from the first reflection (usually from floor). The impulse response shown in the photo on the left can be seen below.

Please see the impulse response's energy time curve below (ETC), showing a reflection at 11ms, and direct sound arrival time at 4ms. The usable impulse response length is only 7ms, and the data only valid at 200Hz and up (after windowing out the reflection).

2.Gating/Windowing Impulse Response

To help clean up the measured impulse response, Filter Hose provided Hann 50% (also known as Tukey 50%) window and multi-zone time window with different length presets. The user can experiment with different types of windowing to clean up the measurement impulse response if it is imported without windowing. Please note that windowing only works when there is measured phase information. Hence this function is not usable for RTA data or averaged frequency response (power response) data.

3.Preparing Loudspeaker Data in Filter Hose

To create a proper FIR filter in Filter Hose, several steps are necessary. The most important step will be matching the sample rate of the measurement to the hardware/software that implements the FIR filter. If the measured data is taken with different target sample rate, Filter Hose Manual Input Data can perform sample rate conversion.

Preparing the Measured Data

An outdoor free-space measurement is loaded to the Filter Hose. The data contains a ground reflection.

The direct sound arrival is approximately at 3.9ms and reflections can be noted at around 11ms. First, we like to move the impulse response cyclically to the beginning, so the very early energy arrival is located at zero. In this sample, the first impulse response energy arrival comes at 4 samples prior to the peak of the impulse response.

Step-by-step:

1. Press Max to 0. This will put the peak of the impulse response at 0ms.
2. Insert a value of 4 (samples) under the manual shift and click the right arrow button. This will cyclically move the impulse 4 samples to the right.
3. Select MZTW medium - Set - Yes.
4. Click Advance - Smooth Input.
5. Click Advance - Smooth Input.

In this example, multi-zone time window medium length is used with two times smoothing (smoothing can be found under the advance menu). The impulse response is now fully prepared and the final result is shown below.

Selecting Target

Flat Magnitude and Phase = 0 is selected as target response.

In this example, the magnitude will be flatten from 80Hz to 17000Hz and the phase will be flatten from 250Hz to 35000Hz.

Tips: User is also able to open more than one window of Filter Hose. This can be useful to experiment with different FIR creation for a same input file.

4.Editing FIR Filter

To create a 512 tap linear FIR filter with 2ms latency, several steps need to be done. First, we need to calculate how many samples are 2ms in a 96000Hz sample rate. This can be done by multiplying the targeted latency with sample rate.

Targeted latency is 2ms, fs is the sample rate which is 96000Hz. The result will be 192 samples.

Step-by-step:

1. Select Hann 50% window. The user can experiment with different windowing to clean up the impulse response's tail of the filter.
2. Input a value of 512 in the tap length. This is the targeted FIR tap length.
3. Click C. This will center the peak of the FIR filter peak to the middle of 512 total tap, which is at sample 256.
4. Because we are targeting 2ms latency (at sample 192), and currently the peak is at 256, we need to cyclically shift the FIR filter 64 sample to the right. Input a value of 64 in shift tap and click right arrow button.
5. Click Set.

The FIR filter final result is shown below.

To export the filter, right click on the time domain graph and click Export text (MiniDSP) and Export text (csv).

5.Predicted Result

To correctly see the predicted result, it is recommended that the filter be saved as a csv temporarily.

Step-by-Step:

1. Load the original measurement (back to step 1), and use only the multi-time window (medium) to clean up.
2. In step 2, select Load from *.csv and select the filter csv file.
3. In step 3, select Rectangular window, input a value of 512 in tap length, click C, input a value of 64 and click the right arrow button. Click Set. This should not alter the loaded FIR filter impulse response at all.
4. Click step 4, convolve input and filter.

The predicted result is shown in the picture below.

6.Importing the Filter to MiniDSP OpenDRC Plug-in

Open the exported .txt file using Notepad or Excel or some other similar program. Select all and copy to clipboard (Ctrl-C).

Open MiniDSP OpenDRC, and click on FIR channel to open the FIR filter import window.

Click Manual Mode, and then select all text in the left table and hit delete on the keyboard. This will clear up all coefficients that was previously loaded. Paste the new coefficient from the clipboard.

Click Process to import the FIR filter and Voila!

Do keep us updated of your progress experiences on the forum so you can share with the HXAudiolabs team!

rePhase is a Windows-based freeware program written by Thomas (aka "pos"), a long time miniDSP community member. rePhase generates finite impulse response (FIR) filters that "reverse" the phase shifts introduced by a loudspeaker crossover. rePhase can also generate linear-phase crossovers. With the aid of a real-time FIR filtering engine or "convolver" such as miniDSP's OpenDRC or miniSHARC, the result is a linear-phase loudspeaker system.

• What you will need
• Basics
• Functional overview
• Example 1: correcting loudspeaker phase
• Example 2: a linear-phase crossover
• Usage notes
• More info

What you will need [Top]

1. The rePhase program. rePhase can be downloaded from SourceForge here: http://sourceforge.net/projects/rephase/.

2. A convolver/filtering engine:

   • An OpenDRC-DI with stereo digital input and output, and the OpenDRC 2x2 plugin
   • An OpenDRC-AN with stereo balanced analog input and output, and the OpenDRC 2x2 plugin
   • An OpenDRC-DA with stereo digital input and balanced analog outputs, and the OpenDRC 2x2 plugin
   • An OpenDRC-DA8 with a stereo digital input and 8 unbalanced analog outputs, and the miniSHARC 4x8 plugin
   • A miniSHARC kit and the miniSHARC 4x8 plugin for up to 4 input channels and 8 output channels (requires implementation of your own I/O via I2S)
   • A miniDSP 2x4 HD (or its kit version)  for a for 2 x IN, 4 x OUT configuration running the 2x4HD1 plugin.

3. Ability to run acoustic measurements. You will need a measurement program such as Room EQ Wizard (REW) and measurement hardware such as the UMIK-1 and a microphone stand.

Basics [Top]

To effectively use this app note, you will need to be familiar with some basics of acoustic measurement, loudspeaker design and DSP. Please see the following app notes for background information:

• FIR vs IIR filtering
• UMIK-1 setup with REW
• Loudspeaker measurement with UMIK-1 and REW
• Time alignment basics
• Building a 2way crossover

Functional overview [Top]

rePhase can be used to:

1. Correct/linearize the phase of conventional loudspeaker crossovers—whether digital or analog

2. Implement a linear-phase loudspeaker crossover "from scratch"

rePhase accomplishes this by generating FIR filters that you load into an OpenDRC or miniSHARC. While it is possible to use rePhase to correct for room issues, it is intended to be a loudspeaker design tool. In this app note we will give an example of each of the two loudspeaker design scenarios above.

The rePhase user interface allows you to import a measurement from your measurement program and then manipulate the amplitude and phase of the correction filter. The interface is quite intuitive and includes a graph of the predicted response and a large number of controls, many accessed via tabs:

The phase and amplitude correction features are accessed from the main set of tabs:

We will cover each of these tabs in the examples below. The key point in understanding how rePhase works is that it allows you to adjust the amplitude and phase responses of the filter independently. Examine the following figure, in which amplitude is shown as the solid blue line and phase as the dotted blue line:

In this figure, there are three filters applied at different frequencies, as follows:

• 100 Hz: a minimum-phase notch in the amplitude response. In addition to the notch in the amplitude, there is a wiggle in the phase response around the center frequency. This is typical of a normal parametric equalizer (digital IIR or analog).
• 1 kHz: a linear-phase notch in the amplitude response. Note that even though the amplitude changes, the phase is completely unchanged around the center frequency.
• 10 kHz: a phase shift generated by rePhase, but with no change in the amplitude response. This enables rePhase to correct or "undo" phase shifts caused by typical crossover filters.

Example 1: correcting loudspeaker phase [Top]

In this example, we will use rePhase to linearize the phase of a loudspeaker crossover. This technique can be used regardless of whether the crossover is passive (inductors and capacitors in the loudspeaker box), an analog active crossover, or a DSP active crossover.

The example speaker has a fourth-order (24 dB/octave) Linkwitz-Riley crossover at 3 kHz and is a sealed box with a rolloff at 80Hz. First, we measure the loudspeaker's response and then import the measurement into rePhase. To apply the phase correction filters in rePhase, go to the Filters Linearizationtab and add correction filters for the crossover and for the 80 Hz rolloff:

In the figure below is the original response in red, with amplitude shown as the solid line and phase as the dotted line. The predicted response is in blue. As you can see, the amplitude is unchanged but the phase is flattened out (because phase "wraps" every 360 degrees, the corrected phase at -360 is equivalent to it being zero):

To generate the FIR filter coefficients for loading into an OpenDRC, set the Impulse Settings as shown here and press Generate:

Typically, the same filter will be used on both channels, so load the file into both FIR - Channel 1 and FIR - Channel 2 of the OpenDRC 2x2 plugin.

You can now measure the speaker again and check the result. Here is the measured phase of the example speaker, with the phase of the uncorrected speaker in red, and the phase after the correction filter has been applied in blue:

In general, a loudspeaker crossover may not always be quite so easy to linearize. (Also, rePhase only directly supports linearization of Linkwitz-Riley crossovers.) For arbitrary phase corrections, use the Paragraphic Phase EQ tab, which has a set of "sliders" that adjust the phase up or down in a bell-shaped curve. It's like a graphic equalizer, but for phase! Unlike a graphic equalizer, the Q (sharpness) and center frequency of each bell can be changed in addition to the amount of phase shift. The result from using the paragraphic phase EQ for additional phase correction at low frequencies is shown in the graph above in green.

The effect of the phase correction can be observed in the time domain in various ways. For example, here are the impulse responses of the loudspeaker before (in red) and after phase correction (in blue):

Here is a 2 kHz square wave as seen by the microphone, before phase correction:

And after phase correction:

Example 2: a linear-phase crossover [Top]

rePhase can also be used to generate linear-phase crossover filters. As an example, we re-did the above speaker with a linear-phase two-way crossover. One OpenDRC per channel can be used to implement a two-way crossover, or the miniSHARC for up to a stereo four-way crossover.

The drivers must first be measured without any crossover filters in place. You should prepare the measurements for use in rePhase by applying some smoothing or gating, as this will make it easier to make amplitude corrections. For each driver, you will then:

1. Export the measurement from the measurement program (in REW, go to File→Export→Measurement as Text)
2. Import the measurement file into rePhase
3. Use the Paragraphic Gain EQ tab with minimum-phase filters to flatten the amplitude response
4. Use the Filters Linearization tab to correct box rolloff phase
5. Use the Paragraphic Phase EQ tab for any remaining phase correction
6. Use the Linear Phase Filters tab to create the desired crossover filter
7. Export the impulse response from rePhase into a file
8. Load the impulse response file into the OpenDRC 2x2 or miniSHARC 4x8 plugin
9. Measure the driver again to check that the expected result is obtained

That may seem like a lot, but it's straightforward once you understand how rePhase works! Here's the woofer measurement of the example speaker loaded into rePhase shown in red, and the predicted response after steps 1 to 5 above in blue:

The next step is to select a crossover filter. rePhase has a wide range of filter types, different to what you will be used to for IIR crossovers. Here is the setting used for the woofer in the example speaker, an overlapping crossover with a steep (96 dB/octave) cut-off:

To export the impulse response, use settings like those shown in the following screenshot. It's recommended that the "centering" options be set to "middle" and "int" so that the peak of the impulse response is at a known location (see note 4 below):

With the woofer done, you will need to repeat the above for the tweeter (and additional drivers if implementing a three-way or four-way). Each driver should be done in a separate rePhase project.

The impulse responses can now be loaded into the OpenDRC or miniSHARC and the drivers re-measured individually and then in combination. Time alignment may be needed to adjust for acoustic offsets of the drivers and differences in the location of the impulse response peak.

The following graph shows the measured amplitude responses of the example speaker, where the individual responses of the woofer and tweeter are shown in blue and green, and the combined response in mauve (see notes 5 and 6 below).

This is the measured phase response of the example speaker:

Usage notes [Top]

rePhase is designed to allow the operator full control over the correction and filtering implemented. Here are some additional notes on usage:

1. While you can put in almost anything you like for the correction or crossover filters, the requested filter may not be realizable depending on the length of FIR available. After pressing Generate, rePhase will show the actual response that can be achieved in red.

2. The generated FIR filters are not causal — that is, the impulse response starts "ahead of time." In practice, this means delaying the impulse response peak, thus effectively delaying the output signal. For stereo audio playback, this is usually not a problem, but it may be a problem when synchronization with video is needed (home theater), for music recording or in live performance.

3. The part of the FIR impulse response before the peak is sometimes referred to as "pre-ringing." While debates rage about how audible this is and under what circumstances, it is something to be aware of when designing with FIR filters.

4. When implementing a multi-way speaker, the impulse response peak of each driver must be aligned in time. That is why the "middle int" centering option was suggested above, so the peak is in a predictable location. If the FIR filters for each driver are of different lengths, then time delay will be needed to re-align the acoustic signals from the drivers.

5. Any digital filter or speaker design program is a tool to aid you in creating a good speaker, but driver selection and cabinet design are also very important. For example, the measurements above show an issue in the 7-8 kHz region that should not be "fixed" digitally. Rather, an investigation is needed into a better design of the cabinet and/or a different tweeter.

6. The on-axis measured response is, of course, taken at only one point in space. Because a loudspeaker uses multiple drivers, there will be interaction between them so that the response will be different at other points in space. The off-axis responses, both horizontally and vertically, should also be measured and optimized to meet your design goals.

Support Free Software

A software of the complexity like rePhase is by no mean a 5min project on a Friday afternoon! Thomas keeps perfecting this wonderful software and all this takes time. So if you like his software, make sure to support it! Here is a paypal donation link found on his soundforge page as well. https://goo.gl/v44mmD

More Info [Top]

If rePhase is not the most suitable for your FIR filtering project, please see these other app notes for more options:

• List of FIR Filter Tools
• Python Open Room Correction
• AcourateDRC

You have an OpenDRC or miniSHARC kit and now wondering how to build your own FIR filters. Over the years that we've been looking into FIR, there are few tools we've found that are worth sharing with you. Here is a non exaustive list. Please feel free to email us to share more info!

rePhase (Freeware)

Designed by Thomas (Pos), rePhase is a neat little freeware tool for building your FIR  filters. Please make sure to take the time to support this great application. 
A complete step by step guide is available on miniDSP website at the following link. 

Filter Hose (Commercial)

Check out the software by HXAudioLab. A well known platform for all FIR audiophiles. Make sure to check out the application note on this website. 

Ultimate EQ (commercial)

Ultimate EQ is the work of art from Bohdan Raczynski. This very powerful software now has support for generating coefficients for the OpenDRC and miniSHARC series. Make sure to check out our application notes. 

FIR Designer

FIR Designer (OsX and Windows) is a comprehensive tool for creating FIR filters for loudspeakers. It provides both manual and automatic methods for manipulating magnitude and phase, and can export filters suitable for many processors including miniDSP, Powersoft & BSS. The regular work-flow starts with importing a loudspeaker measurement, then designing a FIR filter to achieve a desired overall target response. Alternatively the tool has a Direct Design mode for creating FIR filters without reference to a measurement.

Align2 (Freeware)

Align2 is a free audio softwre designed by JL Ohl to calculate loudspeaker correction for OpenDRC hardware or for convolution softwares.
It was first designed as a GUI for DRC (D. Sbragion) or PORC (M. Green) but evolved with many added functions.

Iowa Hills Filter design tools (Freeware)

This free FIR filter design program uses the Parks McClellan algorithm and Fourier Transform (windows) method to synthesize filters. It is capable of synthesizing a wide variety of filter types in the form of linear phase, minimum phase, or an IIR type phase response. Possible filter types include the Raised Cosine, Bessel, Inverse Chebyshev, and others. Hilbert filters, Differentiators, and other specialized band pass filters are available from our Hilbert Filters program. Make sure to check out the wealth of information on the website of Iowa Hills!

ScopeFIR (Commercial)

ScopeFIR is an advance tool for building all your FIR filters. A well known software recommended for any hardcore designer. Note that this software requires a license.

Octave (Open Source)

Here is an open source solution using a Math package. Octave is a well known GNU (open source) package widely used as an alternative to Matlab. Here is a simple step by step for building a FIR filter under Octave.  

T-Filter (Free online tool)

An online FIR filter designer tool. A very basic GUI for a quick to use application. Check it out here.

MatLab (Commercial package)

Last but not least, Matlab and its processing toolbox is certainly very effective at building FIR filters and simulate their response. Here is a summary from the help file. 

Got a software that you could recommend? Please get in touch with us so we can advertise it in this section. 

PORC is intended to be a free command line tool and is an Open Source project ported by Mason A. Green and based based on the work of Dr. Balázs Ban hosted at the following page: http://home.mit.bme.hu/~bank/parfilt/. This package allows one to specify a target curve and generate the corresponding loudspeaker-room correction filters for the OpenDRC platforms. The filters may easily be imported into OpenDRC or other convolution engines. More details about Dr. Bank's parallel filter can be found in his papers at the following links.

Balazs Bank, "Perceptually Motivated Audio Equalization Using Fixed-Pole Parallel 
Second-Order Filters", IEEE Signal Processing Letters, 2008. 

Balazs Bank, "Direct Design of Parallel Second-order Filters for Instrument Body Modeling", 
International Computer Music Conference, Copenhagen, Denmark, Aug. 2007. 

The following sections are extracted from the online Open Source repository setup by Mason on GitHUB. Please refer to webpage for the latest source code.

Required software dependencies to be installed on your PC

IMPORTANT note: This is a command line tool and therefore dedicated to a more advanced crowd. Matplotlib will produce very nice graphs but will require some computer knowledge and basic command line skills. If you want a plug&play solution and are not too computer friendly, this tool is not the easiest way to get your system up and running. PORC has been tested successfully on both Linux and Windows 7 with Python 2.7. Linux depenency install is fairly straightforward. Windows install packages are available for all dependencies.

1) Python 2.7 
2) Scientific Python: SciPy, Numpy, & Matplotlib 
3) Scikits.audiolab 
4) libsndfile
5) Sox for converting files to correct format
6) Room EQ Wizard (REW) 

Measurement:

One needs to measure the log-frequency impulse response of your speakers with a calibrated Electret Measurement Microphone (e.g. Dayton Audio EMM-6 / ECM8000 or other measurement microphones). Software such as Room EQ Wizard (REW) may be used for the purpose of measuring the impulse response. We recommend that you have a look at REW's website for more info. http://www.hometheatershack.com/roomeq/

Usage

porc.py [-h] [-t FILE] [-n NTAPS] I F

python porc.py -t tact30f.txt -n 6148 l48.wav leq48.wav 

Use the -h flag for help!

Target Response

The default target curve for PORC is flat. Included in the data directory are a number of target curves. Experiment to suit your listening preferences (I prefer tact30f.txt, bk-48.txt).

For further reference, the B&K House Curve is a good place to start. Read "Relevant loudspeaker tests in studios in Hi-Fi dealers' demo rooms in the home etc.," Figure 5: http://www.bksv.com/doc/17-197.pdf

OpenDRC Convolution

To convert the end results to a .bin file that can easily be loaded to the OpenDRC, you will need to use sox to convert the output .wav file to a raw 32 bit IEEE floating point mono. The process must be repeated for the left & right channels.

sox leq48.wav -t f32 leq48.bin sox req48.wav -t f32 req48.bin 

You can then take the output file and load it to the OpenDRC using the "Load Bin" feature in the FIR dialog box.

Support Community Work & Credits

- Algorithms are based on the work of Dr. Balázs Bank and the numerous papers he published at AES/IEEE journals. Dr Bank's personal website includes a wealth a knowledge certainly worth a read. Thanks again to Dr Bank for allowing us to publish his work under this website!

- Our community is also very fortunate to include developers like Mason A. Green working on some Open Source projects. All credits goes to Mason for starting up this project and building the project. Please take some time to give a small donation to this project so Mason can eventually keep developing creative software! All profits are linked to his Paypal account.

---

 

The miniDSP community is full of passionate members who love audio and most importantly are willing to share their knowledge! We want to make sure we do our best to make sure we spread the word about their work.
Introducing the Active Crossover Designer (ACD) by Charlie Laub. Based on a excel file loaded with all the equations required to evaluate the optimal filter for your crossver, the ACD tools is a strongly recommended tool for all speaker designers.

NOTE: This project is free for personal DIY Speaker Design use. Commercial use is prohibited without prior authorization of the author. Please email This email address is being protected from spambots. You need JavaScript enabled to view it. for commercial inquiries.

Introduction

The Active Crossover Designer is an active crossover design tool available as Excel spreadsheets and designed for the serious hobbyist. The user can set up the tools from the set of supplied driver and system templates, import FRD files, determine driver offsets for precise phase determinations, and develop crossovers by following the example provided in a tutorial. There is also a comprehensive technical manual. Once the crossover has been designed, rapid implementation is possible using one of the MiniDSP digital crossover products with the advanced biquad programming feature. Analog transfer functions for all filters are also available in the tools.

The conceptual approach for these tools has been to create an "open" platform using commonly available spreadsheet tools that many DIY hobbyists are familiar with. Because all of the calculations are done using spreadsheet formulas that are all visible and accessible, the tools are fully extensible. The user can change any and everything, or add on new functionality, plots, and so on. Some examples of useful add-ons for the CORE TOOLS are provided in the EXTENSIONS section. Currently, only an Excel version is available, however, an OpenOffice Calc version has been developed in parallel and the release is coming (eventually). ACD is written for Excel version 2003 and later, and requires the Analysis Toolpak.

Things you need to get started

• One miniDSP DSP hardware platform (kit/2x4/balanced/2x8/8x8)
• One miniDSP plug-in with the Advanced feature for custom biquad.
• The lates version of ACD tools spreadsheet
• 1 x measurement microphone + sound card to perform your audio measurements and validate your simulations
• 1 x PC/Mac running the latest version of Room EQ Wizard (REW) V5 available for download here.
• A basic understanding of the biquad filter programming as already highlighted in this application note.

Screenshots

ACD tools come up with a fully featured start up guide and user manual. The following pictures are just samples of the type of simulations you can expect out of ACD tools.

Credits

The whole miniDSP team would like to thank Charlie Laub, Author of ACD Tools, for designing such a useful tool! All credits goes to Charlie for putting together this nice tool. If you want to thank him yourself, remember to fire a small email to him or share your experience on our Forum.

Flexibility really matters when it comes to innovative speaker design. As DIYers and speaker designers ourselves, we came to realize that miniDSP platforms could be used for both basic and advanced digital signal processing applications. While textbook filter implementation like Butterworth, Linkwitz Riley, and Bessel work in 80% of configurations, some speaker designers may want to investigate novel IIR filter implementation like:

- Filters currently not implemented inthe basic mode

- Cascaded filters

- Custom filters developed on 3rd party software

The following section will teach you how to use our advance programming.

DISCLAIMER: The following section assumes that you have some basic understanding of filtering and Digital Signal Processing. Testing the accuracy of a filter before loading it to the miniDSP must be performed ahead. Once loaded inside the miniDSP, you should first confirm the overall transfer function to confirm that your biquad computations are correct. miniDSP cannot be liable of incorrect/unstable plug-in biquad settings being loaded into the platform.

Introducing miniDSP Advanced Biquad Filter Programming

miniDSP filtering is based on 2nd order linear recursive filtering, also called a Biquad digital filter. A biquad contains two poles and two zeroes with a transfer function expressed as follow in the Z domain.

All filters of miniDSP (PEQ/Shelf/Graphic EQ/Low Pass/High Pass) are all implemented using the 5 coefficients of the a 2nd order filter. (a1/a2/b0/b1/b2). Note that a1 and a2 need to have sign inverted (times -1) due to the DSP implementation.

Biquad programming on miniDSP plug-ins is allowed on all biquad objects of the DSP structures. There are currently 2 places were custom biquad can be implemented:

• PEQ filters on input/outputs: With 6 biquad on each input and 6 biquad on each outputs

• Crossovers: With 8 biquads on each output, cascaded filters can be created by stacking 2nd order filters in series.

Adding them up together will give you all the flexibility you need to create your custom processing.

Filter ideas...

Advanced Biquad programming trully opens the door to a new range of filtering applications such as:

• Linkwitz transform
• All pass filters
• Cascaded filters (e.g. two low pass filters in series)
• Filters of up to 16th order (96dB/oct) when cascading 8 biquads.. Still not enough? How about adding the remaining 6 biquad filters from the PEQ outputs and building a ( 8 + 6) * 2nd order = 28th order filter with 168dB/oct attenuation as highlighted in this community tutorial.

Where to start?

• Download the Biquad filter spreadsheet
• Read about biquad filters and how to insure they are stable
• Load a simple biquad and test it with a sound card  first to make sure it is correct
• Start simple, get more creative over time!

Ready to build your own biquad filters?
Check out the miniDSP series (IIR) on miniDSP's webstore.

In this application note, we show you how to use the miniDSP custom biquad programming feature to implement a Linkwitz transform.

What's a Linkwitz transform?

A loudspeaker driver in a sealed enclosure has a mathematical description that determines its low-frequency response. At some point, the response of the speaker starts rolling off and reduces in level at 12 dB per octave. (Each time the frequency halves, the output level drops by 12 dB.) The shape of this curve can be characterized by two parameters: tuning frequency F, and quality factor Q.

A Linkwitz transform is a mathematical operation that changes the effective F and Q to different values. Typically, this is used to lower F to get more low bass output, or to lower the Q to make the box behave like a larger box. You can implement this transform in any miniDSP plugin that supports the custom biquad programming feature, such as the Advanced 2-way crossover, Advanced 4-way crossover, the 4x10 and 10x10 plugins, and so on. Check the miniDSP plugins page to find suitable plugins.

Note that the Linkwitz transform only works properly for sealed boxes. It's not suitable for ported boxes or for open-baffle speakers.

1. Download the biquad programing spreadsheet

Download the community-contributed biquad programming spreadsheet from the following link, and select the "LT" tab.

• Biquad calculation spreadsheet

2. Determine your speaker's F and Q

If you have built the speakers yourself, you may already know these parameters from your box design phase. Otherwise, you will need to determine them by performing a nearfield measurement of the woofer - that is, with the microphone placed close to the cone, or about 5 cm (2"). The curves below illustrate a set of curves with a F of 80 Hz and Q of 0.5, 0.7, 1.0, and 1.4 (green, red, purple, black).

With a measurement taken, you will need to guess at F and Q by comparing the measured response with the curves above. Enter your guesses at F and Q into the spreadsheet and compare the curve displayed in blue with your measurement. If needed, adjust F and Q in the spreadsheet until the curve displayed in blue is reasonably close.

It's not super-critical to get F and Q exactly right the first time around. After applying the transform (steps 3 to 5 below), you can adjust and try again if the result is not what you want.

3. Calculate the biquad parameters

In the spreadsheet, enter the box F and Q as f(0) and Q(0), and the desired F and Q as f(p) and Q(p). For example:

The spreadsheet will display the original response, the desired or target response, and the equalization curve:

You will note at this point that you will need to be realistic about the target F and Q as typically the transform will result in significant boost at low frequencies. Set the target F and Q so that the amount of boost is compatible with available amplifier power and the woofer's excursion limits.

4. Enter the parameters

Scroll down on the spreadsheet to find the biquad coefficients section:

to your measurement

In the miniDSP plugin, set a filter to Advanced mode and enter the five parameters a1, a2, b0, b1, and b2, as shown below. (Note that the order in the spreadsheet is different than the order shown by default in the miniDSP plugin. Just copy each number and paste into the correct location.) Then press the Process button. The frequency response graph will update to show the transform.

5. Verify your results

Run the nearfield measurement of the woofer again. If all is well, you will see the target response! If the response is not what is desired, fine-tune your F and Q settings and try again. Here is the before and after response of the example speaker, as measured with Room EQ Wizard:

Have fun, and don't forget to ask on the miniDSP forum if you have questions!

In this application note, we show you how to use Room EQ Wizard (REW) and its integration with miniDSP to equalize your subwoofer.

• What you will need
• 1. Make the initial measurement
• 2. Calculate the correction filters in REW
• 3. Load the correction filters
• 4. Add a room curve
• Wrapping up

What you will need[Top]

• A miniDSP plugin that supports the "REW integration" feature. (Almost all current plugins support this feature. Check the User Manual for the relevant plugin or product for the capability to do Advanced biquad programming.)
• Room EQ Wizard (REW), which is a free download for Windows, Mac OS X, and Linux.
• A calibrated measurement microphone, for which we recommend the UMIK-1.

1. Make the initial measurement[Top]

• In the plugin interface, disable all equalization.
• Usually, you will also want to disable all crossover filters. This is so that the equalization can be applied independently of the crossover.
• Mute all output channels except the one that you want to measure.

Then, in REW, run a measurement sweep in the frequency range of interest. Ensure that your SPL is adequate to get a clean measurement.

2. Calculate the correction filters in REW[Top]

From the REW main screen, open the "EQ" tab.

Set the parameters as shown in the following screenshots. Firstly, set the Equalizer to "MiniDSP" for plugins running at 48 kHz, and "MiniDSP-96k" for plugins running at 96 kHz.

In Target Settings, set Speaker Type to "None", LF Rise Slope to 0.0, and HF Fall Slope to 0.0. (You can experiment with different settings later.) Click on "Set Target Level" to set the target level. This is just an initial value - you will want to adjust it manually later, after performing your initial filter generation.

Now set the parameters for generating the correction filters. Shown below are a typical set of values for equalization in the subwoofer range. You will want to experiment with different values to get the best set of filters.

Then click on "Match Response to Target." REW will generate filters to optimize the response. Here is an example with the measured response in dark green, the corrected response in light green, and the correction filters in light blue:

In the above example, smoothing was set to 1/6th octave. REW will generate different filters depending on the amount of smoothing, so you should experiment with the amount of smoothing as well. The example below shows the result with 1/48th octave smoothing. It is a much "tighter" EQ but also requires more filters and may not work as well over a larger listening area.

You can see the filter settings calculated by REW by clicking on the "EQ Filters" button near the top of the screen:

3. Load the correction filters[Top]

Once you have a correction that looks satisfactory, click on "Save filter coefficients to file" and save the filter settings as a text file (.txt).

Then return to the miniDSP plugin and open the PEQ block that you want to put the filters in. Set it to "Advanced" mode and click on "Import REW File" (sometimes it is just "Import"). Note that some plugin may not support the max number of PEQ that is being used by REW. Check the datasheet of the plugin for more info. Select the file that you just saved, and the plugin will then display the correction filter. Click "PROCESS" to display your imported biquad.

Confirm your result by running your measurement sweep again - it should be much smoother than before! You can then repeat the above procedure in different configurations (*) but with different settings for REW's Smoothing and Filter Tasks. That way you can compare the audible effect of the settings.

(*) Many miniDSP plugins support four different configurations. Different configurations can be selected from within the plugin or with a remote control.

4. Add a room curve[Top]

Depending on what exactly you are equalizing, you may want to add a "room curve". This boosts the bass and also shelves down the treble slightly. The best way to do this is with low-shelf and high-shelf filters on the input channels of the miniDSP. (In the nanoAVR HD, it will have to be done in the output channels.) Here is an example of a house curve:

Wrapping up[Top]

Now you can re-enable your crossover filters, and do some additional measurements to fine-tune your integration. If the results are not satisfactory, experiment with different settings: the target level, the amount of smoothing, and the frequency range of correction.

That's it for this app note! Have fun, and please share your results in our forum.

In this two part article by industry veteran Pat Brown from Synaudcon, we get a full explanation on the benefits and limitations of FIR filters for speaker design.

• Part one of the article goes through the basics of FIR. A very good summary of how/why/when to use FIR filters.
• In Part two, Pat makes use of rephase along with REW study to showcase a real life scenario.

Figure 1 – At the top we see a 60 tap FIR HPF. Note the lack of sharpness due to the small tap size.
Next, adding more taps (more samples or a longer time length) creates a sharper filter, and then at the bottom,
we have a 6000 tap filter that is 136 ms in time duration, with latency of 1/2 of the length (68 ms)

Figure 2 – The polar response at 1 kHz (1/3-oct, 5-degree angular resolution)
using the LR24 crossover (A) and the linear phase brickwall crossover (B)

A must read for all speaker designers out there thinking about using FIR and wanting to learn more.

For more info on Synaudcon and their great learning/training sessions, make sure to check out SynaudCon website.

This application note will show you how to build a stereo three-way or four-way digital crossover. If you are new to the concept of multi-way DSP loudspeakers, please read the overview app note Digital Crossovers Basics first.

What you will need

The miniDSP product line includes a number of options that support q 3/4-way crossover directly, summarized in Selection Guide : 3/4-way digital crossover:

• miniDSP 2x8 (kit) or 4x10 Hd (boxed) for analog inputs and outputs. If building from the kit, add a DIGI-FP for two digital input/output channels and a VOL-FP for volume and remote control.

• nanoDIGI 2x8 K (kit) or nanoDIGI 2x8 B (boxed) for an all-digital input/output solution.

• miniSHARC kit for an advanced solution with FIR and IIR filtering, in which you implement your analog or digital I/O via I2S.

• OpenDRC-DA8 for a Digital IN, Analog out with FIR and IIR filtering
• DDRC-88A + BM option for a room correction + crossover solution. Check out this app note for more details.

All of the above options can also be used to generate a subwoofer output signal summed from the left and right input signals, for a two-way-plus-sub or 3-way-plus-sub configuration.

1. Configure hardware

The specifics of your hardware configuration depend on the nature of your project, the amplification that you have, and the miniDSP option that you chose above. Here is an example of a 3-way-plus-sub configuration.

The example shows a 6-channel amplifier for simplicity, but you can also use three stereo amplifiers, one for each pair of drivers (tweeter, midrange, woofer). Changing to a 3-way or a 4-way speaker simply means leaving off the sub and/or adding another pair of drivers and amplifier channels.

2. Configure the plugin

The plugins for the above options are all similar. Here we will use the nanoDIGI 2x8 plugin as an example. The interface has three tabs labelled Input, Output, and Routing.

1. On the Input tab, rename the two inputs to Left and Right:

   

2. On the Output tab, rename the eight channels as shown here:

   

3. Also on the Output tab, open the Xover and PEQ blocks and link each left output channel to the corresponding right output channel. For example, here is the display on channel 1 (Tweet L) after linking it to channel 4 (Tweet R):

   

4. On the Routing tab, set which input channels go to which output channels:

   

3. Measure and correct the drivers

Once you have built the loudspeaker, you will need to measure each driver individually and use the PEQ block of its output channel to correct its response. Please refer to step 3 of the app note Building a 2way crossover for information on how to do this, as the procedure is the same except that instead of two drivers, you have three or four!

Note: Before doing any measurements, check that all PEQs and Xover filters are disabled. The subwoofer is not measured at this stage.

4. Add the crossover and fine-tune

In the Xover block of each output channel, set the crossover filters to the desired crossover point and slope. You can then measure the drivers in pairs (e.g. mid+tweet, woofer+mid) to check that the crossover has integrated the drivers correctly. See section 4 of Building a 2way crossover for further information.

Then, measure the completed speaker. For a 3-way or 4-way speaker, a microphone distance greater than the normal 1 meter may be needed for accuracy, so the measurement may need to be done outdoors.

5. Set up for listening

If the channel linking has been done as described above, then the crossover and EQ settings for both channels are already in sync. Save the configuration before proceeding.

You can now set the speakers up properly in their intended location in the room, and move the microphone to the listening position. Measurements made from now on are "in room" measurements.

6. Equalize and integrate the subwoofer

If using a subwoofer, measure, equalize, and integrate the subwoofer as described in Subwoofer integration with miniDSP.

Once everything is set up to measure well, the parametric EQ blocks on the input channels can be used to fine-tune the final in-room response. You can use this to tame additional resonances in the bass region, for example, or to provide a small amount of shelving boost in the bass "to taste."

Other options

Here are some other options for the hardware from miniDSP:

• 2 x miniDSP 2x4, with the advantage of small size and lower cost, together with the 4Way Advanced plug-in or 4Way GEQ plug-in. Current 2x4 platforms include: miniDSP kit, miniDSP 2x4 (in a box), miniDSP balanced kit or 2x4 in a box. Add a miniDIGI board in a combo kit if you require digital input. (See all kits.)

• miniDSP 8x8 (kit) or 10x10 Hd (boxed) for multi-channel processing with up to eight analog inputs and outputs. Outputs can be used for active crossover in any manner desired - for example, build a three-way left and right active speaker and a two-way center channel, and use the digital outputs for subs/bass management.

More info

These app notes will help you with your active speaker project.

• UMIK-1 setup with REW
• Loudspeaker measurement with UMIK-1 and REW
• Building a 2way crossover

Don't forget to ask on the miniDSP forum if you have questions!

In this application note, we show you how to use miniDSP to properly integrate a subwoofer into your system.

What's needed?

To integrate an existing pair of two-channel speakers, the miniDSP will act as the crossover between the main speakers and the subwoofer. You will need one of the following items of hardware:

• For a ready-to-go standalone solution, a miniDSP 2x4. Or for balanced inputs and outputs, the miniDSP Balanced 2x4.
• For a kit solution with choice of digital (coax or optical) or analog inputs, a miniDSP/DIGI Combo. Or for balanced inputs and outputs, the miniDSP/DIGI Balanced Combo.

You will also need:

• The miniDSP plugin 2 way Advanced 21.
• A measurement microphone, such as the UMIK-1, and the freeware Room EQ Wizard (REW) measurement program.

(See also "Other options" below.)

1. Hardware configuration

The diagram below shows the hardware configuration for a typical setup, assuming a stereo amplifier and a subwoofer with built-in amplification. Disable the subwoofer's internal crossover if possible, otherwise set the crossover to as high a frequency as possible (because the miniDSP will be doing the crossover instead).

2. Initialize the miniDSP

Double-click on the installed plugin (e.g. "MiniDSP-2waySUBAdv") to open it. If you receive a message about installing Adobe AIR, quit the plugin, allow Adobe AIR to install, and re-start the plugin.

Click on the Sync button (in green). At the dialog box, click on "Reset to defaults."

You will now see the block diagram:

First click on each crossover block and make sure that all crossover filters are disabled by turning on all the Bypass buttons. Then click on the System Settings tab and ensure that the Sub Output Mode is set to Mono Mode:

3. Equalize the subwoofer

To equalize the subwoofer, you will connect your audio interface to input channel 1 only, and do all equalization operations on Parametric EQ - Output 1.

1. Make sure that your main amplifier is turned off (or alternately, you can mute output channels 3 and 4 using the miniDSP plugin interface).
2. If you have a sealed subwoofer, you may want to start by using a Linkwitz transform to extend its bass response. See the application note Linkwitz transform - step by step. (Also see note below.)
3. Now, equalize the subwoofer for in-room response. The best way to do this is to use Room EQ Wizard and its Auto-EQ function. See the application note Auto-EQ tuning with REW for instructions on how to do this. (Remember: load the generated filters into output channel 1.)
4. Finally, once your EQ is all set up, click on the button "Copy to Output 2." This will ensure that the right audio channel is processed the same as the left. (Note that this is necessary, even though output channel 2 is not physically connected to anything.

Note: If loading an REW file when you have an existing filter such as a Linkwitx Transform, you will have to a. ensure that REW is limited in the number of filters it uses for auto-EQ and b. edit the REW file before loading to include the coefficents from the Linkwitz Tranform.

4. Equalize the mains (optional)

This step is optional. If you wish, you can apply equalization to the main loudspeakers using the Parametric EQ blocks on channels 3 and 4. For guidelines on how to do this, see the section "Measure and equalize the drivers" in the app note Building a 2way crossover. (You will need to turn off your subwoofers while doing this, or alternately, mute output channels 1 and 2 in the miniDSP plugin interface.)

5. Integrate the subwoofer

Click on the top crossover block in the miniDSP plugin interface.

1. For output channel 1, enable the low pass filter (turn off the Bypass button). As a starting point, set frequency to 80 Hz and the filter type to LR 24db/octave.
2. For output channel 3, enable the high pass filter (turn off the Bypass button). As a starting point, set frequency to 80 Hz and the filter type to LR 24db/octave.

The crossover graph should look like this:

This is a just a starting point! Now run a measurement sweep (with the microphone at the listening position) to see what the combined response of the left speaker and the subwoofer is. You will probably need to make some further adjustments:

• Adjust the gain of either channel 1 (the subwoofer) or channel 3 (the main speaker), so that the acoustic levels are matched. Often, a slightly higher level on the subwoofer is subjectively pleasing.
• Adjust the time delay of either the mains or the subwoofer. This is equivalent to the phase adjustment on many subwoofer amps.
• Adjust the frequency and slope of the high pass and low pass filters. Note that the filters do not have to be matched. If the mains are already rolling off, then a lower slope in the high pass filter may better match the low pass filter. And in the end, what counts is getting a smooth response.

Once you are satisfied, go to the lower crossover block and copy the settings just done across to channels 2 and 4.

And... that's it! You can also experiment with different subwoofer locations, and even with multiple subwoofers, to get the best response. Don't forget to ask on the miniDSP forum if you have questions.

Other options

Instead of the hardware choices given above, you can also use any of the miniDSP hardware units capable of more output channels. This applies if you are integrating a subwoofer into an active loudspeaker, but also for integration of a regular two-channel setup with a subwoofer. With these options, you will also be able to independently equalize more than one subwoofer channel:

• For two-channel input, a miniDSP 2x8 kit or miniDSP 4x10 Hd boxed unit, together with the 4x10 plugin.
• For multi-channel (i.e. 5.1 or 7.1) input, a miniDSP 8x8 kit or miniDSP 10x10 Hd boxed unit, together with the 10x10 plugin.
• For pure digital input and output, a nanoDIGI 2x8 K kit or nanoDIGI 2x8 B boxed unit, together with the nanoDIGI 2x8 plugin.

In this application note, we will show you how to design an active 2-way loudspeaker. This application note will use the 2x4 series. If you're looking for higher DSP power, please check out the 2x4HD application note instead.

What you will need

• A miniDSP. For a two-way speaker, a miniDSP kit, miniDSP 2x4, or a miniDSP Balanced 2x4 are suitable.
• A miniDSP plugin. For a 2-way loudspeaker, the 2-way Advanced plugin is most suitable.
• Ability to run acoustic measurements. You will need a measurement program such as Room EQ Wizard (REW), and measurement hardware, such as the UMIK-1.
• Four channels of amplification matching your speaker needs

This is the block diagram of the two-way plugin. A good approach is to use the output channel Parametric EQ to correct for the response of the individual drivers, and the input channel Parametric EQ for overall response shaping and taming room issues.

1. Select the speaker drivers

If you are starting from scratch, you will need to select the drivers for your speakers. There are literally hundreds of drivers available for DIY use at all price levels, so it's impossible to give specific recommendations here. For a small two-way loudspeaker, a 5" or 6.5" woofer and a 1" dome tweeter are common choices. Peruse the online forums to see what others are using and to ask for recommendations for your particular project.

If you are modifying an existing speaker from passive to active, then you have the drivers. In this case, you will most likely need to remove the internal crossover and add a second pair of binding posts.

2. Design the enclosure

If you are building your own box, you will need to design it. The most important factor is the internal volume, and if it's a ported box, the size and length of the port. Fortunately, there are a number of free programs that do the complex math for this based on the Thiele-Small parameters of the woofer. For example, a popular Excel-based program is Unibox.

3. Measure and equalize the drivers

Once you have built the box and mounted the drivers, you will need to measure the drivers one at a time. When performing acoustic measurements of a loudspeaker, it's important to try and minimize reflections. Position the speaker so the tweeter is half-way between floor and ceiling, as far away from walls as possible and angled at 45 degrees to the walls. Position the microphone level with the tweeter, and 1 meter (3' 3") away. For a typical 2-way speaker (tweeter and woofer close to each other), you can leave the microphone at the same position for woofer and tweeter.

You will need to use the output parametric EQ blocks to shape the response of each driver so that it is flat over its operating range. Use "Peak" type filters to flatten peaks (with negative gain so they create a notch) and "High-Shelf" and "Low-Shelf" type filters to straighten out the overall response.

Let's start with the woofer. Ideally, this measurement should be done outdoors, but you can still get good results with indoor measurement if you are careful. Note that at low frequencies there will be peaks and notches caused by the room. You should not correct for those at this point.

The figure below shows the before-and-after responses of a typical small woofer in a ported box, annotated with the key features that you can expect to see in such a measurement. In this example, the baffle step loss and the cone breakup peak are corrected for, but peaks due to the room are left untouched.

Then measure and equalize the tweeter, When performing a tweeter measurement, start the sweep at a frequency so as not to strain the tweeter (for example, start at 1 kHz, not 20 Hz). Here is an example tweeter measurement, before and after equalization:

4. Add the crossover and fine-tune

In the Crossover block, set a low pass filter on the woofer and a high pass filter on the tweeter. As a starting point, try using Butterworth (BW) 18 dB/octave or Linkwitz-Riley (LR) 24 dB/octave filters. You can then experiment with lower or higher slopes, from 6 dB/octave up to 48 dB/octave.

Then you can measure the response of the complete speaker! Use the output level controls to match the signal levels from the woofer and tweeter. If you have a dip at the crossover frequency, you may need to invert the phase of one driver. You will most likely need to fine-tune the crossover settings to get the smoothest response around the crossover frequency:

• Time align the drivers
• Move the filter corner frequency for one driver up or down a little
• Use an asymmetrical crossover, for example BW 18 dB/octave lowpass on the woofer and LR 24 dB/octave high pass on the tweeter
• Adjust the equalization of one driver or the other near the crossover point

This REW plot shows the response of the woofer and tweeter in our example speaker with crossover filters in place, and the combined response, all after crossover fine-tuning:

5. Set up for listening

Copy the settings for the channel you have just done to the other channel. Then save the configuration.

You can now set the speakers up properly in their intended location in the room, and move the microphone to the listening position. Measurements made from now on are "in room" measurements. You will notice that the peaks and dips up to a few hundred Hz will have moved - that is why they are not corrected in the measurement made at 1 meter.

The parametric EQ block in front of the crossover can now be used to adjust the final response as you desire. You can use this to tame resonances in the bass region, for example, or to provide a small amount of shelving boost in the bass "to taste."

Wrapping up

Be warned that you may not get it quite right the first time. It's a learning curve, but once you get going with active speakers, you'll never look back! Don't forget to ask on the miniDSP forum if you have questions.

Digital crossovers are one of the key core strengths of the miniDSP range of products. In this app note, we'll provide an overview of digital crossovers and how they differ from passive crossovers and other types of active crossover. With a high degree of flexibility, user-friendly interfaces, and unbeatable value, miniDSP digital crossovers offer the perfect solution for both the DIYer and the loudspeaker/system professional. So without further due, let's get started on how to build an audio crossover?

• Passive crossovers
• Active crossovers
• Digital crossovers
• The miniDSP advantage
• What else is needed?

Passive crossovers [Top]

A passive crossover uses only components such as resistors, coils (inductors), and capacitors to divide the signal from the power amplifier into different frequency bands for the different drivers — woofer, midrange, tweeter, for example. (The term "passive" refers to a device or circuit that is not able to control electron flow supplied from a power source, like a transistor, tube, or opamp.)

The following diagram shows a typical system configuration, where volume control () is done in the preamp and the passive crossover is located in the loudspeaker cabinet. The diagram shows a second-order two-way crossover, which has relatively gradual cutoff slopes on the woofer and tweeter — more components will be needed for steeper cutoff slopes. Also, more complex speakers, such as a three-way or four-way, will have a lot more components than shown here. As you can see, such passive crossover circuit is a combination of capacitors, resistors and inductors.

Block diagram of system with passive crossover

Active crossover [Top]

An active crossover, in contrast, divides the frequency band using the line-level signal and is typically positioned between a preamplifier and the power amplifiers. Each loudspeaker driver has its own dedicated channel of amplification, as shown in the diagram below.

Connecting a driver directly to an amplifier channel improves damping factor and gives the amplifier greater "control" over the driver. (A capacitor is often used in series with the tweeter to protect it from possible low-frequency or DC transients, especially at turn-on or turn-off.) The large — often expensive — passive components between the amplifier channels and the speaker drivers are no longer needed. This advantage is even greater for three-way and 4-way speakers, as they need larger component values for lower crossover frequencies.

Block diagram of system with active crossover (with analog inputs)

Digital crossover [Top]

Until recently, most active crossovers were implemented with analog circuitry, typically using op-amps to realize specific types of circuit topologies. Switches or plugin-modules select different crossover frequencies. This type of active crossover is limited by the fact that each filter has to be realized with a physical circuit. For example, making the crossover slope steeper would require additional analog circuitry - not easy once a unit is in the field.

With modern DSP (digital signal processing) technology, active crossovers can be implemented entirely with digital computation. This means that the audio processing can be changed much more easily, without any hardware changes. The amount of audio processing is limited only by the DSP power available. Digital crossovers also support direct digital input from a digital source, such as a computer. The following diagram shows a typical system configuration, where volume control can be done digitally either in the source or in the crossover itself. (Note that digital crossovers still support analog input as in the diagram above.)

Block diagram of system with active crossover (with digital inputs)

The miniDSP advantage and How to Build a Crossover? [Top]

Digital crossovers from miniDSP incorporate many additional functions, enabled by flexible onboard DSP and our friendly user interfaces. We have also put together an extensive library of application notes to help you make the most of these features. Check out the below app notes to get started!

Flexibility. miniDSP crossovers span the range from simple two-way up to complex four-way (or even five-way) configurations. The following app notes explain how it's done:

• Building a 2way crossover - basic tutorial on how to build a two-way active speaker.
• Stereo 3/4 way Crossover - basic tutorial on how to build a 3/4-way active speaker.
• Building a two-way active loudspeaker with the PWR-ICE125 - tutorial on how to build a complete 2-way active speaker with integrated crossover and amplification.
• Building active speaker with the 2x4HD
• Building active speakers with Dirac Live DDRC-88BM

Parametric equalization. All our crossovers include extensive parametric equalization capabilities, for correction of loudspeaker driver response, addressing problematic room modes, and for tailoring overall system response. Please see the following app notes:

• PEQ vs Graphic EQ - basic explanation of parametric EQ and how it compares to a graphic EQ.
• Equalizing your home theater system with the nanoAVR and UMIK-1 - this app note shows the process for the nanoAVR, but it applies to any of our crossovers.
• Auto-EQ tuning with REW - how to use Room EQ Wizard (REW) to generate correction curves for your system and room.

Advanced biquad programming. The miniDSP crossovers incorporate a feature that allows almost infinite flexibility and customizability of driver and system response.

• Linkwitz Transform - how to implement a Linkwitz transform to improve the low end response.
• Advanced Biquad programming - more about advanced biquad programming.

Time delay/alignment. Time alignment on all output channels is essential to ensure smooth response through the crossover region. See the application note How to time-align speaker drivers.

What else is needed? [Top]

In addition to the active crossover itself, you will need loudspeaker boxes with drivers. You can either build a set from scratch, or convert an existing loudspeaker into an active speaker by removing the passive crossover.

You will need sufficient channels of amplification: one per loudspeaker driver. You may even have enough stereo amplifiers available already. Otherwise, multi-channel amps intended for home theater applications will work well and are available at very reasonable prices, and if you have a DIY inclination then Class D power amplifier circuit boards with two, four, and even six channels are also available. Or, our PWR-ICE series of plate amplifiers might be a perfect match for your project.

You will also need to be able to perform acoustic measurements. We recommend the free program Room EQ Wizard (REW), together with our UMIK-1 USB measurement microphone. We've put together an app note to show you what to do: Loudspeaker measurement with UMIK-1 and REW! (If you have an existing measurement microphone and/or software, that is fine too, as long as you are able to obtain reasonably accurate measurements.)

Finally, you will need to be willing to learn! Mastering active crossovers is an adventure, and we're delighted to be part of it. Please register and join in on our forum to join other like-minded folks building active speaker systems.

The Dirac Series of high-resolution audio processors from miniDSP incorporating Dirac Live® are an advanced hardware/software solution for digital room correction (DRC). In this application note, we'll give a quick run-down on how to use one. We'll assume that you've already obtained your Dirac Live software license and installed the Dirac Live Calibration tool for miniDSP software.

• 1. Get your calibration file
• 2. Get connected
• 3. Configure
• 4. Run the measurements
• 5. Generate correction filters
• 6. Load and listen!

1. Get your calibration file [Top]

Go to the UMIK-1 page and enter your microphone's serial number. It is in the form xxx-yyyy and labeled on the microphone. Use "Save As" in your browser to save the numbers as a text file e.g. UMIK-7001870.txt.

The calibration file ensures that your microphone is as accurate as possible. Each microphone has a unique calibration file, which is why the serial number must be entered.

2. Get connected [Top]

Connections to a Dirac Series audio processor are straightforward:

• Audio input using either digital or analog connections (depending on the model)
• Audio output using either digital or analog connections (depending on the model)
• USB from your computer to the processor
• USB from the UMIK-1 to the computer

The following diagram illustrates a typical setup, in this case for the DDRC-22A. (Refer to the User Manual if you require more information on these connections.)

3. Configure [Top]

Install the downloaded software and verify your license. Upon running it, you will see six clickable icons that select different tabs for configuring, measurement, and filter generation:

Typically, these are worked through in order from top to bottom. On the Sound system tab, verify that DDRC-22 (miniDSP Ltd) is selected:

On the Mic config tab, select the UMIK-1 as the recording device. Click on the Load file button and locate the unique microphone calibration file that you downloaded from the UMIK-1 webpage in Step 1.

On the Output & levels tab, set the output volume to low. Click on the Test button for the left channel and gradually increase the output volume until it is at a moderate level, such that your voice would have to be raised to converse with someone sitting next to you.

Now increase the input gain so that the blue level bar reaches into the green section of the level meter for the left channel. With the UMIK-1, you may need to put the slider all the way up - this is normal. If there is insufficient input level, increase the output volume. (You can also check the microphone gain in the Windows Control Panel.)

Repeat the test signal for the right channel. The level should be correct without any further adjustment needed.

4. Run the measurements [Top]

You are now ready to run the acoustic measurements! You will need to take nine sets of measurements spread around the listening position. On the Measurements tab, select the most appropriate listening area (chair or sofa). Position the microphone at the location indicated by the arrow. For detailed advice and instructions on performing measurements, please read the User Manual.

Then click on the Start button. The program will run three measurement sweeps, through the left speaker, then right, then left again, and display the measurement result as a plot.

The arrow indicating the microphone position will then move to the next location. You can then press Start again to run another measurement. Proceed methodically through all nine measurements:

Note: It is important that the measurement locations be well spread out. Even for the Chair listening setup, spread measurement locations across a circle of at least a meter (three feet) across and up and down by +/- 30 cm (+/- one foot).

At this point, you should save your project by clicking on the Save... button.

5. Generate correction filters [Top]

On the Filter Design tab, you will initially see the average of the measurements for the left and right channels in light blue.

Also displayed is a target curve in red. This is the desired in-room frequency response after correction. Typically, target curves have a small boost in the bass region, and a gentle fall to the extreme treble. You can adjust the target curve by clicking and dragging on the orange anchor points. Double-clicking on the curve will create another anchor point. Drag-selecting a region will zoom in on that region of the graph, and double-clicking will zoom back out again.

For detailed advice and instructions on setting a target curve, please read the User Manual. Once you have the target curve set, click on the Optimize button. This will generate the correction filters and display the predicted response as the green curve.

You will need to experiment with different target curves to determine what works best for your system in your room, as there is no universally "correct" in-room response.

6. Load and listen! [Top]

On the Export tab, you can load up to four different sets of correction filters into the DDRC. Simply drag the box at the top left into an empty slot - this will automatically download the filters to the DDRC hardware unit. To delete a filter, drag it from the slot onto the trash can icon at the top right.

Be sure to set the Filter button to On. Start playing some audio, sit back, and listen! You can now go back to the Filter Design screen, modify your target curve, and download a new set of filters to the DDRC. You will want to experiment with different target curves to determine what works best with your system and room. You can use a remote control to select between different correction filters to easily audition the differences.

Note that if you move your speakers or make other changes to your listening room, you may well want to redo your measurements in a new project. So keep that UMIK-1 handy!

AcourateDRC is the next generation of AcouratePRO software from Dr Ulrich Brüggemann. Designed with simplicity in mind, AcourateDRC exports filters who can be directly imported in the OpenDRC.
An easy to use interface for an efficient result at correcting for the listening environment but also the weakness of a given speaker set. The soundstage becomes more stable and deep by improving the coherence of left and right side speakers. The boom factor caused by standing waves and room modes is removed. The end result is clean sound with a controlled bass. 

What you will need

• A miniDSP OpenDRC-DI,  OpenDRC-DA or OpenDRC-AN
• AcourateDRC to generate the correction filters. A demo version is currently available at the following link. Purchase of the license to extract the filter is available via Paypal through the software.
• A measurement microphone such as UMIK-1

Note that AcourateDRC is a Windows-only program. If you are not sure if you will be able to use your soundcard with AcourateDRC, download the trial version and go through the measurement process described below. This program is a 3rd party program. All support questions should be directed to AudioVero.

System setup

You will first need to get the OpenDRC connected into your system.

OpenDRC-DI: your digital source or sources are connected to the OpenDRC inputs, and the OpenDRC output is connected to your DAC. Since the OpenDRC-DI has a volume control, you can use it as a digital preamp.

OpenDRC-AN: your preamp is connected to the OpenDRC inputs, and the OpenDRC outputs are connected to your power amp. Or, if you have only a single source, you can use the OpenDRC-AN as the preamp, as it has volume control on board. The connections are balanced, so you will need to use adapters if your components don't have balanced signal connections.

OpenDRC-DA: your digital sources are connected to the OpenDRC input, while the analog outputs can directly connect to your amplifiers.

Measuring with AcourateDRC

There are four steps to the process:

1. Measure the system in AcourateDRC, without any processing being performed in the OpenDRC
2. Generate the correction filter in AcourateDRC
3. Load the correction filters into the OpenDRC
4. Play music through the system

Ideally, you will able to perform step 1 with the OpenDRC still connected in the system. This will enable you to verify the result after loading correction filters, and also to take advantage of the parametric EQ function of the OpenDRC. If using OpenDRC-AN, connect your soundcard output to the OpenDRC-AN inputs. If using OpenDRC-DI or OpenDRC-DA, and your soundcard has a digital output, connect its digital output to the OpenDRC-DI input. You can also use a USB-to-SPDIF convertor.

If you have OpenDRC-DI but no digital output from your computer, then disconnect the OpenDRC-DI for measurement and connect the soundcard outputs to your preamp or power amp inputs.

1. Measure system response

In AcourateDRC, first ensure that your soundcard is configured in the Setup screen.

Then, on the Measurement screen, check the signal levels with the white noise generator. Set the level low initially, then adjust the gains on your soundcard and preamp (if present) until you have the test signal playing at a comfortable level and AcourateDRC indicates an adequate input level.

Note: if you have an OpenDRC inserted in the system, ensure that the FIR filters are set to Bypass mode.

Then click on "Logsweep Recording." AcourateDRC will play a sweep tone, first through the left speaker and then the right speaker, and will display the measurement result as it proceeds:

2. Generate correction filters

After completing the measurement, AcourateDRC switches to the Calculation screen. Here you can set the target response (the frequency response desired for the system after correction is applied). To get started, there are a number of predefined target responses that you can select from the Setup screen:

You can adjust the target response using the controls on the right of the window. Then click on the "Correction" button. AcourateDRC computes the correction filters and displays the predicted response after correction:

3. Load correction filters

In the OpenDRC 2x2 plugin interface, enable the FIR filters (by deselecting the Bypass button). Click on "FIR - channel 1" and then on "Browse." Navigate to the location where AcourateDRC has stored the impulse response files (for example, Libraries > Documents > My Documents > AcourateDRC > My Project) and select FilterL.bin. Click on Open and the correction filter is loaded and then displayed:

Before closing the FIR window, click on "Send to DSP". Then repeat the process for "FIR - channel 2" and load FilterR.bin.

4. Play music

If you have disconnected OpenDRC for measuring, reconnect it. You can now play music through your system.

You may wish to measure the system response with and without DRC enabled. Run a measurement sweep, and then disable FIR filtering in the OpenDRC plugin and run another. The following screenshot shows the example system measured with Room EQ Wizard, without DRC in red, and with DRC in blue:

Repeat the above process with different target curves, to find what is best for you and your system. (You can rename the filter files each time so they don't get overwritten.) You can store up to four different sets of correction curves in the OpenDRC, to make it easy to switch between target curves and compare them. Configurations can be selected using a remote control, the front panel control, or in the plugin interface.

Don't forget to post on the miniDSP forum if you have questions!

All Digital Room Correction with Squeezebox + OpenDRC + Sumoh Amps

Introducing the world of Digital Room Correction (DRC), a process where digital filters are used to correct speaker + room aberrations. A 3 step process yet an acoustical challenge where DSP & Math Gurus use all resources to solve this complex problem. The following lines will summarize the basics of room correction in a setup involving a PC/Mac with a measurement/filter generation software, a measurement microphone and an audio processor (OpenDRC).

Step1: Measurement

The first step in any Room correction process is to "listen" to the room + speaker interaction. Using a a software + measurement microphone combo, a set of test sweeping tones are played back through your speaker systems and recorded by the software application. No doubt that your neighbors will be wondering what you're up too! However please don't despair, because providing you do it well, it will lead to great results and they will be the envious one... Back to the technical aspect, the sweeps are used to "measure" the response in both time and frequency of your system. Since the resulting impulse response will be key to the filter generation (step2), this step is key to your success. Depending on the software used, you may have to take one (AcourateDRC) or multiple measurements (Dirac Live DDRC series) in different locations.

Step2: Filter generation

At the center stage of "good to great" room correction fitlering is the array of math and psycho-acoustic models used to "generate" filters based on the measurement. In audio, two types of filters are typically used: IIR (Infinite Impulse Response) or FIR (Finite Impulse Response). In today's realm, most room correction filters are either FIR based or a combination of FIR + IIR filters. While each software has a different interface and its ways to calculate the filters, a comon element is the"target curve". In other words, the expected frequency response and a guideline for the filter generation. 

Step3: Filter processing

Your filter settings are now ready and it's time to listen to the end result. This is where the OpenDRC/DDRC series plays a key role. What used to be a noisy/power hungry PC running in the middle of your Hifi set is now a low cost, low power yet powerful processor thanks to the introduction of miniDSP's room correction units. With a range of all digital, all analog or hybrid, the OpenDRC is a solution for multiple configurations. It is a set and forget where the end user still get a chance to tweak some settings by allowing end users the flexibility they expect in today's world.

Powered by miniDSP room correction units

With miniDSP's leading open architecture FIR engine for room correction, a couple of software developers already came on boards to provide an array of Filter generation software. The following list is to be updated over time as software currently under development become available to our end users.

• Dirac Live: 

Introducing the Dirac Series of 24/96 high-resolution audio processors. miniDSP and Dirac Research bring you this unique hardware/software combination featuring Dirac Live® — the world's premiere digital room correction solution. For more information, please have a look at the following link.

• Python Open Room Correction (PORC)

A free command line tool and an Open Source project ported by Mason A. Green and based based on the work of Dr. Balázs Ban hosted at the following webpage. This package allows one to specify a target curve and generate the corresponding loudspeaker-room correction filters for the OpenDRC platforms. For more information, please have a look at the following link.

• rePhase

rePhase is a  loudspeaker phase linearisation, EQ and filtering tool compatible with the OpenDRC-DI/AN and miniSHARC products. rePhase let you generate finite impulse responses (FIR) specifically tailored to reverse the phase shifts introduced by the crossovers of your loudspeakers (passive or active), resulting in a linear-phase system. It can also generate linear-phase EQ and crossover filters of arbitrary slopes, including Linkwitz-Riley (albeit linear-phase) and Horbach-Keele shapes.

How to get started ?

You're ready to try miniDSP's room correction experience and want to know what's required. Here are the basic core element for a stereo configuration:

• 1 x OpenDRC-DI/AN for processing of the stereo signal. This device will be located between the source and your DAC (SPDIF/Toslink/AES) or inline with your AVR if you have the ability to loop back a signal. Note that if you intend to use Dirac live, only the DDRC series is compatible with Dirac.
• 1 x Measurement microphone such as miniDSP UMIK 1.
• 1 x A compatible software for calculation of the FIR filter parameters (E.g.  Porc, rePhase, DRC FIR open source or your own software)

And that will be it for you to be ready to go!

FuzzMeasure Pro is an easy-to-use acoustic measurement program for Apple Mac computers. In this app note we will show you how to get started with the UMIK-1 and FuzzMeasure Pro.

• 1. Get your calibration file
• 2. Get connected
• 3. Configure system and software
• 4. Calibrate SPL
• 5. Test and measure
• What's next?

1. Get your calibration file [Top]

On the UMIK-1 page, enter your microphone's serial number to download its unique calibration file. The calibration file ensures that measurements made with your microphone are as accurate as possible.

2. Get connected [Top]

Mount the UMIK-1 into the small stand supplied with it, or if you like, you can use any other microphone stand. Connect the UMIK-1 to your Mac using the supplied USB cable.

You will also need to connect your Mac to generate audio output through your system. There are several options:

• The inbuilt analog line output with a 3.5 mm to RCA adapter cable
• Your Mac's optical digital audio output connected to a DAC or A/V receiver
• A USB cable to an external DAC
• An HDMI connection to an A/V receiver

This photograph shows the UMIK-1 connected to a MacBook Pro and a Mini DisplayPort to HDMI adapter cable for audio output:

3. Configure system and software [Top]

Start up FuzzMeasure Pro. The toolbar across the top has several handy buttons that we will use below:

Click on the Capture Settings button. Here are two examples:

Set the output device under Playback Settings to your chosen output device. In the example above, at the left the inbuilt analog or digital output is selected. At the right, the HDMI output device is being used. Select one or more output channels to send the measurement signal to.

Also set the input device under Record Settings to the UMIK-1 and select Channel 01.

Click on the Configure button for the output device to open it in the Audio MIDI Setup program. Set the output device to operate at 48000 Hz (48 kHz), the same sample rate as the UMIK-1:

Finally, load the calibration file for your UMIK-1 into FuzzMeasure. From the Window menu, select Microphone Calibration. Click on the "+" icon and select the calibration file you saved in step one above to add it to FuzzMeasure's list of calibration records. Then near the top of the window, select that calibration file from the drop-down menu.

4. Calibrate SPL [Top]

SPL calibration is optional. Without performing the SPL calibration step, the measurements will be correct but the absolute SPL readings in FuzzMeasure's Sound Pressure Level frequency response display will not be correct. In that case, you may prefer to use the Magnitude Response graph instead to avoid confusion.

If you decide to do SPL calibration, you will need to generate an acoustic signal of 94 dB at the microphone tip. If you have a microphone calibrator, then of course you will use that. Otherwise, place the microphone and an external SPL meter close to a loudspeaker driver. Generate pink noise and increase the volume until the SPL meter reads 94 dB.

(To generate pink noise, you may be able to play a test disc through your system. There are also some nice Mac applications for generating audio test signals, such as SignalSuite.)

Once you have a 94 dB signal at the microphone, click the Level Meter button. Then click on the small "C" button so it lights up green. From then on, FuzzMeasure will take that SPL to be equal to 94 dB.

5. Test and measure [Top]

Before proceeding, turn down the volume on your preamp or A/V receiver to a fairly low level. Position the UMIK-1 where you want to take the measurement, and point it towards the speaker being measured.

Click on the Sweep Settings button and set the parameters as follows:

Then click on the Measure button. You should hear a faint sweep from low to high frequencies through the selected speaker. Turn up the system volume until the sweep is loud enough to halt a conversation, but not so loud as to be uncomfortable or to sound distorted. If you have done the SPL calibration, an average SPL of about 75 dB is generally considered suitable for home measurements.

You can now run as many more measurements as you like! FuzzMeasure Pro allows you to save all current measurements as a "project" and reload it again later. You can view multiple measurements at once by Command-clicking on them.

FuzzMeasure provides a lot of analysis tools, accessible from the menus. Explaining them all is beyond the scope of this app note, but you should be able to quickly find the options to set the amount of smoothing, to display harmonic distortion, and to change the frequency and SPL range displayed on the graph. To share your graphs online, right-click on any graph and select Export Image File.

What's next? [Top] [Top]

Now that you have the ability to run acoustic measurements with FuzzMeasure Pro, you can proceed to optimize and equalize your system. The following app notes contain information that you may find helpful in your journey:

• Loudspeaker measurement with UMIK-1 and REW explains the principles of loudspeaker measurement (but you would use FuzzMeasure instead of Room EQ Wizard.)
• Building a 2way crossover and Stereo 3/4 way Crossover describe the basics of setting up a loudspeaker with an active crossover.

The miniDSP UMIK-1 is the perfect companion to audio analysis programs running on your tablet or even your smartphone. In this application note we will show you how to set up your UMIK-1 with the SignalScope Pro app from Faber Acoustical, running on an Apple iPad. SignalScope Pro is able to set level and frequency response readings (FFT only) from the UMIK-1 calibration data.

Please note: miniDSP cannot provide support for third-party applications or hardware. This app note shows you how to set up the miniDSP UMIK-1 to use with SignalScope Pro but other functions of the Apple iPad or the SignalScope Pro app are beyond the scope of miniDSP support.

• 1. Connect
• 2. Calibrate
• 3. Measure!

1. Connect [Top]

Mount the UMIK-1 into the small stand supplied with it, or if you like, you can use any other microphone stand. To connect the UMIK-1 to your Apple iPad, you will need either the Lightning to USB adapter (for newer iPads with an 8-pin Lightning connector) or the Camera Connection Kit (for older iPads with a 30-pin connector). This photo shows an iPad mini connected to the UMIK-1 via the Lightning adapter:

To generate audio output from the iPad, connect a cable from the iPad headphone socket to the system being tested. On the iPad end, the cable will have a 3.5mm stereo jack, and typically the other end has a pair of RCA connectors, as shown at left in the photograph below. Alternatively, use a 3.5mm stereo to RCA adapter, as shown at right.

2. Calibrate [Top]

If you haven't already, install the SignalScope Pro program on your iPad from the App Store.

Upon opening SignalScope Pro, it will detect the presence of the connected UMIK-1 and query you for its serial number:

Enter your microphone's serial number and press OK. SignalScope Pro will fetch the calibration file from the miniDSP website and load the calibration data. Each microphone has a unique calibration file, which is why the serial number must be entered.

Then open the Configuration screen (the "gear" icon), and tap on I/O Device Configuration and then Input Channels. You will see the UMIK-1 has been selected:

To enable frequency response compensation (FRC) for the FFT display, turn on Apply FR Compensation. (Note that FRC operates on the FFT tool, but not on the Octave analyzer tool.) To see the frequency response compensation graph, click on Frequency Response Data:

3. Measure! [Top]

Tap on Sig Gen at the bottom of the screen. You can experiment with the different types of test signal here. For now, set it as shown below, adjusting the volume so that it plays through the system at a comfortable level. Ensure that the signal generator is enabled at the top right, and turn signal generation on and off with Play/Pause button (next to the Headphones label at the top left. Adjust the level with the iPad volume control and the Noise Amplitude slider, and use the Noise Pan control to send output to only one speaker.

SignalScope Pro has a number of measurement tools, selected by the icons at the bottom of the screen. We will walk through some simple examples. To start with, tap on Octave to open the real-time analyzer (RTA), and top on the Play/Pause button at the top left (next to "UMIK-1") to start the analyzer. This will generate a pink noise test signal and display the spectrum picked up by the UMIK-1. (Note that the Octave tool doesn't use the UMIK-1 frequency response calibration data.)

(If you don't want to generate the test signal but just analyze ambient noise, tap on the Sig Gen icon, disable the signal generator, then go back to the analyzer.)

The Options screen can be used to set options, such as octave or third-octave display, frequency weighting, response time, and whether or not the display scales automatically.

Go back to the signal generator and set it to generate a square wave instead of pink noise. Now tap on the Oscope icon and then the Play/Pause button to display the actual time-domain signal being picked up by the UMIK-1:

Finally, change the signal generator to generate a pure tone (sine wave). Then go to the FFT display and tap the Play/Pause button. The display will show the tone being generated as a high "spike" as well as any distortion components. (For example, if you play a 1 kHz tone, distortion components will be at 2, 3, and 4 kHz):

Note: the distortion components may include distortion from the microphone, so take care in interpreting these results.

Finally, tap on the Meter icon. This will display a powerful and accurate SPL meter with all the features of physical meter. (If you don't want to generate the test signal but simply meter ambient noise, tap on the Sig Gen icon, disable the signal generator, then go back to the Meter.)

The miniDSP UMIK-1 is the perfect companion to audio analysis programs running on your tablet or even your smartphone. In this application note we will show you how to set up your UMIK-1 for use with the Smaart® Tools from StudioSix Digital, running on an Apple iPad.

Please note: miniDSP cannot provide support for third-party applications or hardware. This app note shows you how to set up the miniDSP UMIK-1 to use with the AudioTools Smaart® plugin but other functions of the Apple iPad hardware or the AudioTools app are beyond the scope of miniDSP support.

• 1. Get your calibration file
• 2. Get connected
• 3. Calibrate
• 4. Measure!

1. Get your calibration file [Top]

Go to the UMIK-1 page and enter your microphone's serial number. It is in the form xxx-yyyy and labeled on the microphone. Use "Save As" in your browser to save the numbers as a file e.g. UMIK-7001870.txt.

The calibration file ensures that your microphone is as accurate as possible. Each microphone has a unique calibration file, which is why the serial number must be entered.

"Sens Factor =-7.062dB, SERNO: 7001870"	
10.054	-6.5726 
10.179	-6.3949 
10.306	-6.2205
...

2. Get connected [Top]

Mount the UMIK-1 into the small stand supplied with it, or if you like, you can use any other microphone stand. To connect the UMIK-1 to your Apple iPad, you will need either the Lightning to USB adapter (for newer iPads with an 8-pin Lightning connector) or the Camera Connection Kit (for older iPads with a 30-pin connector). This photo shows an iPad mini connected to the UMIK-1 via the Lightning adapter:

3. Calibrate and set up[Top]

Calibration is the same as described in the app note Using the UMIK-101 with AudioTools on the Apple iPad.

If you haven't already, you will need to purchase the Smaart® plugin from within AudioTools. From the home screen, simply tap on Acoustics then on Smaart® Tools I. You will be asked for your App Store credentials so you can make the purchase.

4. Measure! [Top]

Tap on Acoustics, then on Smaart® Tools I. Click on the spanner/wrench icon near the bottom of the screen to set the parameters. Here is a typical example set of settings:

Below is an example screen showing the Smaart® module in action. Here, we have enabled the dual display mode, and are displaying an RTA, spectrum, and a spectrogram while monitoring music playback:

The miniDSP UMIK-1 is the perfect companion to audio analysis programs running on your tablet or even your smartphone. In this application note we will show you how to set up your UMIK-1 for use with the AudioTools app from StudioSix Digital, running on an Apple iPad.

Please note: miniDSP cannot provide support for third-party applications or hardware. This app note shows you how to set up the miniDSP UMIK-1 to use with AudioTools but other functions of the Apple iPad hardware or the AudioTools app are beyond the scope of miniDSP support.

• 1. Get your calibration file
• 2. Get connected
• 3. Calibrate
• 4. Measure!

1. Get your calibration file [Top]

Go to the UMIK-1 page and enter your microphone's serial number. It is in the form xxx-yyyy and labeled on the microphone. Use "Save As" in your browser to save the numbers as a file e.g. UMIK-7001870.txt.

The calibration file ensures that your microphone is as accurate as possible. Each microphone has a unique calibration file, which is why the serial number must be entered.

"Sens Factor =-7.062dB, SERNO: 7001870"	
10.054	-6.5726 
10.179	-6.3949 
10.306	-6.2205
...

2. Get connected [Top]

Mount the UMIK-1 into the small stand supplied with it, or if you like, you can use any other microphone stand. To connect the UMIK-1 to your Apple iPad, you will need either the Lightning to USB adapter (for newer iPads with an 8-pin Lightning connector) or the Camera Connection Kit (for older iPads with a 30-pin connector). This photo shows an iPad mini connected to the UMIK-1 via the Lightning adapter:

To generate audio output from the iPad, connect a cable from the iPad headphone socket to the system being tested. On the iPad end, the cable will have a 3.5mm stereo jack, and typically the other end has a pair of RCA connectors, as shown at left in the photograph below. Alternatively, use a 3.5mm stereo to RCA adapter, as shown at right.

3. Calibrate [Top]

If you haven't already, install the Audio Tools program on your iPad from the App Store.

Open the app and go to the Settings men, and then select Microphone Setup. You will see see two choices: Low Range and High Range. Select Low Range.

Then click on the "i" icon to the right. This will bring up the microphone calibration screen.

In the screenshot above, the name of a UMIK-1 calibration file is shown. Initially, however, this will be blank, so you will need to load the calibration file to the iPad and select it. Click on the Calibration File link and then on the Files link at the bottom right. This will bring up a screen with information on how to transfer the calibration file from your computer to the iPad:

On your computer's browser, enter the address shown by AudioTools on the above screen and upload the calibration file from step 1 above. Then, back in the AudioTools interface on the iPad, tap on the name of the calibration file and tap Apply. The calibration screen will change to display the name, as shown above.

It is helpful to adjust the Trim parameter so that the SPL readings given by AudioTool are correct. Normally, you need a separate SPL meter to do this, but since each UMIK-1 contains a sensitivity calibration number (Sens Factor) in its calibration file, you can do this without one. Just work out the following formula:

   Trim = 23 – (Sens Factor)

In our example cal file, the Sens Factor is given as -7.1, so the number we enter into the Trim field is:

    Trim = 23 – (–7.1) = 23 + 7.1 = 30.1

(There is no need to use more than one decimal place for this calculation.) Then tap on Done to close the calibration screen, and on Done again to go back to the main AudioTools menu.

(Fortunately, you only need to go through the above procedure once!)

4. Measure! [Top]

AudioTools has a number of modules, some of which require additional "in app" purchase. Here, we will give a couple of quick examples of included modules. Click on the Acoustics icon at the left and then on the RTA icon. The display will show the spectrum of the signal being picked up by the UMIK-1. You can use this to monitor frequency content while music is playing:

The RTA can be changed between octave smoothing and 1/3rd octave smoothing using the selector at the lower left, and the decay time changed using the selector at the lower right.

To show more detail in the measured frequency response, use the FFT tool. This time, let's generate a pink noise signal so we can measure the frequency response of the system (rather than just the spectrum of music playing). First, open the FFT tool. then click on the small sine wave icon at the bottom of the screen. A control panel will pop up; set it for pink noise and tap the button at its top left. (Note: make sure to turn the iPad output volume or your system volume down first, and then increase it gradually.)

With pink noise playing through the system, the FFT shows the frequency response of the system as measured in the room. You can adjust the frequency resolution using the control at the lower left.

Dirac Live software calibration tool is an advanced music processor that corrects for acoustical errors at the listening position caused by reflections from the listening room. It also provides the ability to tailor the response at the listening position using target curves. Dirac Live runs on the Windows and Mac platforms. Note that this application note is for "Dirac Software" only (i.e. running on PC/Mac). If you are owning a DDRC or nanoAVR, please consult the appropriate app notes.

In order to generate the data that Dirac Live uses for its room correction algorithm, acoustic measurements of your speakers in your listening room are required, using the Dirac Live Calibration Tool. The miniDSP UMIK-1 is a low-cost calibrated measurement microphone that is ideal for use with the Dirac Live Calibration Tool.

• What you will need
• 1. Get connected
• 2. Configure Dirac Live Calibration Tool
• 3. Set levels
• 4. Run the measurements
• 5. Generate correction filters
• 6. Load and listen!

What you will need [Top]

• Dirac Live. The Dirac Live download includes two programs that you will need to install: the Dirac Audio Processor, and the Dirac Live Calibration Tool.
• A miniDSP UMIK-1. After receiving the UMIK-1, go to the UMIK-1 page to get the calibration file for your unique serial number, and save it as a .TXT file.
• A microphone stand with boom arm, available in music supply stores. (While the UMIK-1 is supplied with a small "table top" stand, the Calibration Tool program requires that the microphone be positioned in several locations around the listening area).

1. Get connected [Top]

To use Dirac Live, you will need to have audio playing from your computer through your audio system. A basic method is to connect the line out jack from your computer to your preamp, but for higher quality a USB DAC or USB-SPDIF convertor and a DAC will be used. If connecting to an A/V preamp or receiver, you may be able to use HDMI if your computer supports it: on Windows, use an HDMI-to-HDMI cable; on the Mac, use a DisplayPort to HDMI adapter cable into a Thunderbolt port.

On the input side, the UMIK-1 can be simply connected to any available USB port on your computer. Position the computer and stand so that the mic can be moved into several positions, as will be seen in the Dirac Live Calibration Tool interface (below).

2. Configure Dirac Live Calibration Tool [Top]

Install the Dirac Audio Processor and the Dirac Live Calibration Tool. Double-click on the Dirac Live Calibration Tool application to run it. (On Windows, it will be in the Start Menu, under the Dirac folder; on the Mac, it is located in the /Applications/Dirac folder.) On the left are five clickable icons that select different tabs for configuring, measurement, and filter generation:

First, select the output device that you will be using from the Sound system tab, and all sample rates at which you will be playing audio:

On the Mic config tab, select the UMIK-1 as the recording device. Click on the Load file button and locate the unique microphone calibration file that you downloaded from the UMIK-1 webpage.

3. Set levels [Top]

On the Output & levels tab, set the output volume to low. Click on the Test button for the left channel and gradually increase the output volume until it is at a moderate level, such that your voice would have to be raised to converse with someone sitting next to you. (If the output volume slider cannot be moved, adjust the volume with your audio system's volume control.)

Now increase the input gain so that the blue level bar reaches into the green section of the level meter for the left channel. With the UMIK-1, you may need to put the slider all the way up - this is normal. If there is insufficient input level, increase the output volume. You can also check:

• In Windows, go to the Sound pane in the Control Panel, then Recording, then select the UMIK-1 and view its Properties. In Levels, set the gain up to 100.
• On Mac, check in the Audio MIDI Setup app (in Applications/Utilities) and set the UMIK-1 gain up to 24 dB.

Repeat the test signal for the right channel. The level should be correct without any further adjustment needed.

4. Run the measurements [Top]

You are now ready to run the acoustic measurements! Dirac Live Calibration Tool uses nine measurements spread around the listening position to calculate its correction filters.

On the Measurements tab, select the most appropriate listening setup (chair, sofa, or auditorium). Position the microphone at the location indicated by the arrow. Be sure to check both the top and front views using the selector underneath the graph, so that you have the height of the microphone correct.

Then click on the Start button. The Calibration Tool will run three measurement sweeps, through the left speaker, then right, then left again, and display the measurement result as a plot.

Dirac Live Calibration Tool will then ask you to move the microphone to the next location, after which you can press Start again run another measurement. Proceed methodically through all nine measurements.

At this point, you should save your project by clicking on the Save... button.

5. Generate correction filters [Top]

On the Filter Design tab, you will initially see the average of the measurements for the left and right channels. You can also opt to display all individual measurements — this will show you how much variation there is across your listening area.

Also displayed is a target curve. This is the desired in-room frequency response after correction. Typically, target curves have a small boost in the bass region, and a gentle fall to the extreme treble. You can adjust the target curve by clicking and dragging on the circular grab-points. Double-clicking on the curve will create another grab-point. Drag-selecting a region will zoom in on that region of the graph, and double-clicking will zoom back out again.

You may find you need to experiment with different target curves to determine what works best for your system in your room, as there is no universally "correct" in-room response. Once you have the target curve set, click on the Optimize button. This will generate the correction filters and display the predicted response.

Click on Save Filter to save the correction filter as a file. You can then, if you wish, set a different target curve, click Optimize again, and save the result to a different filter.

6. Load and listen! [Top]

Start the Dirac Audio Processor (DAP). (Refer to the Dirac Live documentation for information on how to set up DAP to start automatically on system boot.) You will need to set your system and/or audio/media player to use DAP as its audio output device. (The DAP appears to other programs as a virtual audio device, that goes "in between" the program playing audio and the real audio device.)

Click on the first vacant slot, which will say "Click to load". A pop-up menu will allow you to select one of your correction filters. If you created more than one filter, click on the next vacant slot and load that filter. You can load up to four correction filters at a time, which allows for easy auditioning of different target curves.

Be sure to set the Filter button to On. Start playing some audio, sit back, and listen! You will want to experiment with different target curves to determine what works best with your system and room. Note that if you move your speakers or add acoustic treatments to your room, you will need to redo the measurements, so keep your UMIK-1 handy!

This app note shows you how to set up the UMIK-1 and Room EQ Wizard (REW) on the Mac, together with an HDMI output device such as an A/V receiver or A/V preamp. This enables you to select individual output channels for use in measuring and equalizing a home theater or multi-channel surround system. You can use a Mac with an HDMI output port or a Thunderbolt port.

Your HDMI device will typically be an A/V receiver, but you may also have an A/V preamp or a Blu-ray player with HDMI input.

Please note: this app note requires OSX version 10.7.3 or higher. If you have OSX 10.6 or earlier and cannot upgrade, please try the "LineIn workaround," documented here.

• 1. Get your calibration file
• 2. Download and install software
• 3. Get connected
• 4. Configure system and software
• 5. Test and measure
• What's next?

1. Get your calibration file [Top]

Go to the UMIK-1 page and enter your microphone's serial number. It is in the form xxx-yyyy and labeled on the microphone. Use "Save As" in your browser to save the numbers as a file e.g. 7000343.txt.

The calibration file ensures that your microphone is as accurate as possible. Each microphone has a unique calibration file, which is why the serial number must be entered.

2. Download and install software [Top]

You will need to download and install two items of software:

• Download Room EQ Wizard (REW) from Home Theater Shack. You should get the latest version from the download area.

• SoundFlower is the application that we will use to "route" audio coming out of Room EQ Wizard to the various HDMI channels. You can download it from cycling74.com or directly from the Google Code download page. Please be sure to get the latest version, which as of the time of writing is 1.6.6b.

3. Get connected [Top]

Mount the UMIK-1 into the small stand supplied with it, or if you like, you can use any other microphone stand. Connect the UMIK-1 to your Mac using the supplied USB cable.

Connect your Mac to a suitable HDMI input of your A/V receiver or preamp. If using a Mac with an HDMI output port, you can use a regular HDMI cable. If using a Mac with a Thunderbolt port, you can use a Mini DisplayPort to HDMI adapter cable. Check that your HDMI device is set for multichannel output e.g. 5.1 or 7.1, and not to a stereo downmix.

The photograph below shows the necessary connections made to a Retina MacBook Pro.

4. Configure software [Top]

4.1. Audio MIDI Setup

Double-click on the Audio MIDI Setup app (in Applications/Utilities) to run it. Select the HDMI device, and:

1. Select 48000.0 Hz in the Format drop-down selector.
2. Click on the Configure Speakers... button. Set the HDMI device to the appropriate multichannel setting (e.g. 5.1 or 7.1 Surround), then click Done.

4.2. SoundFlowerBed

Double-click on the SoundFlowerBed app (in Applications/SoundFlower) to run it. This will show up as a little flower icon in the menu bar. Then:

1. Pop the menu down and in the section Soundflower (64ch), select the HDMI device (see screenshot below).
2. In the channel selectors, set the Channel 1 routing selector to HDMI [1] as shown below.
3. Set Channel 2 to None - this ensures that REW will output audio to only one channel at a time.

4.3. Room EQ Wizard

Double-click on the REW app to run it. You will see a screen asking if you want to use the UMIK-1. Click on Yes.

Answer Yes to the next question about the calibration file, and then locate the file that you saved in Step 1 above.

The main functions of REW will now be accessible from a row of buttons at the top of the window. This screenshot highlights the functions we will use in this app note:

Click on the Preferences pane, then:

1. Check that the sample rate is 48 kHz.
2. Set the output device to Soundflower (64ch).
3. Verify that the input device has been set to the UMIK-1.

5. Test and measure [Top]

Before proceeding, turn down the volume on your A/V preamp or receiver.

You can use the REW Signal generator to test that all is working correctly. Click on the Generator button and set the signal generator to "Speaker Cal" pink noise and the level at -12 dB, as shown in the screenshot below. Press the green play button. Gradually increase the volume on your A/V preamp or receiver - you should hear sound coming from the front left speaker.

To check all channels, change the Channel 1 routing selector in SoundFlowerBed to HDMI [2], HDMI [3], and so on. (Remember to keep Channel 2 set to None. There is no need to make or change any settings for Channel 3 and upwards, as REW does not output to those channels.) Here is a typical mapping of channel numbers:

• HDMI [1]: Front Left
• HDMI [2]: Front Right
• HDMI [3]: Center
• HDMI [4]: Subwoofer
• HDMI [5]: Surround Left
• HDMI [6]: Surround Right
• HDMI [7]: Surround Back Left
• HDMI [8]: Surround Back Right

Open the REW SPL Meter, and click on the red button to turn it on. Increase the volume in your A/V receiver or preamp until the level reads about 75 dB.

You are now all set to run a measurement sweep. Select the HDMI [1] output channel again and click on the Measure button. Then use Check Levels and then Start Measuring. You will hear a sweep through the front left speaker, and REW will display its first frequency response graph. Then user SoundFlowerBed to select the next channel and run a measurement on it, and so on.

The UMIK-1, like all measurement microphones, is slightly directional and at high frequencies (6 kHz and above) is slightly less sensitive to the sides. For highest accuracy and repeatability, point the UMIK-1 at each speaker being measured. Alternatively, orient the UMIK-1 vertically (pointed at the floor or ceiling) and use a 90-degree calibration file such as that is supplied when the UMIK-1 is purchased via Cross-Spectrum Labs

After completing all channel measurements, you can view them together in the Overlays window. (Apply some smoothing to the graphs using the pop-down Controls overlay.)

What's next? [Top] [Top]

Now that you have the ability to run acoustic measurements, you can proceed to optimize and equalize your system. The following app notes contain information that you may find helpful in your journey:

• Acoustic measurement with the UMIK-1 and REW
• Auto-EQ tuning with REW
• Home Theater Tuning

Credits

The use of Soundflower for HDMI output routing was first documented by EmagSamurai.

This app note shows you how to set up the UMIK-1 and Room EQ Wizard (REW) on Windows together with an HDMI output device such as an A/V receiver or A/V preamp. This enables you to select individual output channels for use in measuring and equalizing a home theater or multi-channel surround system. You will need a computer with an HDMI output port.

Your HDMI device will typically be an A/V receiver, but you may also have an A/V preamp or a Blu-ray player with HDMI input.

• 1. Get your calibration file
• 2. Download and install software
• 3. Get connected
• 4. Configure Room EQ Wizard
• 5. Configure ASIO4ALL
• 6. Select input and output channels
• 7. Test and measure
• What's next?

1. Get your calibration file [Top]

Go to the UMIK-1 page and enter your microphone's serial number. It is in the form xxx-yyyy and labeled on the microphone. Use "Save As" in your browser to save the numbers as a file e.g. 7000343.txt.

The calibration file ensures that your microphone is as accurate as possible. Each microphone has a unique calibration file, which is why the serial number must be entered.

2. Download and install software [Top]

You will need to download and install two items of software:

• Download Room EQ Wizard (REW) from Home Theater Shack. You should get the latest version from the download area at least 5.01 beta 17 is needed to operate correctly with the UMIK-1.

   

• Download ASIO4ALL from http://www.asio4all.com/. Double-click and follow the installation steps. The optional components don't need to be installed.

3. Get connected [Top]

Mount the UMIK-1 into the small stand supplied with it, or if you like, you can use any other microphone stand. Connect the UMIK-1 to your computer using the supplied USB cable.

Connect the HDMI output of your computer to a suitable HDMI input of your A/V preamp or receiver. Be sure that input is selected, and also check that your HDMI device is set for multichannel output e.g. 5.1 or 7.1 and not to a stereo downmix.

4. Configure Room EQ Wizard [Top]

Start REW. The main functions are accessible from a row of buttons at the top of the window. This screenshot highlights the functions we will use in this app note:

Click on the Preferencesbuttons and make the following selections (in this order):

• Under Drivers, select ASIO.
• Under Sample Rate, select 48 kHz.
• Under ASIO Device, select ASIO4ALL v2.

Then click on the ASIO Control Panel button (still in REW Preferences). This opens up the ASIO4ALL control panel.

5. Configure ASIO4ALL [Top]

In the ASIO4ALL control panel, set advanced mode. This is done by clicking on the wrench or spanner icon at the lower right. In advanced mode, there will be additional controls in the window and the spanner will have a red X over it.

In the list at the left, enable the HDMI audio device and the UMIK-1. You should deselect other audio devices to avoid clutter when selecting input and output channels in REW.

Audio devices don't always appear with obvious names, depending on the device and the version of Windows. In some cases, the UMIK-1 may appear just as "USB Audio Device." By hovering the mouse over the device you will get a popup window with more information that helps to identify the device. The UMIK-1, for example, will be an input device that runs at 48 kHz (only).

6. Select input and output channels [Top]

Quit REW and start it again. This is to ensure that REW has the latest set of channels in the ASIO4ALL driver. In the Input drop-down selector, select the first channel.

REW will at this point ask you to locate the UMIK-1 calibration file that you downloaded in Step 1. Click on Yes and locate the file.

In the Outputdrop-down selector, you should see eight channels - these are your eight HDMI audio channels. Set it to the first channel:

6. Test and measure [Top]

Before proceeding, turn down the volume on your A/V preamp or receiver.

You can use the REW Signal generator to test that all is working correctly. Click on the Generator button and set the signal generator to "Speaker Cal" pink noise and the level at -12 dB, as shown in the screenshot below. Press the green play button. Gradually increase the volume on your A/V preamp or receiver - you should hear sound coming from the front left speaker.

To check all channels, change the Output drop-down selector in the REW Preferences window. (You can leave the Preferences window open to make output channel selection easier.) Here is a typical mapping of HDMI channel numbers:

• Channel 1: Front Left
• Channel 2: Front Right
• Channel 3: Center
• Channel 4: Subwoofer
• Channel 5: Surround Left
• Channel 6: Surround Right
• Channel 7: Surround Back Left
• Channel 8: Surround Back Right

Open the REW SPL Meter, and click on the red button to turn it on. Increase the volume in your A/V receiver or preamp until the level reads about 75 dB.

You are now all set to run a measurement sweep. Select the first HDMI output channel again and click on the Measure button. Then use Check Levels and then Start Measuring. You will hear a sweep through the front left speaker, and REW will display its first frequency response graph. Then change the output channel to run a measurement on the next channel.

The UMIK-1, like all measurement microphones, is slightly directional and at high frequencies (6 kHz and above) is slightly less sensitive to the sides. For highest accuracy and repeatability, point the UMIK-1 at each speaker being measured. Alternatively, orient the UMIK-1 vertically (pointed at the floor or ceiling) and use a 90-degree calibration file such as that is supplied when the UMIK-1 is purchased via Cross-Spectrum Labs

After completing all channel measurements, you can view them together in the Overlays window. (Apply some smoothing to the graphs using the pop-down Controls overlay.)

What's next? [Top]

Now that you have the ability to run acoustic measurements, you can proceed to optimize and equalize your system. The following app notes contain information that you may find helpful in your journey:

• Acoustic measurement with the UMIK-1 and REW
• Auto-EQ tuning with REW
• AcourateDRC

In this application note, we will show you some basic acoustic analysis techniques that you can apply to measurements of your listening room made with the UMIK-1 and Room EQ Wizard (REW).

• Requirements and setup
• Frequency response measurements
• Reverberation time
• Time-frequency plots
• Further Reading

Requirements and setup [Top]

1. A miniDSP UMIK-1.
2. Room EQ Wizard (REW). Be sure to download the latest version from the Downloads Area for UMIK-1 support.
3. Optionally, a microphone stand with boom arm (the UMIK-1 is supplied with a small stand that can be rested on a table or the back of a sofa.)

Please use one of the following app notes to get configured with the UMIK-1 and REW:

• UMIK-1 setup with REW
• UMIK-1 setup with REW and HDMI (Windows)
• UMIK-1 setup with REW and HDMI (Mac)

Frequency response measurements [Top]

If you have not already done so, start REW and click on the Measurement button. Check the sweep settings and click the Start Measuring button. REW will generate the sweep tone and capture the acoustic signal from the UMIK-1. The main REW window will now look something like this:

(If the Phase checkbutton at the lower right is checked, uncheck it. The Phase measurement is a more advanced concept that we won't cover in this app note.)

The main measurement plot shows a frequency measurement graph in red. At the left is a small "thumbnail" version of the measurement - each time you make a new measurement, a new thumbnail will appear here. At the top of the screen are buttons that access various tools and immediately above the measurement is a row of tabs that show different "views" of the measurement.

Let's do a few basic operations on this first measurement:

1. Give the measurement a meaningful label. Click in the text window next to the thumbnail and type in a label such as "Left spkr in-room". (By default, the label is the date and time the measurement was taken.)

2. Change the scale of the measurement. The default scale is typically too large in the vertical direction, so click on the Limits button (towards the top right) and set the dB scale like this:

   

3. Change the color of the plot. Click on the "brush" icon next to the thumbnail and select a color.

4. Add some smoothing to the graph. From the Graph menu, select "Apply 1/6 Octave Smoothing."

The result now looks like this:

Much better! You can click on the little camera icon to create an image that is better for viewing on the web and for sharing on forums. Here is the image generated using this feature:

Now we can examine in more detail what we have here. The scale along the bottom shows the frequency range, in this case 20 Hz to 20 kHz (20,000 Hz). At each frequency, the UMIK-1 has picked up a sound pressure level (SPL), which is the vertical height of the graph at that point. The scale at the left shows the SPL, which we have set to go from 50 to 90 dB. For example, at 100 Hz the SPL is just under 75 dB.

(The scale at the right of the graph is the phase values. Since we are not looking at phase in this app note, this can be ignored.)

Why is the graph so uneven? Shouldn't we have a flat frequency response? Mostly, it's because of the room. Room modes (resonances) and reflections in a room act to make a nice flat loudspeaker response into a roller-coaster.

Recall that we applied smoothing to the graph. An in-room measurement has a lot of reflections that create peaks and dips in the frequency response. Above 250 Hz or so, much of this is not relevant to what we hear, so for full-range measurements, 1/3 or 1/6th octave smoothing is generally used. For graphs limited to the low frequency range (up to say 300 Hz), use 1/24, 1/48 or no smoothing.

miniDSP offers products that can be used to equalize a system or apply full-range room correction, such as the nanoDIGI 2x8 (digital in-out, parametric EQ), miniDSP 10x10 HD (multichannel analog in-out, parametric EQ), and OpenDRC-DI and OpenDRC-AN (digital or analog in-out, FIR filtering and room correction). Note that equalization for room correction should generally make use of multiple measurements over the listening area, as measured response will change with microphone position.

Reverberation time [Top]

Sound generated in a room decays over time. You've probably been in a "live" hall where footsteps echo and a handclap can be heard to decay over seconds. In other rooms, sound decays very quickly. A convenient measure for this is the reverberation time, or the time that a transient sound takes to decay by 60 dB from its initial level. For home listening rooms, this measurement can be considered relevant above about 200 Hz.

In REW, click on the RT60 tab above the main plot. Deselect all plots except T30 and Topt. (For more information on the various plots, see the REW Help.) Set the frequency range from 100 Hz to 10,000 Hz, and the time range from 0.0 to 1.0 seconds. The resulting plot for the example measurement is:

Recommended values of reverberation time depend on the type of system (e.g. music or home theater), the size of the room, and the type of speaker. However, 300 ms to 500 ms (0.3 to 0.5 seconds) is a commonly recommended range. It can be seen that this room is within this range but towards the high end for a good part of the frequency band.

DSP equalization can't change reverberation time. Heavy drapes and carpets will reduce reverberation time, but if you need to go further, then dedicated acoustic treatment may be required.

Time-frequency plots [Top]

REW offers a set of plots that show how different frequencies decay in time. These are generally used in the "modal region" of the room's acoustic response, from say 20 to 300 Hz, where resonances in the room dominate the measured frequency response. A well-known plot of this type is the "waterfall" plot, or cumulative spectral decay:

This plot shows that there are slowly-decaying resonances around 26, 30, 57 and 74 Hz. With this type of graph, you may need to adjust some display parameters on the Controls and Limits popup windows to get the most useful and informative plot.

Resonances like this can be addressed with a range of techniques: by equalization (if they show up as peaks in the frequency response measurement), by moving and placement of speakers or subwoofers, by the use of multiple subwoofers, and by extensive bass trapping.

A plot that shows the same type of information in a different way is the spectrogram. You may find this plot easier to interpret in some cases — in particular, it can be useful to highlight resonances after equalization has been applied to address measured frequency response peaks.

With all of these plots, you can click on the camera button and save the plot as an image. This will make it easier to share your measurements with others and for "before and after" comparisons.

Further reading [Top]

See the following app notes for information on addressing room problems with miniDSP products:

• Auto-EQ tuning with REW
• AcourateDRC
• Using UMIK-1 with Amarra Symphony

The following links contain additional information on room acoustics and measurement that you may find helpful:

• Acoustic Measurement Standards for Stereo Listening Rooms
• Hard proof that equalization kills room modes
• Bass Integration Guide

Using UMIK-1 with Amarra Symphony - Quick-start guide

Using UMIK-1 with Amarra Symphony: Quick-start guide

Amarra Symphony with IRC is an advanced music file player for the Apple Mac platform. It includes a feature called "Impulse Response Correction" (IRC), that corrects for acoustical errors at the listening position caused by early reflections from the speaker and listening room. It also provides the ability to tailor the response at the listening position using target curves.

The UMIK-1 is an ideal low-cost way to perform the acoustic measurements necessary for Amarra Symphony with IRC. In this app note we'll show you how to do this. We will assume that you already have Amarra Symphony configured to play music files through your stereo system.

What you will need

• Amarra Symphony with IRC. The Amarra Symphony with IRC installation includes a program called IRC Measure, which performs acoustic measurements and filter calculations.
• A miniDSP UMIK-1. After receiving the UMIK-1, go to the UMIK-1 page to get the calibration file for your unique serial number, and save it as a .TXT file.
• A microphone stand with boom arm, available in music supply outlets. (While the UMIK-1 is supplied with a small "table top" stand, the IRC Measure program requires that the microphone be positioned in several locations around the listening area).

1. Getting set up

You will need to already have Amarra Symphony playing music files from your Mac through your audio system. Typically, this will be via a USB DAC, a USB-SPDIF convertor and a DAC, or the Mac's optical digital output into a DAC.

On the input side, the UMIK-1 can be simply connected to any available USB port on your Mac. Position the computer and stand so that the mic can be moved into several positions, as will be seen in the IRC Measure interface (below).

2. Configure IRC Measure

Double-click on the IRC Measure application to run it. (It is located in the /Applications/Amarra Symphony folder.) On the left are five clickable icons that select different tabs for configuring, measurement, and filter generation:

First, select the output device that you will be using from the Sound System Sonic tab:

On the Mic config tab, select the UMIK-1. Click on the Load File button and locate the unique microphone calibration file that you downloaded from the UMIK-1 webpage.

3. Set levels

On the Output and levels tab, set the output volume to low. Click on the Test button for the left channel and gradually increase the output volume until it is at a moderate level, such that your voice would have to be raised to converse with someone sitting next to you.

Now increase the input volume so that the blue level bar is about near the bottom of the green section of the level meter for the left channel. With the UMIK-1, you will probably need to set the input volume at maximum - this is normal. The level can be a little lower than the green bar, but if it is too low, IRC Measure will issue a warning when running a measurement. In that case, you will need to increase the output level.

Repeat the test signal for the right channel. The level should be correct without any further adjustment needed.

4. Run the measurements

You are now ready to run the acoustic measurements! IRC Measure uses nine measurements spread around the listening position to calculate its correction filters.

On the Measurements tab, select the most appropriate listening setup for you (chair, sofa, or auditorium). Position the microphone at the location indicated by the arrow. Be sure to check both the top and front views using the selector underneath the graph, so that you have the height of the microphone correct.

Then click on the Start button. IRC Measure will run three measurement sweeps, through the left speaker, then right, then left again, and display the measurement result as a plot.

If there are errors in the measurement, such as clipping or low levels, IRC Measure will warn you and you will need to go back to the Output and levels tab to make adjustments. If all is well, IRC Measure will ask you to move the microphone to the next location, after which you can press Start again run another measurement. Proceed methodically through all nine measurements.

At this point, you should save your project by clicking on the Save... button.

5. Generate correction filters

On the Filter Design tab, you will initially see the average of the measurements for the left and right channels. You can also opt to display all individual measurements, which shows just how much variation there is across your listening area!

.

Also displayed is a target curve. This is the desired in-room frequency response after correction. Typically, target curves have a small boost in the bass region, and a gentle fall to the extreme treble. Click on Auto Target to have IRC Measure estimate a suitable target curve. You can adjust the target curve by clicking and dragging on the circular grab-points. Double-clicking on the curve will create another grab-point. Drag-selecting a region will zoom in on that region of the graph, and double-clicking will zoom back out again.

You may find you need to experiment with different target curves to determine what works best for your system in your room, as there is no universally "correct" in-room response. Once you have the target curve set, click on the Optimize button. This will generate the correction filters and display the predicted response.

Click on Save Filter to save the correction filter as a file for import into Amarra Symphony. The default location is ~/Library/Application Support/Dirac/Filters, but you can save the file to another location of your choosing.

6. Load and listen!

In Amarra Symphony, click on the IRC button:

This opens the IRC window. Double-click on one of the slots to select and load a correction filter. You can load up to four correction filters into Amarra Symphony. This allows for easy auditioning of different target curves and enables you to keep different correction curves loaded for different purposes.

Now start playing a file, sit back and listen! You can now experiment with different target curves to determine what works best with your system and room. Note that if you move your speakers, or add acoustic treatments to your room, you will need to redo the measurements, so keep your UMIK-1 handy!

After getting used to the basics of in-room measurement with the app note UMIK-1 setup with REW, you may want to perform some more advanced measurements. In this app note, we will explain some techniques that will help with your active loudspeaker design. We will assume that you're already up and running with the UMIK-1 and REW.

• What you will need
• Gated measurements
• Viewing phase
• Low-frequency measurement
• More info

What you will need [Top]

1. A miniDSP UMIK-1.
2. Room EQ Wizard (REW). Be sure to download the latest version from the Downloads Area for UMIK-1 support.
3. A microphone stand with boom arm is recommended for loudspeaker measurements.

Gated measurements [Top]

An acoustic measurement with the microphone at the listening position is generally referred to as an in-room measurement. You are measuring the combined effect of the loudspeakers (and/or subwoofers) and the acoustics of the listening room. That is fine for assessing room acoustics, but when doing loudspeaker design, you want to measure the loudspeaker only, without any effects from the room. This figure illustrates the problem:

The sound from the speaker arrives at the microphone first, followed by a reflection from the floor, then from the ceiling, and somewhere along the way from the side walls. These reflections continue to bounce around the room for up to several hundred milliseconds. That is what we hear, but it makes it hard to get an accurate measurement of just the loudspeaker! To illustrate, here's a measurement of a 3-inch full-range driver taken with the UMIK-1 and REW in a typical domestic environment:

Obviously, without a little work, this is fairly hard to use! One way to address the problem is to use an anechoic chamber, which is a very large room with lots of acoustic absorption in it so there are almost no reflections. Most of us don't have one of those. Another technique is to use a gated measurement, where the reflections are simply removed from the measurement. We can do this in REW.

First, let's understand what we are looking for. The measurement was taken with the driver mid-way between the floor and ceiling at 120 cm from each, and the microphone level with the driver and 1 m distant. According to the floor-bounce calculator at mehlau.net, we would expect the first reflection to arrive at the microphone 4.65 ms after the direct signal.

Let's see if we can see this first reflection in our example measurement. In the main REW window, click on the "Impulse" button, and then hover the mouse over the display and select the "%FS" dropdown (you may need to adjust the graph limits to get a good view of the impulse response):

Yep! There it is... To eliminate this reflection (and all subsequent reflections), we will gate the signal at 4.4 ms as shown by the marker. Go to Tools→IR Windows, set the "Right Window" to 4.4 ms, and then click Apply Windows. The "window" is the region of time that REW uses to calculate the frequency response. The display will update to show the new window, which reduces the signal so that it is zero by 4.4 ms (in blue):

Clicking back on the SPL & Phase button will now show the gated measurement — that is, with the reflections removed from the frequency response calculation. Here it is (shown on the Overlay screen) in green:

Another technique that can be used to make the raw measurement graph more usable is smoothing. This function is accessed from the Graph menu when viewing the main window, or from the Control dialog on the Overlays window. For example, the graph above shows, in red, the same measurement with 1/12 octave smoothing (and without gating). While smoothing can sometimes be used to give a similar effect as gating, this example shows that you must be careful: the peaks at 14 and 17 kHz have been smoothed over and now are not accurate.

Gating does have limitations. For one thing, the frequency resolution of the measurement is reduced—the IR Windows dialog shows this resolution. In addition, any frequency below about 1 / (window length) cannot be represented. In the case of the 4.4 ms window, the lowest frequency that can be represented correctly is 1/0.0044 or about 230 Hz - anything on the graph below this frequency must be ignored. So this technique works best at high frequencies and can't be used for low-frequency measurements at all.

Viewing phase [Top]

The phase of the acoustic output from a driver is useful to know in some cases. For example, when looking for the causes of poor crossover integration, or when performing overall phase corrections to obtain a linear-phase loudspeaker (see the rePhase FIR tool app note). To view the phase of a measurement, click on the "Phase" checkbox in the main REW window. (Or alternatively, go to the Overlays window and click on the Phasebutton.)

Once there, use the Limits and Controls dialogs to adjust the graph. Typically, the Unwrap Phase button will be used to remove discontinuities in the phase display, and the +360 and -360 buttons can be used to put the display into a usable range. Here is the result for the measurement above (the phase scale is over on the right):

Note: you may not get a clean phase reading when the default 500 ms window is used. If so, gate the measurement as described above to remove the reflections, and the display should now show the correct phase of the driver or loudspeaker being measured.

Low-frequency measurement [Top]

As noted above, gated measurements are only relevant at higher frequencies, typically above a few hundred Hz. Even with in-room measurements, you may be able to design and equalize your speaker by judicious use of smoothing and taking care to mentally separate room effects from the loudspeaker. For better accuracy, though, here are some techniques you can use:

Nearfield measurement. The microphone is moved close to the driver being measured so that the relative level of room sound is reduced. With the mic right near the cone, this can be effective for subwoofers. For full-range measurement, the response will change with distance, and so the mic typically needs to be a few times the cone diameter away. Even then, the effect of the cabinet on the response may not be fully visible.

Outdoor measurement. Room modes will be eliminated if measurements are taken outdoors. It is still important to minimize reflections though. This can be done by moving the speaker and microphone well away from any walls or other large objects and raising them both as far off the ground as possible.

Ground plane measurement. This is usually done outdoors as well, to avoid room modes. The microphone is placed right on the ground so that there is no reflection from the ground at all. In practice, the mic capsule is still a small distance from the ground, so this technique can be used up to perhaps a few kHz.

More info [Top]

These app notes will help you with your active speaker project.

• UMIK-1 setup with REW
• Building a 2way crossover
• Stereo 3/4 way Crossover

Don't forget to ask on the miniDSP forum if you have questions!

With the miniDSP UMIK-1, you have an easy way to get started with acoustic measurement: to optimize and equalize your subwoofer, to correct for room effects, or even to design your own loudspeakers! In this application note, we'll show you how to get up and running with the UMIK-1.

1. Get connected

Mount the UMIK-1 into the small stand supplied with it, or if you like, you can use any other microphone stand. Connect the UMIK-1 to your computer using the supplied USB cable.

Connect an audio output from your computer to your sound system. With most computers, you can use the line out or headphone output with a suitable cable. It is advisable to connect into your system at the preamp inputs, so that the preamp volume control can be used to manage the signal level from the computer.

2. Get your calibration file

Go to the UMIK-1 page and enter your microphone's serial number. It is in the form xxx-yyyy and labeled on the microphone. Use "Save As" in your browser to save the numbers as a file e.g. 7000343.txt.

The calibration file ensures that your microphone is as accurate as possible. Each microphone has a unique calibration file, which is why the serial number must be entered.

"Sens Factor =-16.273dB, SERNO: 7000343"	
20.396	-0.376769317
22.242	0.434645133
24.255	0.819522375
26.4503	0.869646823
...

3. Set up Room EQ Wizard

Download Room EQ Wizard from Home Theater Shack. You should get the latest version from the download area as at least 5.01 beta 14 is needed to operate with the UMIK-1.

Once installed, start REW. You will see a screen asking if you want to use the UMIK-1. Click on Yes. (If you don't get this screen, go to the REW Preferences window and set the Sample Rate to 48 kHz.)

Answer Yes to the next question about the calibration file, and then locate the file that you saved above.

4. Set levels

The UMIK is automatically calibrated by REW for sound level (this information is in the calibration file). You will want to set your system to generate a suitable signal level, though. Click on the Signal Generator button and set the parameters like this (the top frequency parameter, "24,000" in this example, isn't relevant, you can ignore it):

Position your microphone at the listening position and turn the volume of your system down. Then click on the Play button (green triangle), and turn the volume up until the test signal is at a comfortable level.

Now open the REW SPL Meter. Click on the red button in the lower right corner to turn it on, and adjust your system volume until the meter reads about 75 dB.

5. Run a measurement sweep

Click on the Measure button near the top left of the main REW Screen. Check that the level is set to -12 dB, and click on the Start Measuring button.

REW will make a "whoo-oop" sound through your speakers. A short time later, you should see your first in-room measurement!

Note: if you're measuring primarily bass frequencies, it's better to run both speakers at the same time. If you're measuring full-range and looking at from, say, 200 Hz and up, it's better to run just one speaker - just disconnect one of the cables. Depending on the type of measurement, you may also want to apply some smoothing to it from REW's "Graph" menu.

What's next?

You're well on the way now! Now that you're able to do an acoustic measurement of your speakers and room, there are lots of interesting things you can explore. Check out the following app notes:

• Loudspeaker measurements with REW. Learn how to perform basic loudspeaker measurements. A great starting point for anybody who just acquired the UMIK-1.
• Subwoofer Integration with miniDSP. Use the measurement capabilities you now have to accurately integrate your subwoofer into your system. No more guesswork!
• Auto-EQ tuning with REW. Use the Auto-EQ feature of REW to generate correction curves for your system based on your UMIK-1 acoustic measurement. Just load the result right into a miniDSP plugin.
• Building a 2way crossover. Use the UMIK-1 to measure your speakers accurately in order to make an active 2-way (or 3-way, or 4-way) loudspeaker system.

The miniDSP PMIK-1 is the perfect pocketable companion to audio analysis programs running on your tablet or even your smartphone. In this application note we will show you how to set up the PMIK-1 for use with the AudioTool app for Android.

Please note: miniDSP cannot provide support for third-party applications or hardware. This app note shows you how to set up the miniDSP PMIK-1 to use with AudioTool but other functions of the Android device or the AudioTool app are beyond the scope of miniDSP support.

• 1. Get your calibration file
• 2. Get connected
• 3. Calibrate
• 4. Measure!

1. Get your calibration file [Top]

Go to the PMIK-1 page and enter your microphone's serial number. It is in the form xxx-yyyy and labeled on the microphone.

The calibration file ensures that your microphone is as accurate as possible. Each microphone has a unique calibration file, which is why the serial number must be entered.

*1000Hz	-32.1	
20.00	-3.9
20.55	-3.6
21.11	-3.2
...

2. Get connected [Top]

The PMIK-1 simply plugs into the headphone jack of the Android device. The microphone body is oriented at 90 degrees to the plug, which makes it easy to use your Android device in landscape mode and point the microphone towards the sound source:

As suggested by the photograph, we recommend that for the most reliable and repeatable measurement results, the included wind/pop filter be removed from the PMIK-1. If the pop filter is used, be aware that the response above 10 kHz will read up to a few dB low.

To generate audio output from the Android device for the purpose of measuring system response, connect a cable from the rear of the PMIK-1 to the system being tested. At the PMIK-1 end, the cable will have a 3.5mm stereo jack. A 3.5mm to RCA socket adapter is included with the PMIK-1 to make it easy to use regular RCA cables to connect to the system.

3. Calibrate [Top]

If you haven't already, install the AudioTool app on your Android device from the Google Play Store.

On your computer, you will need to copy the downloaded calibration file into the AudioTool folder. (Before doing this, be sure to start the app once so that the folder is created.)

1. Connect the Android device to your PC or Mac with a USB cable.
2. On Windows, you should be able to access the Android file system directly. On the Mac, you will need to use the Android File Transfer program.
3. Copy the calibration file into the Download folder on the Android device. (If the AudioTool folder is directly visible, then you can copy it directly into there and skip the next step.)
4. Install the OI File manager app on the Android device and move the calibration file from the Download folder into the AudioTool folder.

In the app, click on the options icon (three vertical dots) and select Load Cal. Click on the name of the calibration file. This will load the frequency correction data into AudioTool.

(If you do not see the Load Cal menu entry, click on Use 1/3 Octave Calibration. Re-opening the menu will then show the Load Cal entry.)

Because different devices will have different sensitivity on the microphone input, it is impossible to provide a standard adjustment to get correct SPL readings. You will need to use an external sound source with a known SPL and adjust the Global Offset parameter, as described in the AudioTool documentation.

4. Measure! [Top]

Now we will generate a test signal to feed into the system. Tap on Gener in the bottom right corner. On the signal generator screen, select Pink (pink noise) and then tap on the Off button (which turns the signal generator on). Note: be careful with the volume!

Tap on Analyser to get back to the analyser screen. Click on the third button from the left in the top row to cycle through the different analysers. Here is the 1/3 octave RTA showing the response of the system:

You can now explore the different types of measurement and analysis tools available in AudioTool. Please refer to the AudioTool documentation, and have fun with your PMIK-1!

The miniDSP PMIK-1 is the perfect pocketable companion to audio analysis programs running on your tablet or even your smartphone. In this application note we will show you how to set up the PMIK-1 for use with the AudioTools app and Smaart® Tools from StudioSix Digital, running on an Apple iPad.

Please note: miniDSP cannot provide support for third-party applications or hardware. This app note shows you how to set up the miniDSP PMIK-1 to use with AudioTools but other functions of the Apple iPad hardware or the AudioTools app are beyond the scope of miniDSP support.

• 1. Get your calibration file
• 2. Get connected
• 3. Calibrate
• 4. Measure!

1. Get your calibration file [Top]

Go to the PMIK-1 page and enter your microphone's serial number. It is in the form xxx-yyyy and labeled on the microphone.

The calibration file ensures that your microphone is as accurate as possible. Each microphone has a unique calibration file, which is why the serial number must be entered.

*1000Hz	-32.1	
20.00	-3.9
20.55	-3.6
21.11	-3.2
...

2. Get connected [Top]

The PMIK-1 simply plugs into the headphone jack of the iPad. The microphone body is oriented at 90 degrees to the plug, which makes it easy to use your iPad in landscape mode and point the microphone towards the sound source:

As suggested by the photograph, we recommend that for the most reliable and repeatable measurement results, the included wind/pop filter be removed from the PMIK-1. If the pop filter is used, be aware that the response above 10 kHz will read up to a few dB low.

To generate audio output from the iPad for the purpose of measuring system response, connect a cable from the rear of the PMIK-1 to the system being tested. At the PMIK-1 end, the cable will have a 3.5mm stereo jack. A 3.5mm to RCA socket adapter is included with the PMIK-1 to make it easy to use regular RCA cables to connect to the system.

3. Calibrate [Top]

If you haven't already, install the AudioTools program on your iPad from the App Store.

Open the app and go to the Settings menu, and then select Microphone Setup. You will see see two choices: Low Range and High Range. Select Low Range.

Then tap on the "i" icon to the right. This will bring up the microphone calibration screen.

In the screenshot above, the name of a PMIK-1 calibration file is shown. Initially, however, this will be blank, so you will need to load the calibration file to the iPad and select it. Tap on the Calibration File link and then on the Files link at the bottom right. This will bring up a screen with information on how to transfer the calibration file from your computer to the iPad:

On your computer's browser, enter the address shown by AudioTools on the above screen and upload the calibration file to the iPad.

Back in the AudioTools interface on the iPad, tap Done and then on the name of the calibration file and tap Apply. The calibration screen will change to display the name, as shown above.

The Trim parameter can be adjusted for correct SPL readings. Unfortunately, because different devices will have different sensitivity on the microphone input, it is impossible to provide a standard adjustment to get correct SPL readings. You will need to use an external sound source with a known SPL. With our iPad mini, we used the value 2.5, so you could enter that as an approximation if you wish. Then tap on Done to close the calibration screen, and on Done again to go back to the main AudioTools menu.

4. Measure! [Top]

AudioTools has a number of modules. A powerful tool for measuring system response is the FFT tool. Tap on the Acoustics icon at the left and then on the FFT icon. To measure system response, we will need to generate a pink noise test signal. Tap on the small sine wave icon at the bottom of the screen. A control panel will pop up; set it for pink noise and tap the button at its top left. (Note: make sure to turn the iPad output volume or your system volume down first, and then increase it gradually.)

With pink noise playing through the system, the FFT shows the frequency response of the system as measured in the room. You can adjust the frequency resolution using the selector at the lower left.

To use Smaart® Tools, you will need to purchase the Smaart® plugin from within AudioTools. From the home screen, simply tap on Acoustics then on Smaart® Tools I. You will be asked for your App Store credentials so you can make the purchase.

With that done, click on the spanner/wrench icon near the bottom of the screen to set the parameters. Here is a typical example set of settings:

Below is an example screen showing the Smaart® module in action. Here, we have enabled the dual display mode, and are displaying an RTA, spectrum, and a spectrogram while monitoring music playback:

The miniDSP PMIK-1 is the perfect pocketable companion to audio analysis programs running on your tablet or even your smartphone. In this application note we will show you how to set up your PMIK-1 with the SignalScope Pro app from Faber Acoustical, running on an Apple iPad.

Please note: miniDSP cannot provide support for third-party applications or hardware. This app note shows you how to set up the miniDSP PMIK-1 to use with SignalScope Pro but other functions of the Apple iPad and SignalScope Pro are beyond the scope of miniDSP support.

• 1. Get your calibration file
• 2. Get connected
• 3. Calibrate
• 4. Measure!

1. Get your calibration file [Top]

Go to the PMIK-1 page and enter your microphone's serial number. It is in the form xxx-yyyy and labeled on the microphone.

The calibration file ensures that your microphone is as accurate as possible. Each microphone has a unique calibration file, which is why the serial number must be entered.

*1000Hz	-32.1	
20.00	-3.9
20.55	-3.6
21.11	-3.2
...

2. Get connected [Top]

The PMIK-1 simply plugs into the headphone jack of the iPad. The microphone body is oriented at 90 degrees to the plug, which makes it easy to use your iPad in landscape mode and point the microphone towards the sound source:

As suggested by the photograph, we recommend that for the most reliable and repeatable measurement results, the included wind/pop filter be removed from the PMIK-1. If the pop filter is used, be aware that the response above 10 kHz will read up to a few dB low.

To generate audio output from the iPad for the purpose of measuring system response, connect a cable from the rear of the PMIK-1 to the system being tested. At the PMIK-1 end, the cable will have a 3.5mm stereo jack. A 3.5mm to RCA socket adapter is included with the PMIK-1 to make it easy to use regular RCA cables to connect to the system.

3. Calibrate [Top]

If you haven't already, install the SignalScope Pro program on your iPad from the App Store.

Then open the I/O Configuration screen by tapping on the "Microphone" icon at the top left of the screen. Change the Device Units to FS and set Gain to Low:

Then tap on Input Channels. If under All Channels it does not say "Units (Pa)", tap on "Units" and select Pa from the list and then tap on Back.

You will need to use iTunes to copy the downloaded file to the correct location on the iPad. Use the iTunes Apps window to locate SignalScope Pro, and copy the folder Frequency_Response_Data to your computer. Copy the calibration file into the downloaded folder, then in iTunes, upload the whole folder back to the iPad.

Tap on Frequency Response Data, then Load FR Data From File, and select the uploaded file. The lower part of the screen will fill with the file preview. On that screen, tap on Load Frequency Response Data. You will get a confirmation dialog and the calibration curve of the PMIK-1 will be displayed:

Then tap on Calibrate. Unfortunately, because different devices will have different sensitivity on the microphone input, it is impossible to provide a standard adjustment to get correct SPL readings. You will need to use an external sound source with a known SPL. With our iPad mini, we used the value 23 in Input Sensitivity, so you could enter that as an approximation if you wish.

4. Measure! [Top]

You can now start to measure! We've covered the specifics in our app note for the UMIK-1 with SignalScope Pro, so please jump to that app note to do more with SignalScope Pro and the PMIK-1. Here are a couple of examples.

Introduction

Siegfried Linkwitz is a well known reference in the Audio industry. His work range from filtering to speaker building and has been referenced/used by many designers worldwide for the past few years. The Linkwitz Lab website is a goldmine of information on advanced filtering, speaker building and even basic concepts of acoustic and speaker design. A reference website for any audio designer!

The latest addition to Linkwitz Lab portfolio is the LXmini, an alternative to the Pluto 2.1 design and built around a miniDSP 2x4 platform. Fully engineered by Siegfried himself and his legendary handmade pole/zero diagrams, this new design is yet another

The LXmini is a most remarkable loudspeaker. It converts electrical signal voltages into acoustic pressure variations, which are perceived as completely neutral and detailed even in a reverberant environment. With this design I want to give every music lover the opportunity to build and enjoy a reference quality sound system singing in their own living space.

Siegfried Linkwitz

LXmini in Room

Woofer equalization

Frequency response measurements ON/OFF axis

Design

As you can see from below cost, the ease of construction and the price put the LXmini within reach of many audiophiles and music lovers.

• LXmini Construction Plans:    $105 USD
  
• 4 SEAS drivers:    $320 USD
  
• miniDSP 2x4 in a box:    $115 USD
  
• USB power supply:    $15 USD
  
• PVC/ABS pipes, couplers, cups:    $60 USD
  
• Hardboard, Plywood, Rod:    $25 USD
  
• Screws, spade lugs, cable clamps, terminal blocks, grip handle:    $20 USD
  
• 60 ft of18 Gauge speaker cable, plugs:    $30 USD
  

More Info

• The LXmini Construction Plans can be ordered from This email address is being protected from spambots. You need JavaScript enabled to view it. after 17 July 2014 and a whole lot of information can be found directly from Linkwitz Lab LXmini page. 
• Want to get up and running in no time? Purchase a complete LxMini Kit from Madisound by clicking on the picture. 

John Reekie from Hifizine, the enthusiast's audio webzine, documented an amazing design in the following 2 part articles. This is a prototype speaker dubbed the "Mini Convertible," which uses drivers from Seas of Norway (http://www.seas.no/) and a miniDSP PWR-ICE125 plate amp for amplification and DSP crossover.  The prototype has a removable baffle for experimentation with different baffle profiles and cabinet bracing, but the final speaker will have the baffle glued (i.e. no screw holes).

The DIY speaker project is written up in a series of three articles covering design and prototyping, construction details, and detailed measurements and crossover configuration.

Terms of use

This design is released under the Creative Commons Attribution license. That means, in a nutshell, that you can use it and modify it as you like, including commercial use of the design, provided that you provide attribution to the original source with this link: http://www.hifizine.com/issues/mini-convertible.

• Attribution 4.0 International Licence Deed
• Attribution 4.0 International Legal Code

Siegfried Linkwitz is a well known reference in the Audio industry. His work range from filtering to speaker building and has been referenced/used by many designers worldwide for the past few years. The Linkwitz Lab website is a goldmine of information on advanced filtering, speaker building and even basic concepts of acoustic and speaker design. A reference website for any audio designer!

It's no news that our platforms have already been used many times to perform filtering and processing of few Linkwitz Lab speaker designs. We are always amazed at the creativity of our members and once again give you kudo for your work! Few weeks ago, Dave Reite certainly took it one step further. By building a complete DSP configuration for the LX521 monitor and submitting it to Mr Linkwitz for review, a new wave of oportunities opened up for the miniDSP community trying to build the LX521. Here is a bit of the story of the LX521 powered by miniDSP with all Kudos to Dave Reite and Siegfrid Linkwitz for giving our platform a good challenge! For more info, please check the DSP challenge here on Linkwitz Lab website.

After more listening by myself and an experienced audio professional I can state that Davey's LX521 DSP prototype is without reservations an excellent alternative to the ASP in a stereo system application.

Siegfried Linkwitz

Baffle LX521 monitor	Finished implementation
System Diagram with nanoDIGI 2x8
Complete prototype by Dave Reite using 1 x nanoDIGI 2x8 + FiiO DACS
miniDSP 4x10 stacked on Linkwitz Lab setup

Parts/software used for construction of the All Digital Setup:

• nanoDIGI 2x8 K or nanoDIGI 2x8 B
• nanoDIGI 2x8 plug-in (programming software)
• FiiO D03K DAC (4x), ( FiiO D07 for +3 dB out?)
• Multichannel amplifier
• Power supply, 5V, 500 mA
• RCA cables (4x, short length)
• Proprietary LX521.xml file provided by Linkwitz lab on purchase of plans

Parts/software used for construction of the Analog + Digital setup

• miniDSP 4x10 HD
• miniDSP 4x10 plug-in (programming software)
• Multichannel amplifier
• Audio cables (4x, short length)
• Proprietary LX521.xml file provided by Linkwitz lab on purchase of plans

Credits

All credits for building and engineering this nanoDIGI 2x8 solution goes to Dave Reite. All credits for pictures of LX521 monitor, diagrams and other technical aspects of the designs go to Linkwitz Lab.
We'd certainly like to thank both of these gents for their great support to the miniDSP community! :-)

@miniDSP, we value creative entrepreneurs building products/solutions around the miniDSP platforms. From a garage company to a well known established venture, we love commercial implementations making use of our technology.
The "Powered by miniDSP" program is our way to advertize products and make sure they get the exposure they deserve.

If you have a commercial product based on miniDSP technology and need our help to advertise it, please This email address is being protected from spambots. You need JavaScript enabled to view it. with us.

Music & Design

The NaO Note II RS is the newest offering from Music and Design. Advancing the proven performance of the NaO II, the Note II RS is a 4-way, active, full range dipole system which uses the miniDSP 2x8 or 4x10 HD digital crossover and can be build for a modest cost. The driver set consists of front and rear Vifa OX20SC00 ¾"; dome tweeters; the ScanSpeak Discovery 10F4424 and 22w8534 serve as the upper and lower midrange drivers; a pair of Peerless SLS 830668 woofers handle the bass. Peerless XXLS 835016 woofer may be substituted at additional cost for increased bass capability. The basic driver set can be purchased for as low as $722 and the system can be build for under $1300. There is no expensive active analog crossover to build and test. Just build the frames, mount the drivers, flash the miniDSP with the supplied configuration files, connect your amplifiers and you are ready to listen. The miniDSP allows easy and rapid changes in the system voicing to meet the listeners taste, environment and source material. The performance of the Note II RS equals or exceeds that of system costing many times more. See more information on Music and Design Page.

Click on below images to see zoomed version.

In this application note, we will show you how to design and build your own DSP-controlled active loudspeaker with the miniDSP ICE-PWR125 DSP plate amplifier. Note that a similar project could be built using the PWR-ICE250 if additional power is required.

• What you will need
• 1. Select the speaker drivers
• 2. Design and build the enclosures
• 3. Get connected
• 4. Configure the amplifiers
• 5. Develop the crossover
• 6. Configure the second speaker
• What's next?

What you will need [Top]

1. Two ICE-PWR125 plate amps OR two PWR-ICE250
2. A set of loudspeaker drivers (see text below).
3. Ability to run acoustic measurements. You will need a measurement program such as Room EQ Wizard (REW) and measurement hardware such as the UMIK-1.
4. Optionally, a microphone stand with boom arm (the UMIK-1 is supplied with a small stand that can be rested on a table or the back of a sofa.)

1. Select the speaker drivers [Top]

If you are starting from scratch, you will need to select the drivers for your speakers. You can use drivers rated at 4 ohms or 8 ohms — 4 ohm drivers will provide greater power output from the amplifier but for most applications it's not critical. For a small two-way loudspeaker, a 5" or 6.5" woofer and a 1" dome tweeter are common choices. Or, you might like to try a larger "pro" style or "Econowave" speaker with a 12" woofer and a horn tweeter. Peruse the online forums to see what others are using and to ask for recommendations for your particular project.

It may also be possible for you to convert an existing speaker from passive to DSP active. In this case, you will need to remove the internal crossover and binding posts and re-cut the enclosure to fit the plate amp.

2. Design and build the enclosures [Top]

If you are building your own enclosure, you will need to design it to suit the woofer you have chosen. The most important factor is the internal volume and, if it's a ported box, the size and length of the port. Fortunately, there are a number of free programs that do the complex math for this based on the Thiele-Small parameters of the woofer. For example, a popular Excel-based program is Unibox.

Once designed, you will need to build the enclosure. You can search online for advice on building speaker boxes, or ask on your favorite online forum. When building the enclosure, you may wish to build a separate sub-enclosure for the plate amplifier. A rebate for the amplifier is optional, but does result in a neater finish.

The drivers will need to be connected to the leads from the amplifier. A convenient way to do this for most drivers is to solder push-fit terminals onto the leads. Or, you can solder directly to the terminals on each driver.

3. Get connected [Top]

Once the speakers are assembled, make the connections to the amplifiers:

• Power, via the supplied IEC cable
• Ethernet, to your local area network, using the supplied Ethernet cable
• Audio, using either analog RCA, analog XLR, or digital XLR

If using digital input, a normal S/PDIF (RCA/coax) digital source can be connected with an RCA to XLR adapter cable. Alternatively, a transformer adapter such the Canare BCJ-XP-TRB with a BNC-RCA adapter can be used. If you have an AES/EBU digital source (pro balanced connection), then an XLR-XLR cable is used. To connect the second speaker, simply connect an XLR-XLR cable from the digital output of the first amp to the digital input of the second amp.

Note that, even if your audio connection is digital, you may need to use analog connections to hook the amplifier input to your measuring system. If you do, make sure to disconnect any digital sources and the link cable while doing your measurements.

Finally, the Ethernet connection is only needed for configuring the amplifiers, and is not needed for normal operation. Once everything has been configured to your satisfaction, the Ethernet cable can be disconnected.

4. Configure the amplifiers [Top]

Double-click on the PWR_ICE2_1x2 application to run it. After a short time, you should see the two amplifiers appear in the "Device Tree".

Initially, the amplifiers will have generic names but you can rename them so that you know which is which. Click on one of the amps in the Device tree and wait for it to load. Then click on the entry box next to Now connected to:, type in the desired name for that amp, and press Return. You will get a message saying that you need to reboot the amplifier - turn the amplifier off and back on again, and it will reappear in the Device Tree with the new name.

While you're there, set both amplifiers for either analog or digital input, depending on how you have them connected. And set one for left channel input, and the other for right channel input. Note that, for a two-way speaker, the Amplifier mode selector must be set to Stereo.

5. Develop the crossover [Top]

The equalization and crossover is done in the two output channel sections to the right of the main screen. We will use Channel 1 for the woofer and Channel 2 for the tweeter - you can provide each channel with individual names by clicking on the default "Output 1" and "Output 2" text and typing over it.

There are two main parts to developing a DSP crossover. The first is to equalize each driver individually, to a flat response, using the PEQ block in each output channel. The second is to add the crossover filter using the Xover block of each channel. The procedures are described in detail in the following app notes:

• Building a 2way crossover
• Loudspeaker measurement with UMIK-1 and REW
• Auto-EQ tuning with REW

6. Configure the second speaker [Top]

Once you are happy with the crossover and equalization for the first speaker, you will need to configure the second speaker. The simplest way to do this is to first save the configuration to a file:

Then, in the Device Tree, click on the other amplifier, and load the configuration file into it. You will then need to change the channel selection button back to the correct channel. Then set the two speakers up in their normal positions in your listening room and listen to the results of your work!

What's next? [Top]

Loudspeaker design is not a simple "straight line" process. You can continue to tweak the EQ and crossovers on your speakers over time. And here are some more advanced things that you can also try:

• Measure room acoustics with your DSP speakers in it. You can use the input PEQ block of each amplifier to add EQ for room acoustics. See the app note Acoustic measurement with the UMIK-1 and REW.

• Build a subwoofer. A third PWR-ICE125/250 amplifier can be used to power a subwoofer that will integrate into a complete DSP-controlled system. See the app note Building a DSP-controlled subwoofer with the PWR-ICE125.

• Linearize the phase of your complete system. This can be done with a miniDSP OpenDRC-DI or OpenDRC-AN. For more information, see the app note The rePhase FIR tool.

In this application note, we will show you how to design and build your own DSP-controlled subwoofer with the miniDSP PWR-ICE125 DSP plate amplifier. Note that a similar project could be built with the PWR-ICE250 if added power is required.

• What you will need
• 1. Select the subwoofer driver
• 2. Design and build the enclosure
• 3. Get connected
• 4. Configure the amplifiers
• 5. Measure and equalize the subwoofer
• What's next?

What you will need [Top]

1. One ICE-PWR125 plate amp OR PWR-ICE250
2. A subwoofer driver (see text below).
3. Ability to run acoustic measurements. You will need a measurement program such as Room EQ Wizard (REW) and measurement hardware such as the UMIK-1.
4. Optionally, a microphone stand with boom arm (the UMIK-1 is supplied with a small stand that can be rested on a table or the back of a sofa.)

1. Select the subwoofer driver [Top]

If you are building your DSP-controlled subwoofer from scratch, you will need to select the driver. There are dozens if not hundreds of good quality subwoofer drivers available for DIY use ranging in size from 6" to 21", but drivers in the 10-15" range are typically a good match for the PWR-ICE125. For higher rating, we'd recommend you to select the PWR-ICE250. It can be used with a 4 ohm driver, but if it will be driven hard or in a pro sound application, an 8 ohm driver is recommended.

It may also be possible for you to convert an existing subwoofer, either from a passive configuration (external amplifier) or to replace an existing plate amplifier. In this case, you will need to remove the internal amplifier or binding posts and modify the enclosure to fit the miniDSP plate amp.

2. Design and build the enclosure [Top]

If you are building your own enclosure, you will need to design it for the subwoofer driver you have chosen. The most important factor is the internal volume and, if it's a ported box, the size and length of the port. Fortunately, there are a number of free programs that do the complex math for this based on the Thiele-Small parameters of the woofer. For example, a popular Excel-based program is Unibox.

Once designed, you will need to build the enclosure. You can search online for advice on building subwoofer boxes or ask on your favorite online forum. When building the enclosure, we recommend that for most applications, a separate sub-enclosure be built for the plate amplifier, and this sub-enclosure have air-flow vents above and below the amplifier for maximum cooling.

For subwoofer use, the PWR-ICE125/250 is run in BTL (bridge-tied load) mode. This means that both channels of the amplifier are used to power the one subwoofer driver, so the two red wires from the amplifier must be connected to the subwoofer terminals. The two black leads can be cut, insulated with electrical tape, or a pointed tool used to remove the center two pins from the 4-way connector that plugs into the amp.

3. Get connected [Top]

Once the subwoofer is assembled, make the connections to the amplifier:

• Power, via the supplied IEC cable
• Ethernet, to your local area network, using the supplied Ethernet cable
• Audio, using either analog RCA, analog XLR, or digital XLR

For use in a conventional two-channel system, the subwoofer amplifier will take its audio signal from both left and right stereo channels. For analog input, use Y-splitters and connect both left and right channels to the subwoofer inputs. If connecting to an A/V receiver or preamp with a Sub Out connection, only a single cable is needed.

For digital input, a normal S/PDIF (RCA/coax) digital source can be connected with an RCA to XLR adapter cable. Alternatively, a transformer adapter such the Canare BCJ-XP-TRB with a BNC-RCA adapter can be used. If you have an AES/EBU digital source (pro balanced connection), then an XLR-XLR cable is used. Note that, even if your audio connection is digital, you may need to use analog connections to hook the amplifier input to your measuring system. If you do, make sure to disconnect any digital sources while doing your measurements.

Finally, the Ethernet connection is only needed for configuring the amplifier and is not needed for normal operation. Once everything has been configured to your satisfaction, the Ethernet cable can be disconnected.

4. Configure the amplifier [Top]

Double-click on the PWR_ICE2_1x2 application to run it. After a short time, you should see the amplifier appear in the "Device Tree" under "Ethernet Device." To rename it, click on the amp in the Device tree and wait for it to load. Then click on the entry box next to Now connected to:, type in the desired name for that amp, and press Return. You will get a message saying that you need to reboot the amplifier - turn the amplifier off and back on again, and it will reappear in the Device Tree with the new name.

If the subwoofer is being driven from a dedicated subwoofer signal, such as the Sub output from an A/V receiver or A/V preamp, select the analog input that you connected to above. If the subwoofer is being integrated into a two-channel system and you connected both left and right input channels, select "Mixed L&R."

Also select analog or digital input, depending on whether you have connected an analog or digital source. Finally, you must make sure to set the Amplifier mode to "BTL." This will reconfigure the DSP to drive the power amplifier as a single channel.

5. Measure and equalize the subwoofer [Top]

With the subwoofer in position in the room, you can now measure its response! Use one of the following app notes to get started with measurements:

• UMIK-1 setup with REW
• UMIK-1 setup with REW and HDMI (Windows)
• UMIK-1 setup with REW and HDMI (Mac)

Equalization can be done in either of the PEQ blocks, using the Auto-EQ function of REW. Make sure to set the Equalizer type in REW to "MiniDSP-96k," as the PWR-ICE125/250 runs at 96 kHz. See the following app note for details on how to use the Auto-EQ function (with the change of equaliser type as just noted):

• Auto-EQ tuning with REW

You can use the Xover block to add a low pass filter, to remove all but subwoofer frequencies. This will be necessary if integrating into a two-channel system, and may or may not be helpful if integrating into a home theater system. You can also add a high pass filter at any frequency down to 10 Hz to remove extreme low frequencies, which can help to protect the driver from over-excursion if you have built a ported subwoofer.

What's next? [Top]

Once you have your subwoofer set up and working to your satisfaction, there are still lots of things you can do! Here are some ideas:

• Experiment with positioning. The location of a subwoofer in the room can have a dramatic effect on the measured response.
• Build another one! Using more than one subwoofer can help to even out room modes, as well as increase levels and headroom.
• Use a Linkwitz transform to extend its low end response (applies to sealed subwoofers only). For more information, see the app note Linkwitz Transform.

This app note shows you how to use the DDRC-88BM plugin with the powerful Multi-Sub Optimizer (MSO) freeware. Multi-sub optimization should be done before performing your Dirac Live calibration or performing any bass management in the plugin.

Please note that MSO is third-party software. miniDSP is not able to directly provide support for this software. For questions or issues specifically related to MSO, please refer to the AVS Forum MSO discussion thread.

• Overview of multiple subwoofers
• What you will need
• Multi-sub system configurations
• 1. Choose subwoofer locations
• 2. Set up signal routing
• 3. Take your measurements
• 4. Export measurements
• 5. Configure MSO
• 6. Load measurements and create a configuration
• 7. Set optimization parameters and run
• 8. Refine optimization
• 9. Export filters
• 10. Import filters into the plugin
• 11. Set gains and delays
• 12. Confirm your results
• Wrapping up
• Appendix A. Notes on using an external 2x4 for the subs

Overview of multiple subwoofers [Top]

The main reason for using multiple subwoofers is to improve evenness of bass response across the whole listening area. While Dirac Live will optimize the response to be the best it can across the listening area, it cannot correct for spatial variation. For example, if the level of 40 Hz in one seat is 10 dB different to the level in the next seat, the difference between the two seats will always be 10 dB, no matter how much EQ is applied.

In this app note we will use Multi-Sub Optimizer (MSO) to minimize the spatial variation across the listening area. The overall shape of the frequency response will then be taken care by the user-defined target curve in your subsequent Dirac Live calibration.

What you will need [Top]

• A miniDSP DDRC-88A with DDRC-88BM plugin (add it as an option when purchasing the DDRC-88A).

• Room EQ Wizard (REW). This freeware acoustic measurement program is used to take the measurements for use in multi-sub optimization. You will use your existing UMIK-1 (supplied by default with the DDRC-88A) microphone with it. REW runs on Windows, Mac and Linux.

• Multi-Sub Optimizer (MSO). This freeware program is used to optimize the measurements taken with REW. MSO runs on Windows only.

Please note that while the method described here delivers excellent results, it is quite involved. You may prefer to start with a simpler multi-sub method, as described in the related app note Using multiple subwoofers with the miniDSP DDRC-88A.

The example system [Top]

The diagram below illustrates a multi-sub system for a "5.1" configuration. Since two output channels of the DDRC-88A are normally unused in a 5.1 system, they can be used for subwoofers. The subwoofers can be either self-powered or driven from external amplifiers (not shown on the diagram).

A 5.1 system with three subwoofers

In other systems, more (or less) output channels may be available. For example, if using two DDRC-88A units for an Atmos system, four or six channels may be available.

If there are not enough output channels to control each sub, you can add an external miniDSP 2x4 (HD/Balanced) unit. See Appendix A. Notes on using an external 2x4 for the subs for full details.

1. Choose subwoofer locations [Top]

There are no hard and fast rules about where to place subs when doing a multi-sub system. In the app note Using multiple subwoofers with the miniDSP DDRC-88A, two distinct ways of choosing subwoofer locations were outlined. Here are two examples:

2. Set up signal routing [Top]

Set up the Mixer tab of the DDRC-88BM plugin to route the signal produced by Dirac Live to output channels 3, 7 and 8. The channels can be renamed as well – SUB 1, SUB 2, SUB 3:

Note that a single Dirac Live channel is used for all subwoofers. This is so that Dirac Live optimizes for the combined response of all subwoofer channels. (If separate Dirac Live channels were used, each one alone would be optimized but the combined response of all three would not.)

3. Take your measurements [Top]

On the Analysis tab of REW Preferences, set it to use an acoustic timing reference:

When performing these measurements, it is simplest to connect the audio from the computer as follows:

• Left audio channel to input channel 1 of the DDRC-88A. This is used to drive the left front speaker, which is used as the acoustic timing reference.
• Right audio channel to input channel 3 of the DDRC-88A. This is used to drive the subwoofer/s.

Place the UMIK-1 in the middle of the listening area at ear height. The orientation (vertical or horizontal) doesn't really matter for these measurements. You can use either of the downloaded calibration files (0 or 90 degrees). In the DDRC-88BM plugin, check that channels 3, 7, and 8 are not muted.

1. Click on the Measure button. Select Right as the measurement output and Left as the timing reference:

   

2. Click Check Levels and then Start Measuring. When the measurement completes, rename it "ALL Pos 1".

3. In the DDRC-88BM plugin, go to the Outputs tab and mute SUB 2 and SUB 3. (You can also turn off the cross-points in the Routing or Mixer matrix.) Run another measurement and rename it "Sub 1 Pos 1."

4. Mute SUB 1 and unmute SUB 2. Run another measurement and rename it "Sub 2 Pos 1."

5. Mute SUB 2 and unmute SUB 3. Run another measurement and rename it "Sub 3 Pos 1."

Move the microphone to another position. You should also change its height from the floor. Repeat steps 1 to 5 above but this time name the measurements "ALL Pos 2", "Sub 1 Pos 2," "Sub 2 Pos 2," and "Sub 3 Pos 2." Repeat for the remaining positions.

Generally, at least as many measurement positions as subs are required. (3 positions for 3 subs, 4 positions for 4 subs, etc.) However, you can do more. The goal is to "sample" the response of the subs around the listening area. Too many measurements will however make for a lot of work and slow down the optimization. In our example run, we are using five measurement positions.

Click "Save All" and save your measurement as a project.

4. Export measurements [Top]

In REW, drop down the File menu and select Export and then "All measurements as text.".

Choose a new/empty folder and click Open.

Create a graph of your baseline measurement to compare with your later results. On the Overlays window, turn off all measurements except for the "ALL" measurements, then click the camera icon to save an image file. Here is ours:

We seem to have been fairly lucky with our initial placement, as the response over the listening area is very similar up to 55 Hz. Above that, however, there is a quite a large difference in the plots. All positions also tend to have a null just above 100 Hz.

5. Configure MSO [Top]

Open Multi-Sub Optimizer (MSO). Drop down the Tools menu and select Application Options, then click on Hardware. For the DDRC-88BM, set the options as shown here:

6. Load measurements and create a configuration [Top]

You will now need to load your measurements into MSO and set up the configuration. This is essentially the same procedure as the MSO tutorial. In brief, you will need to take the following steps:

1. On the Data View tab, right-click on the Subs node and select "Import Sub Measurements...". Change the file type selector to '*.txt' and select the measurements of the individual subs (that is, "Sub 1 Pos 1" and so on). You should not import the "ALL" measurements. Here, for example, are the SUB 1 measurements imported (you will need to import all three subs.)

   

2. Switch to the Config View tab. If there is a config there already, right-click on it and select "Delete Configuration." Drop down the Config menu and select "Add New Sub-Only Configuration".

   

3. Right-click on Subwoofer Channels and select "Add Filter Channel". Do this once for each sub (i.e. for three subwoofers, create three filter channels).

4. Under Sub Channel 1, right-click on "Filters" and select "Add Gain Block". Do the same for "Add Delay Block," then three times for "Add Parametric EQ".

5. Right-click on "Measurement Associations" and select "Associate Measurements...". In the dialog that pops up, select all measurements for SUB 1 and click OK. Here is the result of steps 3–5:

   

6. Repeat steps 4 and 5 for Sub Channel 2 and Sub Channel 3. (You can make this faster by copying and pasting the filters with Ctrl-C and Ctrl-V.) On the last channel, you won't be able to add a delay – this is fine.

7. Under Optimization Parameters, right-click on "Measurement Groups" and select "Add Measurement Group." In the dialog that pops up, select all measurements for Pos 1. Here's how it looks:

   

   Note the difference to step 5: in step 5, you chose all positions for a single sub, whereas in this step, you choose all subs for a single position.

8. Repeat step 7 for Pos 2, Pos 3, and so on.

9. Switch back to the Data View. Right-click on Graphs and select "New Graph..." Under Data at the left, click on Measurement Groups and select all 5 groups:

   

   Click OK. The summed graph of all five measurement positions should appear:

   

   If this graph is very similar to your baseline, you have imported and set up your measurements correctly! (You may need to adjust the graph axes.)

10. Create another graph, but this time select "Filter Channels" and add all three channels to the graph. This graph will show the EQ that is applied to each sub.

7. Set optimization parameters and run [Top]

Drop down the Tools menu and select "Optimization Parameters."

1. Under Method, select the options as shown here. The Reference Level should be set to the "floor" of the measurements. (The Parametric EQs are by default limited to not boost, only cut. You can experiment with adding boost later on.)

   

2. Under Criteria, uncheck Auto and set the frequency range to a range like 20 to 120 Hz. Set the optimization to run for a minute or two (later on, you should increase this).

   

3. Under Group Weights, leave all groups set to 1.0. This is the best option when selecting for minimum variation.

   

Click on the Start Optimization button. Let the optimization run until it completes.

8. Refine optimization [Top]

You can now refine the results obtained so far. For example:

• Add a Polarity Inversion block to one or more subwoofers (but not all).
• Add a LF Shelf filter.
• Add an all pass filter.
• Add a HPF to subwoofers with limited low-frequency output.

You can also change the properties of individual filter blocks. For example, you could change the maximum gain of some parametric EQ filters. Eventually, you should run the optimizer for a much longer period of time – for example, 30 minutes.

Here is our final result for all 5 measurement positions:

Here are the filter responses of the three sub channels:

9. Export filters [Top]

Once you are happy with your results, right-click on the Subwoofer Channels and select "Normalize Gains." Then select "Normalize Delays." (This ensure that all gains are zero or less, and that all delays are positive.)

Drop down the Config menu and select "Save Biquad text file...". Click on the line "Sub Channel 1" and then the Save button, and select a folder to store the file in.

Repeat the above for Sub Channels 2 and 3.

10. Import filters into the plugin [Top]

In the DDRC-88BM plugin, go to the Outputs tab and click on PEQ for the SUB 1 channel. Select Advanced mode, then IMPORT. Select the file you saved for SUB 1/Sub Channel 1, and the plugin will update to show the filter plot:

Check that the plot matches the expected response from your MSO filter response graph.

Repeat the above for the SUB 2 and SUB 3 channels.

11. Set gains and delays [Top]

1. In MSO, if Sub Channel 1 has a Gain block:

   1. Read its value from the properties window.
   2. In the DDRC-88BM plugin, write that value into the gain entry field of the SUB 1 output channel.

2. In MSO, if Sub Channel 1 has a Delay block:

   1. Read its value from the properties window.
   2. In the DDRC-88BM plugin, write that value into the delay entry field of the SUB 1 output channel.

3. In MSO, if Sub Channel 1 has a Polarity Inversion block:

   1. In the DDRC-88BM plugin, click on the Invert button of the SUB 1 output channel (so that it says "Inverted").

Repeat steps 1–3 above for Subwoofer Channel 2 (SUB 2/output channel 7 in the DDRC-88BM plugin) and Sub Channel 3 (SUB 3/output channel 8 in the DDRC-88BM plugin).

Here is the result for our example:

12. Confirm your results [Top]

It is a good idea to confirm that your measurements reflect the results predicted by MSO. Re-run the measurements for all subs, with the microphone in each of the measurement positions. (There is no need to re-run the measurements of the individual subs.) If your results are not similar to the predicted results shown in MSO, you may have made a mistake in transferring the data from MSO to the DDRC-88BM plugin.

Here is the result from our example system. The variation above 55 Hz is much lower than before. The "hole" around the 100 Hz area is mostly gone. While there are some deep notches, these are very narrow and each appears in only one of the measurement positions.

Wrapping up [Top]

Once you are satisfied with your multi-sub optimization, save your DDRC-88BM configuration to a file. Then proceed as follows:

1. Run your Dirac Live calibration as described in the User Manual (pdf).

2. Implement bass management (if desired). See the application note Bass management with the DDRC-88BM.

Then play music and listen to the results! You deserve it! Have fun, and let us know about your experiences in our forum.

As a closing statement, we'd like to thank Andy C (Andyc56) for his hard work on this amazing freeware. Multi-Sub optimizer is nothing short of providing amazing results. We hope that this app note will help you navigate his very powerful software and make sure to thank him online!

Appendix A. Notes on using an external 2x4 for the subs [Top]

If you do not have enough channels on the DDRC-88A to control each sub, you can add another miniDSP unit such as a:

• 2x4 Balanced (for subwoofers with balanced inputs) with the 2x4 Advanced plugin.
• 2x4 HD (2.0V maximum output) with the 2x4 HD1 plugin.
• 2x4 (0.9V maximum output) 2x4 Advanced plugin.

A typical connection is shown in the diagram below:

A 7.1 system using a miniDSP 2x4 (HD/Balanced) to drive up to four subwoofers

In this case, the Mixer tab of the DDRC-88BM plugin can just route Dirac channel 3 to output channel 3. In addition, set the external 2x4 (HD/Balanced) to route from input channel 1 through to all four outputs. In the 2x4 Advanced plugin:

In the 2x4 HD1 plugin:

When setting the Application Options screen, use these values:

If using the 2x4 Advanced plugin, you will need to limit the value of the Delay blocks. Initially, set "Maximum value" to 7.2 and "Minimum value" to -7.2. If, after using "Normalize Delays" (Step 9. Export filters), any Delay block has a delay greater than 7.2, you will need to:

1. Change the "Maximum value" to 3.6 and "Minimum value" to -3.6.

2. Re-run the optimizer.

In this app note we will show you how to perform bass management with the miniDSP DDRC-88BM plugin. Before starting on this app note, you will need to have performed a Dirac Live calibration. If you are implementing an active speaker or a multi-sub system with the DDRC-88BM, these will also need to be done before setting up bass management – see the separate app notes Implementing active speakers with the DDRC-88BM and Optimizing multiple subwoofers with the DDRC-88BM and Multi-Sub Optimizer.

• The need for bass management
• 1. Establish your baseline
• 2. Configure routing
• 3. Configure one crossover
• 4. Configure remaining channels
• 5. Wrapping up
• Appendix A: using analog inputs for measurement
• Appendix B: the 5.1 and 7.1 system types

The need for bass management [Top]

When movies are mixed for the cinema, each speaker channel is specified as full bandwidth i.e. 20 Hz to 20 kHz. The Low Frequency Effects (LFE) channel is used for additional low-frequency content – rumbles, booms, and the like – and is fed to dedicated subwoofers in order to avoid overloading the speakers with high-energy low-frequency content.

In a typical home theater, the speakers are typically not capable of reproducing the full frequency range, down to 20 Hz (or thereabouts). The solution is bass management, where low frequencies are filtered out from the speaker channels and instead are sent to the subwoofer.

While bass management can be performed by the A/V receiver or processor of the home theater system, the DDRC-88BM provides very flexible control over crossover slopes and frequencies, more so than most A/V processors.

Figure 1 illustrates. The seven speaker channels are high pass filtered to remove low frequencies. The speaker channels are also low pass filtered and these bass frequencies summed with the LFE signal to be sent to the subwoofer. All of these signals then pass through the Dirac Live room correction and output channel processing.

Figure 1. Bass management with the DDRC-88BM ("Custom" system type)

Figure 1 includes the note "with 10 dB gain" on the LFE channel input. This is because the LFE channel on the disc is recorded 10 dB lower than the other channels to avoid overloading the recording medium. This 10 dB must be regained somewhere in the playback chain, and typical A/V receivers and processors apply this gain internally. The Dirac Live calibration will, in this scenario, have used the "Custom" system type. The main part of this app note assumes this is the case.

(If your source equipment does not provide 10 dB gain internally – as may be the case if using a Blu-ray player with multichannel analog outputs or a home theater PC with USB DAC – then your approach will be slightly different. In that case, see Appendix B.)

1. Establish your baseline [Top]

The starting point for setting up bass management is a set of output channels that have been equalized and time-aligned with Dirac Live Calibration Tool (DLCT), as described in the User Manual (pdf).

Bass management is most accurately set up by using acoustic measurement. You can use your UMIK-1 together with any suitable acoustic measurement program such as Room EQ Wizard (REW). This diagram shows the recommended measurement setup:

In this setup, the computer is connected to the A/V receiver or processor via HDMI. This enables the measurement program to send audio to each HDMI channel and thence to the DDRC-88A. See the app note "Using the UMIK-1 and REW with HDMI output" for Windows or Mac. For this to work properly, you must disable all processing in the AV receiver/processor, including bass management, up/down-mixing and effects of any kind.

If you prefer (or if you cannot get HDMI output working), you can instead connect an analog line out from the computer directly to each input of the DDRC-88A in turn. However, in that case, see Appendix A.

Before proceeding, check that the Routing matrix is set to the default i.e. that the "LFE Mgt" signal is not used. See the User Manual.</>

Position the UMIK-1 at the center of the listening area, pointed at the ceiling, and measure the response of each channel. Make sure that you are using the correct calibration file (the "90 degree" version). You should end up with a result similar to the following graph (LFE channel in red, left front speaker in blue and left surround speaker in green, all with 1/6th octave smoothing):

Note how the LFE channel measures 10 dB higher than the speaker channel. (It's actually a bit more than 10 dB because of the effect of the target curve that was set in DLCT.) Also, you will probably not get perfectly smooth and equal measured responses. That is because you are measuring in one location only whereas DLCT optimizes over the whole listening area (nine measurement locations).

2. Configure routing [Top]

On the Routing tab, route the LFE Mgt signal to Dirac channel 3. Make sure to also turn off the routing of the LFE In 1 signal:

3. Configure one crossover [Top]

On the LFE Mgt tab, click on the HPF (high pass filter) and LFP (low pass filter) blocks to configure the crossover between the left front speaker and the subwoofer. The composite screenshot below illustrates with a typical 80 Hz typical crossover frequency.

Check also that the mix levels are all set to 0 dB.

Re-measure the front left channel. You should see that the measured response now includes the subwoofer (blue plot in the graph below). If you do not have a smooth transition between the subwoofer and the speaker, change the high pass or low pass frequency or filter type and re-measure. (You do not need to have the same frequency and type for both high pass and low pass filters.) Getting the best response through the crossover region may require several attempts.

Note: it may be tempting to change the delay on the output channels to get a smooth response through the crossover. This should be avoided: remember that Dirac Live has already set delays on all channels, so subsequently changing the delay on the output channels will corrupt its calibration.

4. Configure remaining channels [Top]

Copy the high pass and low pass filter settings to the other channels, and measure them. You may need to change the settings for different speakers. Surround speakers will often have different settings, but where possible, try and use the same settings for the corresponding left and right speakers.

You should also apply a low pass filter to the LFE channel (e.g. Butterworth 24 dB/octave at 120 Hz), in case the disc contains high frequency content on that channel.

This graph shows the result after applying bass management in our test system (LFE channel in red, left front speaker in blue, and left surround speaker in green, all with 1/6th octave smoothing):

In our example, we have used the default Dirac Live target curve. If you have used a custom target curve, you will of course expect to see this reflected in your measurements.

5. Wrapping up [Top]

Now sit back and listen to the results! You may find that you wish to "tweak" the settings further based on listening – either by generating Dirac Live filters with different target curves or by fine-tuning the crossover settings in the plugin. In any case, have fun, and let us know about your experiences in our forum.

Appendix A: using analog inputs for measurement [Top]

If you perform your measurements by connecting directly to the analog inputs of the DDRC-88A, the 10 dB alignment gain will not be applied to the LFE channel. The effect that this will have on your measurements is that the measured LFE channel response will be at the same level as the speaker channels. Your settings in the DDRC-88BM plugin will however be the same.

Appendix B: the 5.1 and 7.1 system types [Top]

If your source equipment does not apply the 10 dB alignment gain, you will perform your Dirac Live calibration using the "5.1" or "7.1" system type. In that case, Dirac Live calibrates the subwoofer channel for a 10 dB higher acoustic output level than the speaker channels. (To optimize the gain structure of this type of system, set the subwoofer level several dB higher than the speakers in the Output&Levels tab of DLCT.)

Because the LFE gain has not been applied when the signals reach the LFE Mgt tab, the speaker channels must be reduced in level by 10 dB before mixing with the LFE channel, as illustrated in this diagram:

Accomplishing this is simple. On the LFE Mgt tab, set the mix level of the speaker channels to −10 dB, but leave the LFE channel at 0 dB:

Everything else proceeds as described in the body of this app note for the "Custom" system type. Even if using the analog inputs for measurement, the plots will show the LFE level 10 dB higher than the speakers.

The DDRC-88BM plugin includes miniDSP's flexible and powerful signal processing functionality on each output channel, along with full signal routing/mixing. This can be used to implement active speakers.

Note: configuring the DDRC-88BM plugin for active speakers should be done before performing your Dirac Live calibration.

The example project [Top]

Figure 1 illustrates a multichannel active system in which four channels from an A/V receiver or processor are input into the DDRC-88A: left front, right front, center, and LFE/subwoofer. The front and center speakers are active two-ways. A great many other combinations are also possible – with a single DDRC-88A you can implement four two-way active speakers, two four-ways, or two three-ways and a two-way. You can even add on multiple subwoofers and optimize them as described in Optimizing multiple subwoofers with the DDRC-88BM and Multi-Sub Optimizer. For more complex systems, multiple DDRC-88A units can be used.

Figure 1. Connection of an example active multichannel system

Note: when connecting power amplifiers directly to speaker drivers, a protection capacitor in series with each tweeter is strongly recommended.

1. Select the speaker drivers and design the enclosure [Top]

If you are starting from scratch, you will need to select the drivers for your speakers. There are literally hundreds of drivers available for DIY use at all price levels, so it's impossible to give specific recommendations here. Peruse the online forums to see what others are using and to ask for recommendations for your particular project.

If you are building your own box, you will need to design it. The most important factor is the internal volume, and if it's a ported box, the size and length of the port. Fortunately, there are a number of free programs that do the complex math for this based on the Thiele-Small parameters of the woofer. For example, a popular Excel-based program is Unibox.

If you are modifying an existing speaker from passive to active, then you have the enclosure and the drivers already. In this case, you will most likely need to remove the internal crossover and add a second pair of binding posts.

2. Set up the Mixer tab [Top]

The Mixer tab is the key to implementing an active speaker system. This is how to set it up for the example system:

The output from Dirac Live channel 1 is sent to output channels 1 and 2. These will be used for the woofer and tweeter of the left front speaker, as indicated by the labels at the top. (The names are changed on the Outputs tab.) The right and center speakers are set up in the same way. The two remaining output channels are used to drive two subwoofers.

The Routing tab doesn't have to be set up specially for this application. However, since only four input channels are used, this can be indicated like this:

3. Measure and equalize the drivers [Top]

You will now need to measure the drivers one at a time. (You only need to do this for one speaker.) For information on how to measure a loudspeaker driver, please see the app note Loudspeaker measurement with UMIK-1 and REW.

Then use the PEQ blocks on each output channel to shape the response of each driver so that it is flat over its operating range. Use "Peak" type filters to flatten peaks (with negative gain so they create a notch) and "High-Shelf" and "Low-Shelf" type filters to straighten out the overall response.

Below is an example measurement for a woofer. The various features of the measurement and the areas to correct are marked on the graph, along with the corrected response in light blue. If using an indoor measurement (as in the example), be careful not to correct for peaks and notches caused by the room.

Here is a graph of the tweeter, measured before and after correcting its response. When performing a tweeter measurement, start the sweep at a frequency so as not to strain the tweeter (e.g. start at 1 kHz, not 20 Hz).

You will generally want to use the linking feature on the PEQ blocks to make the same settings for corresponding left and right speakers. With the example system, link channels 1 and 3, and 2 and 4.

4. Add the crossover

Open the crossover settings screen by clicking on the Xover button on the Outputs tab. For a two-way speaker, set a low pass filter on the woofer and a high pass filter on the tweeter. For a three-way speaker, the midrange driver will need both high pass and low pass filters. As a starting point, try using Linkwitz-Riley (LR) 24 dB/octave filters. You can then experiment with lower or higher slopes, from 6 dB/octave up to 48 dB/octave.

Now measure the response of both drivers together. Use the output level controls to match the signal levels from the woofer and tweeter. You may need to fine-tune to get the smoothest response around the crossover frequency:

• Time align the drivers
• Move the filter corner frequency of one driver up or down a little
• Use an asymmetrical crossover, for example BW 18 dB/octave lowpass on the woofer and LR 24 dB/octave high pass on the tweeter
• Adjust the equalization of one driver or the other near the crossover point

This REW plot shows the response of the woofer and tweeter with crossover filters in place, and the combined response after crossover fine-tuning:

Wrapping up [Top]

Once that's all done, you can position the speakers in their proper positions, play music and listen to the results! You may wish to continue to experiment – for example, by trying a different crossover frequency or slopes. It is easy to do A-B auditions by setting up different configurations and switching between them with a remote control.

If you have spare output channels, you may also want to consider using them for additional subwoofers and optimizing as described in our app note Optimizing multiple subwoofers with the DDRC-88BM and Multi-Sub Optimizer.

Once that's all done perform perform your Dirac Live calibration as described in the User Manual (pdf). Then set up your bass management as described in the app note Bass management with the DDRC-88BM. Have fun, and let us know about your experiences in our forum.

In this app note, we will show you how to use the DDRC-88A to apply Dirac Live® room correction in multiple rooms. In this app note we will assume that there are four stereo zones to be set up.

Note: please follow the instructions below carefully. We recommend that you work fully through a single room calibration using the instructions in the comprehensive User Manual before attempting a multi-zone setup.

• Zone 1
• Zone 2
• Zones 3 and 4
• Wrapping up

Zone 1[Top]

1. Start Dirac Live Calibration Tool for miniDSP (DLCT).

2. On the Sound system tab, choose a Custom System with 2 channels.

   

3. On the Mic config tab, select the UMIK-1 as usual. Use the 90-degree calibration file if the microphone is to be oriented at 90 degrees to the speakers, and the on-axis (0-degree) calibration file if the microphone is to be pointed in the general direction of the speakers.

   

4. On the Output & levels tab, set the output channels to be channels #1 and #2, and name the two channels appropriately. With the output volume set low to start with, click on the Test button for a channel and adjust output volume and input gain so that the meter reads in the middle of the green area. Repeat for the second channel.

   

5. Proceed to the Measurements tab and perform the first measurement. The screenshot below shows the result. Proceed to perform the remaining eight measurements. It does not matter too much which type of listening area you select, but the Sofa option is probably a good choice for many situations. The important thing is to spread the measurement locations around the full listening area and to also vary the microphone height vertically.

   

6. Proceed to the Filter Design tab, adjust your target curve, and click on the Optimize button to generate the correction filters. This screenshot shows the measurements before optimization, and the predicted response.

   

7. Save the project. Use a meaningful name. In our case, we will use "Zone 1."

8. Proceed to the Export tab and drag the box labeled with the project name onto the first empty slot (slot 1 shown in the example). The slot will show the project name and the names of each channel.

   

Zone 2[Top]

9. Move the microphone to the second zone. If a long USB cable (> 5m) is required, an active USB repeater may be needed.

10. Go back to the Sound system tab. Change the first output channel to Channel #3. You will get a warning like this:

    

    Click on OK. Change the output channel for the second channel to Channel #4.

11. Give the two channels meaningful names:

    

12. Reduce the Output volume, and again use the Test buttons to set suitable measurement levels.

13. Proceed through the Measurement and Filter Design tabs, as before.

14. Save your measurements as a new project. Use a meaningful name. In our case, we will use "Zone 2."

15. Proceed to the Export tab and drag the box labeled with the project name onto the same slot as before (slot 1 in our example). The new filters will load into Channel #3 and Channel #4, as shown here:

    

Zones 3 and 4[Top]

Repeat the steps listed above for zones 3 and 4. Here are the channel assignments:

Don't forget to save each project with a meaningful name! Here are all four zones loaded into the DDRC-88A:

Wrapping up[Top]

If you haven't already, turn on Dirac Live filtering and play audio. Note that the selection/switching of which zones to play through will need to be done in the source equipment—the DDRC-88A is only performing room correction for the zones, not zone switching.

That's it for this app note! If you do set up multiple zones with the DDRC-88A, please let us know your experience in our forum.

This short app note gives a worked example of the gain structure optimization procedure described in the DDRC-88A User Manual.

Preliminaries[Top]

If the Output & Levels tab in Dirac Live Calibration Tool for miniDSP shows a large difference in the measured sound level for different channels, it would be a good idea to run the gain structure optimization procedure described in Section 6.3 of the DDRC-88A User Manual. The purpose of the procedure is to adjust the output gains of the DDRC-88A and following equipment to avoid problems with low output levels or clipping.

To illustrate the procedure, we used a 5.1-channel test system. The DDRC was internally configured on all channels as follows:

• Input sensitivity: 2 V
• Output gain setting: 2 V

This is a summary of the amplification:

Procedure[Top]

1. On the Output & levels tab, run the test signal on the first channel. Set Output volume and Input gain so that the meter reads half-way on the green bar, at −12 dB.

   

   Figure 1. DDRC-88A Output and Levels tab.

2. Run the test signal on the other channels and note the level. Don't touch the Channel volume sliders! (Leave them at maximum.) In Figure 2 below, the meter readings are shown in the "Before" column and the corresponding levels are written down in the "Reading" column.

3. Now decide on the target level for each channel. Ours are shown in the "Target" column in Figure 2. Since the front left and right are at −12 dB, we used that for the speakers. Because we are using the 5.1 system configuration, we set the target for the sub 6 dB higher i.e. at −6 dB. (This is as advised in the User Manual. For the Custom system configuration, use the same target level as the speakers.)

4. Now we can calculate the amount we want to adjust each channel: Change = Target − Reading.

   

   Figure 2. Measured levels and gain adjustments.

5. Here's the tricky bit: adjust the gains on each channel by the amount noted for Change. There are two tables in the User Manual that show all the options for making gain changes, and we need to just pick the closest. In this case, we were able to get the system pretty much spot-on. (Note: turn off power to the DDRC-88A and amplifiers while making these changes.) In this system, we:

   1. Changed the connection to the subwoofer from unbalanced to balanced. This gave us an extra 12 dB of output gain, so we then adjusted the gain control on the amplifier down by 8 dB.
   2. Dropped the volume control on the center channel monoblock just a tad. (Totally unnecessary, but since we could...)
   3. Flipped the output DIP switches for the surround channels from the "UP" position to the "DOWN" position (changing output gain setting from 2V to 0.9V).

6. With that all done, going back to the Output & Levels tab gave the meter readings shown in the "After" column of Figure 2.

Wrapping up[Top]

It's not necessary to get the meter readings as close as we were able to in this example. The aim of the procedure is to get the gain structure more consistent across channels and avoid major issues — it's not necessary or useful to get the readings down to the last dB.

Enjoy your DDRC-88A! If you run into problems with this procedure, feel free to ask on our forum.

In this app note, we will show you how to use multiple subwoofers with the DDRC-88A.

• Why use multiple subwoofers?
• Start with one subwoofer
• Connecting additional subwoofers
• Method A
• Method B
• Wrapping up

Why use multiple subwoofers? [Top]

The primary reason for using multiple subwoofers is to improve evenness of bass response across the whole listening area. While Dirac Live will optimize the response to be the best it can across the listening area, it cannot correct for spatial variation. For example, if the level of 40 Hz in one seat is 10 dB different to the level in the next seat, the difference between the two seats will always be 10 dB, no matter how much EQ is applied. The only solutions are to find a better position for the (single) subwoofer, or to use multiple subwoofers.

Multiple subwoofers can also help with other problems. For example, a persistent null in the response can be filled in by using multiple subwoofers. (A null is not just an area of low output, but an area of zero output. No amount of EQ can properly correct for a null.) Because real listening rooms are limited to where subwoofers can be placed (due to furniture, doors, and aesthetics), it may not be possible to locate a single subwoofer in the optimum position. Use of more than one subwoofer provides a greater degree of freedom with subwoofer location.

Start with one subwoofer[Top]

The best place to start is with a single subwoofer. This will ensure that you have your DDRC-88A operating correctly and you are able to perform a calibration. It will also provide a "baseline" against which you can measure improvements from adding additional subwoofers.

For reference, this diagram shows the setup with a single subwoofer:

Figure 1. Single subwoofer connection to the DDRC-88A.

We want to focus on the response of just the subwoofer(s), so let's configure DLCT to run Dirac Live on only the subwoofer. On the Sound System tab, create a Custom System configuration with just one channel:

On the Output & Levels tab, assign the output to channel 3:

Now we can run some measurements. We want to look at variation in bass response across the listening area, so more than one measurement location is needed. Three is probably the minimum, although the full nine would be excessively time-consuming. For this app note, we will use three locations: one in the middle of the listening area; one to the right, up and back; and one to the left, down and forward.

After taking three measurements, go to the Filter Design tab and:

1. Set a flat target curve, like this:

   

2. Click on the Optimize button.

3. Select the All (after) checkbox (and turn off all the others).

4. Zoom in on the graph to view a frequency range from 20 to 200 Hz and a magnitude range from -20 to +10 dB.

Here is the result we obtained for a subwoofer located in the center of the front wall (the best location we have found for a single subwoofer in this room):

Graph 1. Illustrating spatial variation in equalized subwoofer response.

This set of graphs gives us an indication of spatial variation — that is, how much variation there is across the listening area. In this example, it's very good up to 55 Hz, but from 60 to 95 Hz one of the measurement locations is about 6–8 dB lower. (We don't care much about what happens above 100 Hz.)

So that's our baseline! Let's see what improvements we can get with more subs.

Connecting multiple subwoofers[Top]

To connect a second subwoofer to the DDRC-88A, all we need is a Y-connector or adapter cable, like this:

Figure 2. Connecting two subwoofers to the DDRC-88A (powered).

Or if using passive subwoofers and external amplifiers, your setup may look something like this:

Figure 3. Connecting multiple subwoofers to the DDRC-88A (passive).

Note that all subs are connected to a single channel of the DDRC-88A. The behavior of low-frequency sound in a listening room is very complex, but this is a relatively straightforward way of obtaining a good result together with Dirac Live, as we will see in the following sections.

Method A [Top]

We will demonstrate two different methods of setting up multiple subs. Method A uses two or four identical subwoofers positioned symmetrically in the room, and is based on research at Harman International (PDF download). The goal of the research was to obtain the most consistent bass response across the listening area. The best subwoofer locations were:

Figure 4. Symmetrical layouts for multiple subwoofers.

To test Method A, we placed two identical sealed subs in our test room as per Layout 1 above. The result (with the microphone in the same three locations as before) is shown in Graph 2 below.

Graph 2. Equalized response of two subwoofers in the center of the front and rear walls.

Below 60 Hz and from 85–100 Hz, the amount of spatial variation is low. Between 60 and 85 Hz, there are deep notches in some locations, but there isn't a deep wide "hole" anywhere as there is in Graph 1. So that's an improvement! Note also that our test room is not actually rectangular, so in a rectangular room the result may be better.

Method B [Top]

Method A is not always feasible: real rooms are often not rectangular and the needed locations for Method A may not be available for practical reasons (furniture, equipment, doors/walkways, etc). Method B is based heavily on (but is not the same as) the technique described by Dr. Earl Geddes in two short papers here. It uses three subwoofers with asymmetrical or "random" placement. Here are a couple of example layouts (Geddes also recommends lifting one sub closer to the ceiling if possible):

Figure 4. Asymmetrical layouts for multiple subwoofers.

Method B is as follows:

1. With only S1 turned on, run your baseline measurement. It's best to record the result, so save the project and/or take screenshots. Our result is in Graph 3 below.
2. Turn on S2 (leave S1 turned on) and run the measurements. If the result is better than step 1, proceed to the next step. Otherwise, change the parameters (*) of S2 and repeat. Our result is in Graph 4 below.
3. Turn on S3 (leave S1 and S2 turned on) and run the measurements. If the result is better than step 2, you're done! Otherwise change the parameters of S3 and repeat. Our result is in Graph 5 below.

To test Method B, we used Layout 4 in our test room with three sealed subwoofers (similar but not all identical). Here are the graphs we obtained:

Graph 3. Equalized response of a single subwoofer in front left corner.

Graph 4. Equalized response after adding a second subwoofer at the center of the rear wall.

Graph 5. Equalized response after adding a third subwoofer on the side wall.

Each additional sub reduced the variation between the three measurement locations. We re-ran the measurements for each additional subwoofer only a couple of times, so it's possible that more time spent on varying sub parameters and measuring would yield even better results.

Wrapping up[Top]

After completing your subwoofer measurements, you will need to redo your Dirac Live calibration with your regular (5.1, 7.1, or Custom) system configuration. If you are using bass management, you will also need to completely redo/reconfigure your bass management so that the new response with multiple subs is taken into account. (Follow the instructions of your equipment manufacturer.)

Use of multiple subwoofers is not an excuse to use low-quality subwoofers! Low distortion, sufficient frequency extension and output, minimal cabinet vibration and driver self-noise, high quality electronics (if powered), and low port noise (if ported) are necessary attributes for good results, regardless of the number of subwoofers used.

That's it for this app note! Have fun, and if you try this experiment with multiple subs and the DDRC-88A, please let us know your results in our forum.

In this app note we will show you how to equalize a surround sound or home theater system using the miniDSP 10x10 Hd. This unit has eight channels of analog input, eight channels of analog output, one stereo digital input, and one stereo digital output. Whether you have a traditional 4.0 quadraphonic surround sound setup, a 5.1 setup, or a 7.1 setup, the 10x10 Hd will fit into your system.

• What you will need
• 1. Getting set up for measurement
• 2. Set up your routing
• 3. Establish your baseline
• 4. Equalize the subwoofer and speakers
• 5. Apply bass management
• Advanced routing

What you will need [Top]

• A miniDSP 10x10 Hd. Alternatively, if you are into DIY, you can use the 8x8 kit.

• The miniDSP 10x10 plugin software.

• Room EQ Wizard (REW). Be sure to download the latest version from the Downloads Area for UMIK-1 support.

• A miniDSP UMIK-1. After receiving the UMIK-1, go to the UMIK-1 page to get the calibration file for your unique serial number, and save it as a .TXT file.

• A microphone stand with boom arm, available in music supply stores. (While the UMIK-1 is supplied with a small "table top" stand, best results will be obtained if the microphone is placed in free space, away from sofas or tables.)

Note: if you already have a measurement microphone and/or measurement software, those can be used instead of the UMIK-1 and REW. The auto-EQ function will only work on REW though.

1. Getting set up for measurement [Top]

The diagram below illustrates the setup used for acoustic measurement. Typically, running the computer measurement signal through the 10x10 Hd will require disconnecting the analog connections from the source and connecting the computer output to the appropriate 10x10 Hd input.

Setting up the computer to use REW and the UMIK-1 is covered in the following app note:

• UMIK-1 setup with REW

2. Set up your routing [Top]

The miniDSP 10x10 HD has three screens that are used to configure all of its settings, called Inputs, Routing, and Outputs. To begin with, let's just give all of our channels useful names. On the Inputs screen, click on the name of each input channel and type in a new name. Here is a typical setup for a 7.1 system:

Do the same for the Outputs screen:

On the Routing screen, use the On/Off buttons to set up which input channels are routed through to which output channels. To start with, pass each input straight through to the corresponding output, like this:

In the above example, we also set the digital inputs to pass through to the left and right output channels, which is appropriate if you wish to play stereo content through the digital input.

3. Establish your baseline [Top]

Use REW to measure the in-room frequency response of each of your speaker and subwoofer channels, one at a time. Position the microphone at the center of the listening area, and for best results and consistency, point the UMIK-1 towards the speaker being measured.

You will end up with a set of frequency response measurements. Use the REW Overlays window and 1/3rd, 1/6th, or 1/12th octave smoothing to view them. Here is an example showing the left channel in red, the center channel in blue, and the subwoofer in green. (The other speakers are omitted from the example to avoid cluttering the graph.)

Performing this baseline measurement ensures that your setup is ready to proceed with equalization. In some cases, the baseline measurements may reveal issues that will need to be corrected before proceeding. Remember also that the subwoofer response will vary a lot with the location of the subwoofer. Experiment with different subwoofer locations in order to get the best response prior to equalization.

You may also wish to try making additional measurements with the microphone in different locations around your seating area. This is a "sanity check" so that you can understand how the frequency response varies around the lsitening area. It can help you to ensure that you are not equalizing for issues that vary a lot over the seating area and are not "over-equalizing."

4. Equalize the subwoofer and speakers[Top]

Start by equalizing the subwoofer. This is most easily and quickly done using the REW auto-EQ function. The method for doing this is described in the app note:

• Auto-EQ tuning with REW

For the speaker channels, you may be interested in trying manual equalization, where you set the EQ settings directly in the user interface. It is more time-consuming but gives you more control over the final result.

The sequence when doing manual equalization is 1. measure, 2. perform an EQ adjustment, 3. measure again. Of the various types of EQ available in the 10x10 plugin, there are two that we suggest as the ones to start with. The first is the PEAK type filter with a negative gain, used to "flatten" peaks in the response. Here is an example setting:

The second is shelving filters. These are very useful for "shaping" the response of your speakers, as they elevate or lower the whole spectrum above or below the specified frequency. For example, you can use a HIGH_SHELF filter to create a gentle high-frequency slope that adjusts the "tilt" of the in-room frequency response curve:

You may find that the left and right speakers measure a bit differently. In that case, you will need to pick a compromise. Overall, you are looking to get to an in-room curve that is reasonably smooth, and is consistent between all speakers. The in-room curve should typically have a gentle slope down in the treble (sometimes called a "house curve").

At this time, you will also want to roughly match the levels of the subwoofer and the other speakers. This is easily done using the output gain controls (it's usually better to use negative gain than positive). Here are the left, center, and subwoofer channels of our example system after completing this process:

Here are some additional guidelines when equalizing the speaker channels:

• Avoid using boost filters (filters with positive gain). As a general rule, negative gain is preferred as it ensures that clipping will not be introduced.

• If boost filters are used, they should have a fairly low Q (1 or less) and a moderate amount of gain (4-5 dB max). Don't try and use a narrow boost to fill in a deep notch in the response.

• View measurements at 1/3rd or 1/6th octave smoothing as a general rule. Switch to 1/12th octave smoothing to check that you are not attempting to correct a narrow peak with a broad filter.

• Even with negative gain, use high-Q filters (Q>4) carefully if at all.

• Link the PEQ filters for the various left and right channels (generally, it's best to avoid using different EQ settings for the left and right speakers):

5. Apply bass management [Top]

Bass management is used to direct low-frequency content to the subwoofer or subwoofers. For example, if the center and surround speakers are not able to produce low frequency content, they can be routed to the subwoofer with the Routing matrix, like this:

The subwoofer output channel must have a steep low pass filter applied (typically at 80 Hz) to remove all high-frequency content. Usually, high pass filters will also be used on the bass-managed channels to remove low-frequency content from them. The signal flow path being implemented with this routing and is illustrated in this diagram (EQ is left off the diagram to simplify it):

Here are the left and center channels of our example system after applying bass management, annotated to show our "house curve":

For more detailed information on bass management and how to apply it, see the related app note Bass management with the nanoAVR. Be aware that summing multiple channels can potentially cause clipping of the signal. If you detect clipping, reduce the input gains and if necessary, increase output gains on selected channels.

Advanced routing [Top]

More advanced routing configurations than those described so far can be set up in the Routing matrix. There are dozens if not hundreds of possible combinations, so we will illustrate just a few here.

Re-routing LFE content

This is like bass management "in reverse". In a system in which there is no subwoofer, but the left and right front speakers are able to produce low frequencies, the LFE channel input signal can be routed to those speakers like this:

(Adjust the level of the LFE content using the input gain control on that channel. Note that it is not possible to use this method to send low-frequency content from the center and surrounds to the front left and right speakers.)

Multiple subwoofers

If you have unused output channels (e.g. you are running 4.x or 5.x), then you can use those spare channels to independently equalize and delay multiple subwoofers. Here is an example routing matrix with three independent subwoofer output channels:

To send low-frequency signals from other channels to these subwoofers (for bass management), turn on the routing to those subs as described above and add high pass filtering as appropriate.

Active crossovers

Above we assumed that each output channel was driving a full-range speaker. The 10x10 Hd can also be used to implement active crossovers, in any combination within the limits of the number of output channels (eight analog + two digital). For example, here is a routing matrix that implements two-way active crossovers on the front left and right speakers of a 5.1 system:

You can of course combine this with bass management by sending some of the input channels to the subwoofer output channel. Note also that in this example, the left and right rear inputs (if used) are routed to the left and right surround outputs.

With use of the digital outputs, a total of ten output channels are available, leading to even more advanced configurations. Here is a routing matrix for two-way active crossovers on the front left, front right, and center, with two separate subwoofer outputs on the digital outputs (just add an inexpensive DAC):

Here is a routing matrix for a 4.0 system with two-way active speakers on each of the four main channels, as well as two derived subwoofer output channels:

If you need more output channels, you can use two (or more!) 10x10 Hd units. For example, a pair of 10x10 Hd units would allow for a three-way active crossover on the front left and right, two-way crossovers on the center and four surround channels, and a number of LFE/sub channels. (Note: only one miniDSP can be connected to the computer for programming/configuration at a time.)

For more information on setting up active crossovers, see these App notes:

• Building a 2way crossover
• Stereo 3/4 way Crossover

In this app note we will show you how to perform bass management with the nanoAVR and the nanoAVR_BM bass management plugin. Before starting on this app note, you will need to have connected the nanoAVR into your system and equalized each output channel following the app note Equalizing your home theater system with the nanoAVR and UMIK-1.

• The need for bass management
• 1. Check your baseline
• 2. Configure mixing/routing
• 3. Configure crossovers
• 4. Measure and tune
• 5. Additional notes

The need for bass management [Top]

When movies are mixed for the cinema, each speaker is specified as a full bandwidth channel—that is, 20 Hz to 20 kHz. In practice, a bandwidth of 40 Hz to 18 kHz for the speakers is considered acceptable in cinema and the mixing studio (see the paper "Recommendations For Surround Sound Production"). The Low Frequency Effects (LFE) channel is used for additional low-frequency content. It contains rumbles, booms, and the like, and is fed to dedicated subwoofers in order to avoid overloading the speakers with high-energy low-frequency content.

In a typical home theater, some or all of the speakers are not capable of reproducing frequencies down to 40 Hz, let alone 20 Hz. Even if using hifi speakers with a specified low frequency roll off of 40 Hz, they may not be capable of producing this frequency at the volumes required for home theater. The solution is bass management, where low frequencies are filtered out from the speaker channels and instead are sent to the subwoofer.

The figure below provides an overview of how bass management is implemented with the nanoAVR and the nanoAVR_BM plugin. At the top in the blue highlight, the seven speaker channels are high pass filtered to remove low frequencies from the signal sent to the speakers. In the green highlight, the signal from the speaker channels is low pass filtered and then summed with the LFE signal. The sum of the LFE channel and the low-frequency content from the speaker channels is sent on to the subwoofer.

There is a twist: the LFE channel on the disc is recorded 10 dB lower than the other channels to avoid overloading the recording medium. When we mix the LFE channel with content from the speaker channels, we therefore have to adjust either one or the other by 10 dB to ensure that the mix is correct. In the recommended approach given in this app note, the level of all speaker channels into the subwoofer mix is reduced by 10 dB.

(Note: it may appear from the diagram as though the LFE and/or bass levels will be too low. However, as per the app note Equalizing your home theater system with the nanoAVR and UMIK-1, the gain on the subwoofer amp should already be set so that the LFE channel measures 10 dB higher than the speaker channels. The net result is that the LFE level and the level of bass content from the speaker channels is correct.)

Compared to using the bass management built into your A/V receiver or processor, the advantages of the miniDSP nanoAVR for bass management are:

• Precise control over the slopes and frequencies of the low pass and high pass filters
• Steeper filter slopes (up to 48 dB/octave)
• Ability to fine-tune every single channel independently
• More accurate subwoofer-speaker integration by using nanoAVR features like time delay
• Combined with measurement and equalization, a much more accurate system!

1. Check your baseline [Top]

The starting point for setting up bass management is a set of equalized output channels (with down-mixing if necessary), as described in Equalizing your home theater system with the nanoAVR and UMIK-1. If it's been a while since you did that, you may like to run those measurements again. (In particular, check that you have your subwoofer gain set correctly, so that the LFE channel measures 10 dB higher than the speaker channels.) Otherwise, you're all set to go!

2. Configure mixing/routing [Top]

The nanoAVR_BM plugin has a special screen for performing the mixing and filtering shown in the diagram above. But first, change the Routing screen so that the LFE Mgt input is sent to SUB Out. The LFE In input should not be routed to SUB Out here:

(If you have less than a full 7.1 system, you will also have some down-mixing, as described in Equalizing your home theater system with the nanoAVR and UMIK-1.)

With the main routing set up, proceed to the LFE Mgt screen. As mentioned above, leave the LFE mix level at 0 dB, and set each of the speaker channels to mix in at -10 dB (right-click on each gain box to set the mix level):

(Note: since we are mixing several signals together here, it is possible that you will clip the subwoofer output on heavy material IF you have the SUB Out level (on the Output screen) set higher than about –6 dB. After completing the procedure below, keep an eye on the SUB Out level meter with typical program material. If you notice clipping on the meter, simply reduce the SUB Out level and correspondingly increase the gain on your subwoofer amp or in your AVR.)

3. Configure crossovers[Top]

Now we will set the crossovers between the subwoofer and the speakers. Click on the LPF block for the front left speaker and link this channel to the front right channel (it's best to start assuming that both channels will have the same crossover frequency).

The nanoAVR allows complete flexibility in setting your crossover frequencies. For the purposes of an example, we will use an 80 Hz crossover frequency. As a starting point, set a 24 dB/octave Linkwitz-Riley low pass filter at 80 Hz:

Now we need to set the corresponding high pass filter for the speaker channel. Return to the Output screen and click on the Xover button for the FL Out channel, and link it to the FR Out channel. As a starting point, choose one of the following:

1. For a smaller sealed speaker with a rolloff frequency of around 80 Hz, set a 12 dB/octave Butterworth high pass filter. This rolloff will add to the acoustic low end rolloff of the speaker and produce a net response that is approximately a 24 dB/octave Linkwitz-Riley filter:

   

2. For a larger sealed speaker or a ported speaker with a rolloff frequency of 60 Hz or below, set a 24 dB/octave Linkwitz-Riley high pass filter:

   

If neither of the above apply, pick the closest. This is only a starting point and will most likely need adjustment based on measurement anyway.

4. Measure and tune [Top]

With the initial setup done, run a measurement sweep through HDMI channel 1. You may wish to run three sweeps:

1. With SUB Out muted, to see the response of only the front left speaker.
2. With FL Out muted, to see the response of only the subwoofer.
3. With neither muted, to see the complete response of the front left channel.

Here is an example showing the separate responses in blue and green, and the combined response in red:

Initially, you may not get a smooth combined response through the crossover region. You will therefore need to make some adjustments:

• Adjust the frequency, type, or slope of the subwoofer low pass filter. If changing the slope, you can make it steeper but do not make it shallower!

• Adjust the frequency, type, or slope of the speaker high pass filter.

• Add delay to either the subwoofer channel or the speaker channel.

With the front left speaker done, measure the front right speaker (HDMI Channel 2). Because of the different location in the room, you may not get as good a response with this channel—if not, make some further adjustments to get the best compromise between both channels. If necessary, you can unlink the channels and set different crossover frequencies or slopes.

With the front left and right channels done, you can proceed to set up crossovers and repeat the measurements and adjustments for the center channel and the surround channels.

5. Additional notes [Top]

The flexibility of the nanoAVR's bass management feature gives you almost infinite control over your signal processing. Above we showed an example with an 80 Hz crossover. Here are some additional notes to help you get the best from your system.

• While it's recommended that corresponding left and right channels have the same high pass frequency and slope, different sets of speakers may benefit from different crossover frequencies. For example, your surround speakers may need a higher crossover frequency than your front speakers.

• You can use a mix of bass-managed channels with channels that do not have bass management. If some of your speakers are large enough with sufficiently deep bass response, simply click on the level setting box on the LFE Mgt screen to turn off mixing of those channels.

• Although the LFE channel on the soundtrack should already be filtered at a frequency no higher than 120 Hz (see "Recommendations For Surround Sound Production"), this is not always true in practice. To ensure that any high frequency content that may be present on this channel is removed, set a Butterworth low pass filter on the Sub In channel at 120 Hz, with a slope of 24 dB/octave or greater.

Here are the final responses of our bass-managed example system with the responses of the front left channel in green, center channel in blue, and LFE channel in red:

You can see that the left and center channels from the disc now have a very flat response down to low frequencies, thanks to bass management! The LFE channel is 10 dB higher, as it should be. You can now proceed to listen to the system and adjust and fine-tune your settings accordingly. Remember, it's really all about how the system sounds to you. Have fun!

References [Top]

• Recommendations For Surround Sound Production (PDF), The Recording Academy's Producers & Engineers Wing.
  
  

In this app note we will show you how to equalize all channels of your home theater system using the miniDSP nanoAVR. Before starting on this app note, you will need to connect the nanoAVR into your system and ensure that it is working correctly. To do so, please read through the User Manual.

• Why equalize?
• What you will need
• 1. Getting set up for measurement
• 2. Set your routing and downmix
• 3. Establish your baseline
• 4. Equalize the front left and right speakers
• 5. Equalize the center and surround speakers
• 6. Equalize the LFE/subwoofer channel
• What's next?

Why equalize? [Top]

For the purposes of home theater, equalization can be divided into two areas. Subwoofer equalization is used to smooth the low-frequency response of the system, and covers the frequency spectrum from 20 Hz (or lower) up to about 120 Hz, depending on the specifics of the system. The goal is greater accuracy: what you hear at the listening position is closer to what the recording engineer intended.

Full-range equalization covers the frequency range above that handled by the subwoofer. It compensates for response variations caused by acoustic reflections and by compromised speaker placement, such as against a wall or into a cabinet. Center and surround speakers in particular are prone to response errors caused by design and placement limitations. Equalizing all speakers in a home theater system to have the same in-room frequency response will ensure optimum creation of soundstage and envelopment.

What you will need [Top]

• A miniDSP nanoAVR HD or nanoAVR HDA (analog outputs). This diminutive unit is the "powerhouse" that we'll be showing you how to use in this app note.

• Room EQ Wizard (REW). Be sure to download the latest version from the Downloads Area for UMIK-1 support.

• A miniDSP UMIK-1. After receiving the UMIK-1, go to the UMIK-1 page to get the calibration file for your unique serial number, and save it as a .TXT file.

• A microphone stand with boom arm, available in music supply stores. (While the UMIK-1 is supplied with a small "table top" stand, best results will be obtained if the microphone is placed in free space, away from sofas or tables.)

1. Getting set up for measurement [Top]

The diagram below illustrates a typical setup used for acoustic measurement. The computer can be connected to HDMI Input #2, so there is no need to disconnect the player connected to HDMI Input #1. The computer will need an HDMI output port, or if you have a Mac with a Thunderbolt port, then a Thunderbolt to HDMI adapter cable can be used.

Setting up the computer to use REW with HDMI output and UMIK-1 input is covered in these app notes:

• Using the UMIK-1 and REW with HDMI output - Windows
• Using the UMIK-1 and REW with HDMI output - Mac

2. Set your routing and downmix [Top]

To start with, go to the Routing screen and ensure that you have a straight-through routing with no bass management. Here is how it looks:

You want the AVR to pass the signal through unaltered, so:

1. Turn off all bass management and down-mixing. Typically this is done by setting all speakers to "large" (even the ones that don't exist).
2. Set all channels to the same level (typically, 0 dB).
3. Turn off all forms of matrix decoding (e.g. Dolby Pro Logic, DTS Neo:6) and any surround simulation or enhancement modes.

Now confirm that you are able to play a test signal through each HDMI channel in turn. A useful tool here is the REW signal generator set to play pink noise (pink noise is easier to localize than a pure sine wave):

Set the volume control on your AVR fairly low to start with, and use the HDMI output channel selector (REW Preferences on Windows, SoundFlowerBed on Mac) to send audio to channel 1. You should hear a hissing sound from the front left speaker. Then use the HDMI channel selector to proceed through each channel in turn.

If you have less than a full suite of 7.1 speakers, you will need to "downmix" some channels. (While your AVR can do this, equalization is more straightforward if the nanoAVR is set to do it.) For example, if you have only two surround speakers, mix the rear left input to the left surround output and the rear right input to the right surround output, like this:

If you don't have a center channel speaker — that is, you are running a "phantom center" — mix the center channel input to the front left and front right outputs at a level of -3 dB, like this:

3. Establish your baseline [Top]

Use REW to measure the in-room frequency response of each of your speakers and subwoofer, one at a time. (You don't need to run a measurement on channels that are down-mixed to other channels.) For best results, point the microphone towards the speaker being measured.

You will end up with a set of up to eight frequency response measurements. Use the REW Overlay window and 1/3rd, 1/6th, or 1/12th octave smoothing to view them. Here is an example showing the left channel in blue, the center channel in green, and the subwoofer in red. (The other speakers are omitted from the example to avoid cluttering the graph.)

Performing this baseline measurement ensures that your setup is ready to proceed with equalization. In some cases, the baseline measurements may reveal issues that will need to be corrected before proceeding. For example, if the subwoofer response has a deep notch, no amount of equalization will be able to correct it and you should try some other positions of the subwoofer in the room.

4. Equalize the front left and right speakers[Top]

Start by equalizing the front left and right speakers. These are usually the highest quality speakers in the system and have fewest placement problems. The goal when equalizing these speakers with an in-room measurement is to obtain a fairly even response that gently slopes down from low frequencies to high frequencies.

To perform an automatic equalization with REW, follow the steps detailed in the app note Auto-EQ tuning with REW. You will need to alter the settings slightly, as follows:

Important items to set are the MiniDSP-96k button (the nanoAVR operates internally at 96 kHz), the settings related to the HF Fall, and the Match Range frequencies.

You may also be interested in trying "manual" equalization. This is where you set the EQ settings directly in the nanoAVR user interface. It is more time-consuming but gives you more control over the final result. You can do this in a different configuration if you like, so you can easily switch between them to compare. Link the PEQ filters for the front left and right channels (generally, it's best to avoid using different EQ settings for the left and right speakers):

The sequence when doing manual equalization is 1. measure, 2. perform an EQ adjustment, 3. measure again. Of the various types of EQ available in the nanoAVR interface, there are two that we suggest as the ones to start with. The first is the PEAK type filter with a negative gain. Use this to "flatten" peaks in the measured response. Here is an example setting:

The second is a high-shelf filter. One use of this type of filter is to correct for the response of the speaker in the mid-range, where placement may alter the designed response. The center frequency is typically a few hundred Hz and the gain may be positive or negative, depending on your particular acoustic situation:

The second use is a gentle high-frequency slope to adjust the "tilt" of the in-room frequency response curve at high frequencies:

You may find that the left and right speakers measure a bit differently. In that case, you will need to pick a compromise. Here are some additional guidelines:

• Use boost filters sparingly. If used, they should have a fairly low Q (1 or less) and a moderate amount of gain (4-5 dB max). Don't try and use a narrow boost to fill in a deep notch in the response.

• View measurements at 1/3rd octave smoothing as a general rule. Switch to 1/12th octave smoothing to check that you are not attempting to correct a narrow dip or peak with a broad filter.

• Even with negative gain, use high-Q filters carefully. If there is a persistent resonance that you can measure in different locations in the listening room, it may be an architectural issue (such as a corridor or stairwell), in which case a narrow filter can help.

• Avoid using EQ to increase the amount of bass from your speakers — leave that to the subwoofer. Your speakers have a natural low-frequency rolloff and trying to change this by too much is a recipe for overdriving amps and woofers.

• Make some additional measurements with the microphone in different locations around your seating area. This can be used as a "sanity check" to ensure that you are not equalizing for issues that vary a lot over the seating area and that you are not "over-equalizing."

This plot shows the front left channel of our example system before any EQ in light blue and afterwards in dark blue:

5. Equalize the center and surround speakers[Top]

Now move on to the center and surround speakers. The goal with these is not only to even out the frequency response, but to also get approximatelythe same in-room frequency curve as the front left and right speakers. You may find that high-shelf and low-shelf filters are particularly useful for this purpose. Also, use the gain adjustment on these channels so that the levels match the front speakers.

This plot shows the result of equalizing the center channel speaker in our example system compared to the front left speaker (the center is in green, the front left in blue):

Due to placement issues, the frequency response measurements of the center and surround speakers, even after equalization, will likely not be as smooth as the front left and right speakers. This is OK — you don't need to get the curves perfectly smooth, just get to a reasonable compromise.

6. Equalize the LFE/subwoofer channel[Top]

To equalize the LFE/subwoofer channel, use the auto-EQ function of REW. This is described in detail in the app note Auto-EQ tuning with REW, so that procedure should be followed carefully. The key difference is that the Equalizer setting must be set to "miniDSP-96k", since the nanoAVR operates internally at 96 kHz.

As the final step, adjust the level of the subwoofer so that the LFE channel measures 10 dB higher than the speaker channels. This plot shows the LFE/subwoofer channel response in our example system before equalization in orange and afterwards in red, with the front left speaker in blue used as the reference for setting the subwoofer level:

You may also want to make use of the target curve feature of REW to create a response that is more emphasized at lower frequencies:

Remember also that the better the response of your subwoofer before equalization, the better the result will be afterwards. Subwoofer response is very sensitive to placement within the room, so experiment with different subwoofer locations in order to achieve the best final result.

What's next? [Top]

If you have full-range (20 Hz to 20 kHz) speakers all around (front left and right, center, and all surrounds), then your job is done! You can sit back and listen. In practice, very few home theaters have a full suite of full-range speakers, and so bass management is required to send low frequency content to the subwoofer. We will cover this in the app note Bass Management with the nanoAVR.

You can continue to fine-tune your system while listening. For example, set up some high-frequency shelving filters to get the overall tonal balance that you prefer. Adjust the level of the subwoofer channel to get the best balance of bass. And so on. In the end, it's really all about how the system sounds to you. Have fun!

The miniDSP U-DAC8 is a compact high-resolution eight-channel DAC with a myriad of uses, from PC-based home theater, multichannel audio, to computer-based active loudspeakers. In this application note we will show you how to use the U-DAC8 with JRiver Media Center to play back multichannel audio on your Mac.

Please note: miniDSP cannot provide support for third-party software. This app note shows you how to set up the miniDSP U-DAC8 with JRiver Media Center for multichannel audio playback but any functions of JRiver Media Center are beyond the scope of miniDSP support.

• 1. Get connected
• 2. Configure
• 3. Play!

1. Get connected [Top]

Nothing could be simpler than connecting the U-DAC8 into your system:

• Analog outputs to your multichannel amplifier/s and subwoofer (or to another device with multichannel analog inputs such as an A/V preamp or receiver)
• Power connector for 5 VDC
• USB connector to your Mac

2. Configure [Top]

We will assume that you already have an JRiver Media Center library set up, which includes multi-channel audio files. These files can be in multichannel (typically 5.1) PCM format up to 192 kHz, or in multichannel DSD format (again typically 5.1).

From the Player menu, select Playback Options.... Under Audio Device, drop down the selector and choose U-DAC8 Output:

Then click OK.

If you want JRiver to perform bass management, from the Player menu, select DSP Studio:

Configure as follows:

1. Click on Room Correction to enable it and display the control panel.
2. Select each speaker in turn to configure it (as in steps 3 and 4).
3. For each speaker, set the highpass crossover.
4. For each speaker, set the lowpass routing and slope.

You will need to set the items in steps 3 and 4 for your speakers and room. You can also adjust the other parameters on this screen like speaker distance and level.

3. Play! [Top]

Now you can browse to an album or file in the JRiver Media Center library (screenshot at top of page) and play by clicking on the Play button. You will hear multichannel audio playing through your system.

You can check the sample rate of the file and the DAC by going to the Player menu and viewing the Audio Path entry. Here is an example for 5.1 channel FLAC:

When playing multichannel DSD files, JRiver Media Center will automatically convert them to PCM format at 176.4 kHz. Here is an example:

That's it, have fun listening to multichannel audio with your miniDSP U-DAC8 and JRiver Media Center! Let us know how you go on our forum.

The miniDSP U-DAC8 is a compact high-resolution eight-channel DAC with a myriad of uses, from PC-based home theater, multichannel audio, to computer-based active loudspeakers. In this application note we will show you how to use the U-DAC8 with Audirvana+ to play back multichannel audio on your Mac.

Please note: miniDSP cannot provide support for third-party software. This app note shows you how to set up the miniDSP U-DAC8 with Audirvana+ for multichannel audio playback but the functions of Audirvana+ are beyond the scope of miniDSP support.

• 1. Get connected
• 2. Configure
• 3. Play!

1. Get connected [Top]

Nothing could be simpler than connecting the U-DAC8 into your system:

• Analog outputs to your multichannel amplifier/s and subwoofer (or to another device with multichannel analog inputs such as an A/V preamp or receiver)
• Power connector for 5 VDC
• USB connector to your Mac

2. Configure [Top]

We will assume that you already have an Audirvana+ library set up, which includes multi-channel audio files. These files can be in multichannel (typically 5.1) PCM format up to 192 kHz, or in multichannel DSD format (again typically 5.1).

In the Audirvana Preferences, select the Audio System tab:

Configure as follows:

1. Click on Change and select U-DAC8 Output from the list.
2. Confirm that the U-DAC8 supports the expected set of sample rates (from 44.1 up to 192 kHz).
3. Ensure that Native DSD Capability is set to None: convert to PCM.
4. The low-level playback options may need to be adjusted depending on your version of Mac OS X.
5. Allocate a good amount of memory for buffering.

(Note that Audirvana does not do bass management. So when playing content without a subwoofer channel (e.g. 5.0), it will be necessary for all speaker channels to be capable of full-range response sufficient for music. Alternatively, if connected to the analog inputs of an A/V receiver, the receiver may be able to perform bass management.)

3. Play! [Top]

Now you can browse to an album or file in the Audirvana+ library (screenshot at top of page) and play it by clicking on the Play button near the top left. You will hear multichannel audio playing through your system.

You can check the sample rate of the file and the DAC in the status area at the top of the window. To the right of the track display is the file format and sample rate, and underneath the play buttons at the left is the sample rate and format being sent to the DAC. For example, for multichannel 5.1 PCM (in FLAC), these two areas are:

When playing multichannel DSD files, Audirvana+ will automatically convert them to PCM format at 176.4 kHz. Here is the display for a 5.0 channel DSD recording:

That's it, have fun listening to multichannel audio with your miniDSP U-DAC8 and Audirvana+! Let us know how you go on our forum.

In this application note, we will show you how to design an active 2-way loudspeaker with the miniDSP 2x4 HD. Using the 2x4 HD, you can either create a conventional (Linkwitz-Riley or Butterworth) crossover, or a linear phase crossover.

• 1. What you will need
• 2. Select the speaker drivers and design the enclosure
• 3. Getting connected
• 4. Set up routing
• 5. Measure and equalize the drivers
• 6. Add the crossover (conventional)
• 7. Add the crossover (linear phase)
• Wrapping up

1. What you will need[Top]

• A miniDSP 2x4 HD (boxed or kit) together with the matching 2x4 HD1 plugin.
• Ability to run acoustic measurements. You will need a measurement program such as the freeware Room EQ Wizard (REW), and measurement hardware for which we recommend the UMIK-1.
• Four channels of amplification. Two stereo amplifiers can be used or a single multichannel amplifier.

This is the block diagram of the 2x4 HD1 plugin. A good approach is to use the output channel PEQ (parametric EQ) to correct for the response of the individual drivers, and the input channel PEQ for overall response shaping and taming room issues. The Xover (Crossover) block can be used to implement a conventional (Linkwitz-Riley or Butterworth) crossover, or the FIR block used to implement a linear phase crossover.

2. Select the speaker drivers and design the enclosure [Top]

If you are starting from scratch, you will need to select the drivers for your speakers. There are literally hundreds of drivers available for DIY use at all price levels, so it's impossible to give specific recommendations here. For a small two-way loudspeaker, a 5" or 6.5" woofer and a 1" dome tweeter are common choices. Peruse the online forums to see what others are using and to ask for recommendations for your particular project.

If you are building your own box, you will need to design it. The most important factor is the internal volume, and if it's a ported box, the size and length of the port. Fortunately, there are a number of free programs that do the complex math for this based on the Thiele-Small parameters of the woofer. For example, a popular Excel-based program is Unibox.

If you are modifying an existing speaker from passive to active, then you have the enclosure and the drivers already. In this case, you will most likely need to remove the internal crossover and add a second pair of binding posts.

3. Getting connected [Top]

This diagram shows how to connect everything up. (Please follow the procedure in the User Manual on initializing the miniDSP and familiarizing yourself with the 2x4 HD1 plugin before doing this.)

It's recommended to put a capacitor in series with each tweeter as shown. This will help to protect the tweeter from any turn-on and turn-off surges from the amplifier, or if you accidentally send low frequency test signals to it.

4. Configure routing [Top]

The 2x4 HD1 plugin allows any input to be routed or mixed to any output. This a key element of its extraordinary flexibility. To implement a two-way crossover, set up the routing as shown in this screenshot. (See the User Manual for how to rename the input and output channels.)

5. Measure and equalize the drivers [Top]

Once you have built the box and mounted the drivers, you will need to measure the drivers one at a time. (You only need to do this for one speaker.) For more information on how to measure a loudspeaker driver, please see our app note Measuring a loudspeaker with the UMIK-1.

Use the PEQ blocks on each output channel to shape the response of each driver so that it is flat over its operating range. Use "Peak" type filters to flatten peaks (with negative gain so they create a notch) and "High-Shelf" and "Low-Shelf" type filters to straighten out the overall response.

Below is an example measurement for a woofer. The various features of the measurement and the areas to correct are marked on the graph, along with the corrected response in light blue. If using an indoor measurement (as in the example), be careful not to correct for peaks and notches caused by the room.

Here is a graph of the tweeter, measured before and after correcting its response. When performing a tweeter measurement, start the sweep at a frequency so as not to strain the tweeter (for example, start at 1 kHz, not 20 Hz).

6. Add the crossover (conventional)

There are two options for implementing the crossover: IIR (conventional) and FIR (linear phase). In this section we will describe the IIR version.

To open the crossover settings screen, click on the Xover button on the Outputs screen. Set a low pass filter on the woofer and a high pass filter on the tweeter. As a starting point, try using Linkwitz-Riley (LR) 24 dB/octave filters. You can then experiment with lower or higher slopes, from 6 dB/octave up to 48 dB/octave.

Then you can measure the response of the complete speaker! Use the output level controls to match the signal levels from the woofer and tweeter. If you have a notch at the crossover frequency, you probably need to invert the polarity of one driver. You will most likely need to fine-tune the crossover settings to get the smoothest response around the crossover frequency:

• Time align the drivers
• Move the filter corner frequency of one driver up or down a little
• Use an asymmetrical crossover, for example BW 18 dB/octave lowpass on the woofer and LR 24 dB/octave high pass on the tweeter
• Adjust the equalization of one driver or the other near the crossover point

This REW plot shows the response of the woofer and tweeter in our example speaker with crossover filters in place, and the combined response after crossover fine-tuning:

7. Add the crossover (linear phase)

Instead of a conventional crossover, the 2x4 HD can be used to implement a linear phase crossover. Each output channel has 1024 taps accessible by clicking on the FIR button on each output channel.

Use the excellent freeware program rephase. See "Example 2: a linear-phase crossover" in the app note The rePhase FIR tool for a tutorial example. Note that for the 2x4 HD1 plugin, you will need to set the taps parameter to 1024 and the rate parameter to 96000. (Instead of 2048 and 48000 as shown in the app note example.)

Here is an example two-way crossover from that app note:

Wrapping up[Top]

That's it for this app note! Have fun, and please let us know about your active loudspeaker experience in our forum.

In this application note, we will show you how to integrate a subwoofer with your existing loudspeakers by using a miniDSP 2x4 HD. You can use an existing preamp, or replace a preamp and a DAC with the miniDSP 2x4 HD. (In the latter case, you will need to program a remote control to adjust volume in the miniDSP 2x4 HD.)

• 1. What you will need
• 2. Get connected
• 3. Configure routing
• 4. Measure and equalize the subwoofer
• 5. Add the crossover
• Wrapping up

1. What you will need[Top]

• A miniDSP 2x4 HD (boxed or kit) together with the matching 2x4 HD1 plugin.
• Ability to run acoustic measurements. You will need a measurement program such as the freeware Room EQ Wizard (REW), and measurement hardware for which we recommend the UMIK-1.
• The subwoofer. In this app note, we will assume that the subwoofer is self-powered (has an integrated amplifier).

This is the block diagram of the 2x4 HD1 plugin. Outputs 1 and 2 will drive the two loudspeakers (via your existing power amplifier) and output 3 will be connected to the subwoofer. (Optionally, a second subwoofer can be connected to output 4.)

2. Get connected [Top]

This diagram shows how to connect everything up. (Please follow the procedure in the User Manual on initializing the miniDSP and familiarizing yourself with the 2x4 HD1 plugin before doing this.)

3. Configure routing [Top]

The 2x4 HD1 plugin allows any input to be routed or mixed to any output. This a key element of its extraordinary flexibility. To implement subwoofer integration, set up the routing as shown in this screenshot. (See the User Manual for how to rename the input and output channels.)

4. Measure and equalize the subwoofer [Top]

Measurements for subwoofer integration are done with the microphone at the listening position and pointed between the two speakers. That way, the in room response of the speakers and subwoofer are measured. See the app note Acoustic measurement with the UMIK-1 and REW.

To measure the subwoofer, mute output channels 1 and 2, and check that both crossover filters in the Xover block of channel 3 are bypassed. Set the subwoofer's crossover frequency control to maximum and set any phase control to the flat position.

Use REW's Auto EQ to generate a set of parametric filters to "flatten" the response. The procedure is described in the app note Auto-EQ tuning with REW.

Then load the generated filters into the PEQ block of channel 3 and run another measurement. Here is an example "before and after" measurement:

5. Add the crossover

In the Xover block of channels 1 and 2, add a high pass filter. There is no hard and fast rule about what frequency and slope to use, but here is a typical example:

(If your speakers are sealed and roll of naturally at 80 Hz, use a BW 12dB/octave filter instead.)

In the Xover block of channel 3, add a low pass filter. Here is a typical example:

Unmute all output channels, mute the right input channel, and run a measurement. Initially, you may not get a smooth combined response through the crossover region. In that case, you will need to make some adjustments:

• Adjust the frequency, type, or slope of the subwoofer low pass filter. If changing the slope, you can make it steeper but do not make it shallower!
• Adjust the frequency, type, or slope of the speaker high pass filter.
• Add delay to either the subwoofer channel or the speaker channel.

Then unmute the right channel and mute the left channel. Repeat the fine-tuning for the right speaker. (Note that if you added delay to the left speaker, you must add the exact same delay to the right speaker.)

Wrapping up[Top]

That's it for this app note! Have fun, and please let us know about your subwoofer integration experience in our forum.

This example shows how to create a FIR filter to adjust the magnitude and phase of a full-range loudspeaker. The measurement, for this example, can be downloaded below, and is taken from a 12" + horn install loudspeaker. 12in_2way_ir_44p1k.wav

About measurements

Taking good loudspeaker measurements is difficult. Nearby objects and room acoustics all impact a loudspeaker measurement, and therefore how representative the measurement is of the loudspeaker itself. This example assumes that you are experienced in taking measurements, and that you understand the limitations of your measurement environment and are familiar with concepts including measurement gating/windowing, averaging and time alignment. Accurate, representative measurements are very important when using the Auto Mag and Auto Phase features in FIR Designer. Inaccurate measurements can result in these automatic functions adjusting the loudspeaker's magnitude and phase response in ways that measure well well for one specific location in the room only, and measure poorly and sound worse everywhere else.

Terms

• LF : Low frequency
• HF : High Frequency
• LPF : Low pass filter
• HPF : High pass filter

Step#1

Start FIR Designer. On the "Import" tab, click "Load." Find and select the measurement file and press "Open." The Import tab is used to load a measurement and adjust its time alignment to remove any bulk delay, and get the phase response in a range we can work with on the following tabs. The measurement isn't time aligned and so the upper plot shows significant phase wrapping.

Step#2

Press the "Find Peak" button. The response is time shifted to bring the peak to time=0. In this loudspeaker, the HF driver lags the LF driver due to the depth of the horn. When aligned to peak, the measurement has much of its phase relatively close to 180 degrees. We have a few choices in moving the measurement to make it usable.

Step#3

Use the "Delay" slider to shift the impulse response so that the peak of the first bump is at time=0. This aligns the impulse response to the upper frequency range of the LF driver, however the HF delay, and resulting high phase wrap, is difficult to correct with filtering. Alternatively, we can align the impulse response to the inverted HF peak.

Step#4

Check the "Flip Polarity" setting and then press "Find Peak." Now most of the phase is near 0 degrees.

Step#5

Select the "Target" tab. Here we can specify the magnitude and phase we would like the filtered loudspeaker to have. If the "Design" radio button is selected, the "Target Design" upper section is used to define a magnitude or EQ profile using three segments and a bass shelf. This section assumes we want flat, zero phase. If the "File" radio button is selected, an impulse response or transfer function can be loaded in the lower section. This "Target File" can be any of the same file formats that can be loaded on the "Import" tab. Here we can use another loudspeaker as the target response, or use a "FIR Designer Target File," created using FIR Designer in "Direct Design" mode to have the magnitude and phase curve we want (including crossovers). If the "Design + File" radio button is selected, the upper and lower responses are combined to make the target response. Leave the "Target" tab set as shown.

Step#6

Select the "Magnitude Adjustment" tab. Here we can use common filter prototypes, like parametric bandpass and shelf filters, to shape the loudspeaker response. The light blue line in the upper plot is the inverted loudspeaker magnitude and with the target response added. The aim is to use the magnitude filter prototypes, on the left, to create a composite magnitude filter - the green line - that approximately matches the light blue line. Here three filters are used to approximately match the light blue line. More filters could be used, however here we will match the response coarsely and sue the "Auto Mag" tab later to address all the ripples. The two red lines in the middle plot show the loudspeaker phase, before and after the green curve filter is applied. The prototype filters can be minimum phase (like regular IIR based filtering), linear phase or maximum phase. It is worth noting the effect of the filtering on the phase (in the middle plot) and choosing filters that result in the phase curve tending towards the target phase. In this example the target phase is flat.

Step#7

Select the "Phase Adjustment" tab. Here we can use all-pass filter prototypes to shape the phase of the loudspeaker. The light red line in the upper plot is the loudspeaker phase, after filtering from previous tabs, then inverted and with the target phase added. (In this example, the target phase is 0 degrees or flat.) This loudspeaker has a 360 degree phase rotation due to the crossover. The aim is to use the phase filter prototypes, on the left, to create a composite phase filter - the green line - that approximately matches the light red line. Here four filters are used to approximately match the upper plot between 100 Hz and 10 kHz, and therefore move the phase closer to the target phase. We will use the "Auto Phase" tab later to address the phase ripples.

Step#8

Select the "Auto Mag" tab. Again, the light blue line in the upper plot is the loudspeaker magnitude after filtering from the previous tabs, then inverted and with the target response added. Here FIR Designer can calculate a magnitude filter to automatically follow the light blue line within a chosen frequency range. Move the "Zero Adjust" slider to bring the light blue line close to 0 dB at approximately 60 Hz and 16 kHz. Then enable the first two auto mag bands as shown.

Step#9

Select the "Auto Phase" tab. Again, the light red line in the upper plot is the loudspeaker phase, after filtering from all previous tabs, then inverted and with the target phase added. (In this example, the target phase is 0 degrees or flat.) Here FIR Designer can calculate a phase filter to automatically follow the light red line within a chosen frequency range. Enable the auto phase band between 80 Hz and 10 kHz, as shown. These end points are close to where the light red line is near 0 degrees.

Step#10

Select the "Export" tab. The upper plot shows the FIR filter in two ways. The dark green line is a plot of the actual coefficients that will be exported or saved to file. The lighter green line shows the absolute magnitude of these coefficients, in dB. The ideal filter needs to be truncated and windowed to make it practically usable in a processor. The lower plot shows both the ideal and the windowed filter. With the default "Filter delay" of 200 samples and "Filter length" of 400 samples, there are large differences between the ideal and truncated filter, especially near 100 Hz. Reducing the differences involves balancing, and likely increasing, the filter delay and the filter length, and possibly changing the choice of window function. The light green dB display in the upper plot can help with this. Generally keeping the filter magnitude at or below -60 dB near the ends of the filter, will keep the error relatively low.

Step#11

To minimise the error between the ideal and windowed filter, adjust the "Filter delay" and "Filter length" so that the ends, in the upper plot, are below approximately -60 dB. Here we have chosen a "Filter delay" of 400 samples and a Filter length of 1600 samples.

Step#12

The upper plot can also show the loudspeaker impulse response before and after convolving with this filter design. Since this design has focussed on making the phase flat, much of the energy should pile up to make a sharper impulse, which is what we see.

Step#13

Since the imported loudspeaker measurement was inverted (on the "Import tab) to move its phase closer to 0 degrees, it may be necessary to invert the FIR filter. Check the "Invert filter polarity when saving", then in the "Format" drop-down, select the desired output file format, click "Save" and save the file as "2-way filter". (The appropriate file extension is added automatically.) Here we have chosen "Binary file (32 bit, float)" which is used by MiniDSP processors.

Step#14

The greatest difference or error will always be at the lower frequency end of the filter; here near 100 Hz. To see the fine difference between the ideal and truncated filter, look at the "Total Error" on the previous tabs. Finally, in the "Project" menu, select "Save", choose a filename and save the project. (The file extension *.fdp is added automatically.)

Further comments

This example uses the default  "Design sample rate" of 48 kHz however the sample rate can be changed at any time. Imported measurements and target responses are stored in the their native sample rate and resampled to the "Design sample rate." The magnitude response or "voicing" of the loudspeaker can be further adjusted by creating a target response, using FIR Designer in "Direct Design" mode, exporting the target filter file and importing the target file on the "Target" tab. The phase rotation, in the low frequency roll-off around 50 Hz, can also be linearised using a combination of filters on the "Phase Adjustment" tab (see Figure 7) and the "Auto Phase" tab (see Figure 9). This will, however, require larger "Filter length" and "Filter delay" settings. 

This example shows how to create FIR filters for a 2-way loudspeaker using FIR Designer by Eclipse Audio. The FIR filters include the crossover responses and they linearise the loudspeaker phase everywhere except at the loudspeaker's low frequency roll-off. Measurements, for this example, can be downloaded below, and are taken from a 5" coaxial loudspeaker.

• high_mag_phase.txt
• low_mag_phase.txt

About measurements

Taking good loudspeaker measurements is difficult. Nearby objects and room acoustics all impact a loudspeaker measurement, and therefore how representative the measurement is of the loudspeaker itself. This example assumes that you are experienced in taking measurements, and that you understand the limitations of your measurement environment and are familiar with concepts including measurement gating/windowing, averaging and time alignment. Accurate, representative measurements are very important when using the Auto Mag and Auto Phase features in FIR Designer. Inaccurate measurements can result in these automatic functions adjusting the loudspeaker's magnitude and phase response in ways that measure well well for one specific location in the room only, and measure poorly and sound worse everywhere else. When designing filters for multi-way loudspeakers, it is important to maintain the relative acoustic time delay between the drivers in the measurements, as this affects how the loudspeakers acoustically sum in the crossover frequency ranges. Most measurement software can remove the bulk acoustic and sound-card delay from a measurement with an "align to peak" or "auto delay" feature. Do this once, say for the highest frequency driver, then fix the delay for all measurements of all drivers. Finally, when taking measurements with a Microsoft Windows PC, choose a sound-card that has ASIO drivers. Many Microsoft Windows MME/WDM sound-card drivers are notorious for having inconsistent latency, which can cause time offset errors between measurements.

Note that the following terms will be used regularly through out this app note:

• LF : Low frequency
• HF : High Frequency
• LPF : Low pass filter
• HPF : High pass filter

LF Driver Response

Start FIR Designer. On the "Import" tab, click "Load." Find and select the measurement file for the LF driver and press "Open." Uncheck "Normalise magnitude to max." This may make the magnitude response - the blue line - disappear out of view. Adjust the "Magnitude offset (dB)" value until the lower-frequency, flatter portion of the spectrum - indicated by the arrow - is at approximately 0 dB. Here the offset value is -76 dB. Note this value for use later with the HF driver. In the "Project" menu, select "Save" and save the project as "LF Filter". (The *.fdp file extension is added automatically.) We will load this project file again later.

HF Driver Response

On the "Import" tab, click "Load." Find and select the measurement file for the HF driver and press "Open." Uncheck "Normalise magnitude to max" and enter the "Magnitude offset (dB)" value, noted from the LF driver. Here the value is -76 dB. Using the same value for the LF and HF driver ensures that the relative levels of the two drivers are maintained. Note here that the HF driver has a higher average magnitude, since it is more sensitive than the LF driver. This is typical of HF drivers. In the "Project" menu, select "Save" and save the project as "HF Filter". (The *.fdp file extension is added automatically.) We will load this project again later.

LF Crossover Target

Considering the LF and HF responses above, here we have chosen a crossover frequency of 2 kHz. (Good loudspeaker designs also takes into account other factors including the directivity of the drivers and their maximum level capabilities, but this is beyond the scope of this example.) In the "Project" menu, select "New."

Check "Direct Design." The tabs will change to the three shown. On the "Magnitude Design" tab, enable a filter and set it to a 4th order Linkwitz-Riley LPF with linear phase; as shown. Make sure that the "Filter delay" and "Filter length" are long enough to ensure there is essentially no difference between the ideal and windowed filter. Here the defaults of 200 and 400 samples are fine since the sides of the filter, in the upper plot, are well below -100 dB. Also check that the "before windowing" and "after windowing" lines in the lower plot are identical. On the "Export" tab, in the "Format" drop-down, select "FIR Designer target file," click "Save" and save the file as "LF Target". (The file extension *.fdt is added automatically.) In the "Project" menu, select "Save" and save the project as "LF Target". (The file extension *.fdt is added automatically.)

HF Crossover Target

Select the "Magnitude Design" tab and change the filter type to Linkwitz-Riley HPF.

Again, make sure that the "Filter delay" and "Filter length" are long enough to ensure there is essentially no difference between the ideal and windowed filter. On the "Export" tab, in the "Format" drop-down, select "FIR Designer target file," click "Save" and save the file as "HF Target". (The file extension *.fdt is added automatically.) In the "Project" menu, select "Save" and save the project as "HF Target". (The file extension *.fdp is added automatically.)

LF FIR filter design

In the "Project" menu, select "Open" and open the previously saved project file "LF Filter.fdp".

On the "Target" tab, select "Design + File." Here a target response is created by combining the upper plot EQ curve with the target file response from the lower plot. Click "Load" and select the target file "LF Target.fdt". The setting "Use Magnitude only" tells FIR Designer to ignore the phase of the loaded target response. Here the phase of the target file doesn't matter since the target file is already linear-phase.

Select the "Magnitude Adjustment" tab. The light blue line in the upper plot is the inverted loudspeaker magnitude and with the target response added. The aim here is to use the magnitude filter prototypes, on the left, to create a composite filter magnitude filter - the green line - that approximately matches the light blue line. Here two LPF filters are used to approximately match the light blue line. More filters could be used, however here we will primarily match the response near 4 kHz, and use the "Auto Mag" tab later to address all the ripples between 200 Hz and 4 kHz.

Select the "Phase Adjustment" tab. The light red line in the upper plot is the loudspeaker phase, after filtering from previous tabs, then inverted and with the target phase added. (In this example, the target phase is 0 degrees or flat.) The aim here is to use the phase filter prototypes, on the left, to create a composite phase filter - the green line - that approximately matches the light red line. Here one 4th order filter is used to move the phase, near 300 Hz, to 0 degrees - shown by the arrow. We will use the "Auto Phase" tab later to address the phase ripples between 300 Hz and approximately 10 kHz.

Select the "Auto Mag" tab. Again, the light blue line in the upper plot is the loudspeaker magnitude after filtering from the previous tabs, then inverted and with the target response added. Here FIR Designer can calculate a magnitude filter to automatic follow the light blue line within a chosen frequency range. Enable the 1st auto mag band between 200 Hz and 4 kHz, as shown. 4 kHz is chosen to be just sufficiently into the LPF stop band - down to -30 dB - to ensure a good summation with the HF driver around the crossover frequency.

Select the "Auto Phase" tab. Again, the light red line in the upper plot is the loudspeaker phase, after filtering from all previous tabs, then inverted and with the target phase added. (In this example, the target phase is 0 degrees or flat.) Here FIR Designer can calculate a phase filter to automatically follow the light red line within a chosen frequency range. Enable the auto phase band between 600 Hz and 6 kHz, as shown. 6 kHz is chosen to be just sufficiently into the LPF stop band - down to just below -30 dB - to ensure a good phase match with the HF driver around the crossover frequency.

Select the "Export" tab. The ideal filter needs to be truncated and windowed to make it practically usable in a processor. To minimise the error between the ideal and windowed filter, adjust the "Filter delay" and "Filter length" so that the ends, in the upper plot, are below approximately -60 dB. Here we have chosen a "Filter delay" of 150 samples and a Filter length of 500 samples. The lower plot shows both the ideal and the windowed filter. To see the fine difference between the two, look at the "Total Error" on the previous tabs.

The greatest difference or error will always be at the lower frequency end of the filter; here near 300 Hz - indicated by the arrow. Shortening the "Filter delay" and "Filter length" will increase the error, but shortening may be necessary to fit the filter into some processors. On the "Export" tab, in the "Format" drop-down, select the desired output file format, click "Save" and save the file as "LF Filter". (The appropriate file extension is added automatically.) Here we have chosen "Binary file (32 bit, float)" which is used by MiniDSP processors. In the "Project" menu, select "Save" and save the project as "LF Filter". (The file extension *.fdp is added automatically.) The save will prompt to confirm overwriting of the previously created project file.

HF FIR filter design

In the "Project" menu, select "Open" and open the previously saved project file "HF Filter.fdp". On the "Target" tab, select "Design + File." Here a target response is created by combining the upper plot EQ curve with the target file response from the lower plot. Click "Load" and select the target file "HF Target.fdt".

Select the "Magnitude Adjustment" tab. Here a HPF and a shelf filter are used to approximately match the light blue line. More filters could be used, however here we will primarily match the response at between near 900 Hz and near 16 KHz, and use the "Auto Mag" tab later to address all the ripples in between.

Select the "Phase Adjustment" tab. Here two 4th order filters are used to move the phase, near 900 Hz and near 9 kHz, to 0 degrees. We will use the "Auto Phase" tab later to address all the phase ripples in between. This particular HF driver has a notch at approximately 11 kHz, possibly due to breakup or possibly due to it's particular coaxial mounting. Since this notch could move in frequency, and since it's quite narrow and not very audible, attempts to flatten it with filtering could sound worse than leaving it alone. Here we will leave it alone.

Select the "Auto Mag" tab. Since we don't want to correct the ~11 kHz notch, enable two auto mag bands - one each side of the notch - as shown. The lower limit of 900 Hz is chosen to be just sufficiently into the HPF stop band - down to -30 dB - to ensure good summation with the LF driver around the crossover frequency.

Select the "Auto Phase tab. Enable the auto phase band between 700 Hz and 9 kHz, as shown. As mentioned previously, we won't attempt to correct the magnitude and phase at the notch frequency.

Select the "Export" tab. To maintain time alignment with the LF driver, it is important to use the same "Filter delay" as used for the LF driver - here 150 samples. Since HF frequencies have shorter wavelengths, HF filtering generally can be done with shorter FIR filters. However it is recommended to start with the LF driver first, to determine the shortest "Filter delay" that will work for both drivers. In the "Format" drop-down, select the desired output file format, click "Save" and save the file as "HF Filter". (The appropriate file extension is added automatically.) Here we have chosen "Binary file (32 bit, float)" which is used by MiniDSP processors. In the "Project" menu, select "Save" and save the project as "HF Filter". (The file extension *.fdp is added automatically.) The save will prompt to confirm overwriting of the previously created project file.

FIR filters inherently cannot make a perfect HP response that goes to -INF dB at DC. To see the LF performance of the filter, check "View range: -100 to 20 dB." Lengthening the "Filter delay" and "Filter length" will make the windowed filter better match the ideal filter at low frequencies.

HF Series Capacitor

In active loudspeaker designs, it is common to use a series capacitor to provide both DC blocking and some LF protection to the HF driver. FIR Designer includes an "IIR Filters" tab where IIR filters that are used inline - analog or digital - can be entered. FIR Designer takes these into consideration in the FIR creation process, so that the combination of IIR and FIR filters gives the desired, filtered loudspeaker response. The "IIR Filters" tab also includes a simple 1st order series capacitor calculator.

Select the "IIR Filters" tab. In the "Series Capacitor Calculator," enter the HF driver impedance and either a capacitor value or frequency, and the calculator calculates the frequency or capacitance. Here for an 8 ohm driver and 5 uF capacitor, the 1st order HPF cutoff is approximately 4 kHz. Enable the first IIR filter and set the filter to a 1st order Butterworth HPF at ~4 kHz. Note how the 1st order filter brings the loudspeaker magnitude response closer to the target, mostly in the 800 Hz to 2 kHz range.

Select the "Magnitude Adjustment" tab. With the IIR filter enabled, the HPF magnitude filter can be lowered in frequency so that the filtered response matches the target at approximately 900 Hz.

Select the "Phase Adjustment" tab. Now with the IIR filter in enabled, the loudspeaker phase has moved further away from the target. Here we add an additional 4th order filter to help bring the filtered phase back to 0 deg. (Each 4th order phase filter has limits of ±90 degrees.) Also, note that less phase shift is needed at 9 kHz to bring the response close to 0 deg.

Select the "Auto Mag" tab. Similar to previously, we use two auto mag bands, however the "Min Freq" of the lower band has moved to 800 Hz.

Select the "Auto Phase" tab. Change the "Min Freq" of the band to 820 Hz so that it starts approximately where the phase is already 0 deg.

Select the "Export" tab. In the "Format" drop-down, select the desired output file format, click "Save" and save the file as "HF Filter w IIR". (The appropriate file extension is added automatically.) Here we have chosen "Binary file (32 bit, float)" which is used by MiniDSP processors. In the "Project" menu, select "Save" and save the project as "HF Filter w IIR". (The file extension *.fdp is added automatically.)

Further comments

This example uses the default  "Design sample rate" of 48 kHz however the sample rate can be changed at any time. Imported measurements and target responses are stored in the their native sample rate and resampled to the "Design sample rate." The magnitude response or "voicing" of the loudspeaker can be adjusted by enabling additional linear-phase EQ filters on the "Magnitude Design" tab for both the HF and LF target projects, re-exporting the target filter files, then re-importing the target files into the LF and HF filter projects. The phase rotation, in the low frequency roll-off around 50 Hz, can also be linearised using a combination of filters on the "Phase Adjustment" tab (see Figure 9) and the "Auto Phase" tab (see Figure 11). This will, however, require larger "Filter length" and "Filter delay" settings.

Wrapping up[Top][Top]

We hope that this great tutorial from Eclipse Audio will serve its purpose of providing a great step by step for your application. We'd like to thank once again Michael @ Eclipse Audio for his constant support and great software. All credits for writting this great tutorial goes to Eclipse Audio. Make sure to purchase their software and support their impressive effort!

Have fun, and please let us know how you go in our forum.

The miniDSP UMA-8 microphone array, with onboard direction detection, echo cancellation, and noise reduction, has a wide variety of applications. In this app note we'll run through its use with Google Assistant and a Raspberry Pi.

This app note is based on this article published by Google. You will need to follow that article closely.

• Why use the UMA-8?
• 1. Getting connected
• 2. Setting up audio
• 3. Setting up Google Assistant
• 4. Running
• Wrapping up

Why use the UMA-8?[Top]

In Google's hardware setup page, it's suggested that a cheap USB microphone be used. (The Raspberry Pi does not have an inbuilt microphone.) This is not an optimum solution. Instead:

• The UMA-8 has beam-forming running across the 7 microphones, which improves voice detection.

• The UMA-8 also has echo cancellation and noise reduction, to reduce the effects of non-voice sounds (like music playing) and noise (traffic, kitchen noises etc).

1. Getting connected[Top]

Connect your Raspberry Pi to a keyboard, mouse, and an HDMI monitor or TV. Plug the UMA-8 into one of the USB ports. (It is powered over USB).

(Google's hardware setup page provides a number of different setup methods, including "headless." For simplicity we suggest starting with a connected monitor, keyboard and mouse. Later on you can try a headless version.)

Install Raspbian onto a micro-SD card and insert it into the Pi. The Google page linked above suggests using NOOBS. We just downloaded Raspbian Stretch with Desktop and burnt it onto an SD card.

Power on the Raspberry Pi. After a short time you should see a desktop appear on your monitor. Click on the Raspbian icon at the top left, select Preferences then Raspberry Pi Configuration. Set your timezone on the Localization tab. You may at this time also wish to enable SSH.

You may like to explore a little further. If you are using a Raspberry Pi 3 with inbuilt Wifi, use the Settings (top right of screen) to join your wireless network. You may also want to change your keyboard layout to U.S., as it defaults to a U.K. layout.

2. Setting up audio[Top]

We connected our Raspberry Pi to the second input of a nanoAVR HD so the audio output from the Pi goes through a home theater system with room EQ. We bet Google Assistant never sounded so good :)

The instructions on the Google page Configure and Test the Audio aren't quite in the right order. You will need to open a terminal on the Pi to do this. First, list your output audio devices by typing

  alist -l

Here is our output. Note that the HDMI output is card 0 and device 0:

**** List of PLAYBACK Hardware Devices ****
card 0: ALSA [bcm2835 ALSA], device 0: bcm2835 ALSA [bcm2835 ALSA]
  Subdevices: 8/8
  Subdevice #0: subdevice #0
  Subdevice #1: subdevice #1
  Subdevice #2: subdevice #2
  Subdevice #3: subdevice #3
  Subdevice #4: subdevice #4
  Subdevice #5: subdevice #5
  Subdevice #6: subdevice #6
  Subdevice #7: subdevice #7
card 0: ALSA [bcm2835 ALSA], device 1: bcm2835 ALSA [bcm2835 IEC958/HDMI]
  Subdevices: 1/1
  Subdevice #0: subdevice #0
card 1: UAC20 [miniDSP micArray XVSM UAC2.0], device 0: USB Audio [USB Audio]
  Subdevices: 1/1
  Subdevice #0: subdevice #0

Now list your input audio devices by typing

  arecord -l

Here is our output. Note that the UMA-8 output is card 1 and device 0:

**** List of CAPTURE Hardware Devices ****
card 1: UAC20 [miniDSP micArray XVSM UAC2.0], device 0: USB Audio [USB Audio]
  Subdevices: 1/1
  Subdevice #0: subdevice #0

Now you will need to edit the file .asoundrc in your home directory /home/pi so that it looks like this (note that the card and device number for pcm.mic and pcm.speaker are as detected by the above commands):

pcm.!default {
  type asym
  capture.pcm "mic"
  playback.pcm "speaker"
}
pcm.mic {
  type plug
  slave {
    pcm "hw:1,0"
  }
}
pcm.speaker {
  type plug
  slave {
    pcm "hw:0,0"
  }
}

To test that the UMA-8 is working, enter this and then speak aloud near the UMA-8:

  arecord --format=S16_LE --duration=5 --rate=16000 --file-type=raw out.raw

To play the file back, enter this:

  aplay --format=S16_LE --rate=16000 out.raw

3. Setting up Google Assistant[Top]

You will now need to Configure a Developer Project and Account Settings. (Follow the instructions on that page. There are a lot of steps, so take your time and make sure you get everything done. Also, you might want to create a separate Google account for this purpose using e.g. a new gmail address.)

Now it's time to install the application. Follow the instructions on the Download the Library and Run the Sample page.

4. Running[Top]

Now you can start the sample application (finally!). In a terminal window, enter

  google-assistant-demo

Then say "Hey Google" followed by a question, like this: "Hey Google, where am I?" or "OK Google, what is my name?".

By default, the UMA-8 is configured for automatic direction detection. You can alter the behavior with the switches on the board as follows:

• SW1: enables and disables microphone beam-forming.
• SW3: manually select direction of beam-forming (if beam-forming is turned on and automatic detection is turned off).
• SW4: enables and disables automatic detection of microphone direction.

(The LEDs around the edge of the board will indicate the direction of beam-forming.)

After some experience with the setup, you may wish to tweak the audio processing parameters of the UMA-8. To do this, you will need to unplug it from your Raspberry Pi and plug it into your Windows or Mac computer. Start the miniDSP micArray configuration program and click the Connect button to change parameters. For example, to turn on the automatic gain control and increase mic signal levels, try settings like this:

Wrapping up[Top]

That's it for this app note! Have fun, and please let us know about your UMA-8 and Raspberry Pi with Google Assistant experience in our forum.

The miniDSP UMA-8 microphone array, with onboard direction detection, echo cancellation, and noise reduction, has a wide variety of applications. In this app note we'll run through its use with Cortana, the "intelligent virtual assistants" from Microsoft.

• Why use the UMA-8?
• Using the UMA-8 with Cortana
• Further experimentation
• Wrapping up

Why use the UMA-8?[Top]

By default, Microsoft Cortana uses the inbuilt microphone in your computer. This is fine if you are sitting at the computer, but for use of these programs as a "far field" hands-free assistant a more sophisticated microphone is needed. The UMA-8 has:

• Beam-forming running across an array of 7 microphones to improve voice detection and eliminate extraneous noises.

• Echo cancellation and noise reduction to reduce the effects of non-voice sounds (like music playing) and noise (traffic, kitchen noises etc).

The UMA-8 is "plug and play" – you do not have to configure anything to make it work with Cortana. If you wish, however, you can use the miniDSP plugin to tweak the processing parameters of the microphone array (recommended for advanced users only!)

Using the UMA-8 with Cortana[Top]

Plug the supplied USB cable into the micro-USB port on the UMA-8, and plug the other end into a spare USB port on your Windows 10 computer. Open the Control Panel and go to Manage Audio Devices. Select the Recording tab and set the miniDSP micArray as the default device:

To enable Cortana to respond to hands-free voice commands ("Hey Cortana") on Windows 10, you will need to enable it. This is described in this article on c-net:

• How to enable 'Hey, Cortana' on Windows 10

And that's it! You should now be able to say "Hey Cortana" from anywhere in the room to activate it. Check out Microsoft's help page for more details on how Cortana can help you.

Further experimentation[Top]

By default, the UMA-8 is configured for automatic direction detection. You can alter the behavior with the switches on the board as follows:

• SW1: enable and disable microphone beam-forming.
• SW3: manually select direction of beam-forming.
• SW4: enable and disable automatic direction detection.

Note that the LEDs around the edge of the board (on the underside, if the switches are on the top) will indicate the direction of beam-forming.

After some experience with the setup, you may wish to tweak the audio processing parameters of the UMA-8. Start the miniDSP micArray configuration program and click the Connect button to change parameters. For example, to turn on the automatic gain control and increase mic signal levels, try settings like this:

Wrapping up[Top]

That's it for this app note! Have fun, and please let us know about your UMA-8 and Cortana experience in our forum.

The miniDSP UMA-8 microphone array, with onboard direction detection, echo cancellation, and noise reduction, has a wide variety of applications. In this app note we'll run through its use with Siri, the "intelligent virtual assistant" from Apple loaded on OSx.

Note: this app note applies only to desktops and laptops running MacOS. The UMA-8 typically requires more power than can be supplied by phones and tablets.

• Why use the UMA-8?
• Using the UMA-8 with Siri
• Further experimentation
• Wrapping up

Why use the UMA-8?[Top]

By default, Siri would use the inbuilt microphone in your computer. This is fine if you are sitting at the computer, but for use of these programs as a "far field" hands-free assistant a more sophisticated microphone is needed. The UMA-8 has:

• Beam-forming running across an array of 7 microphones to improve voice detection and eliminate extraneous noises.

• Echo cancellation and noise reduction to reduce the effects of non-voice sounds (like music playing) and noise (traffic, kitchen noises etc).

The UMA-8 is "plug and play" – you do not have to configure anything to make it work with Siri. If you wish, however, you can use the miniDSP plugin to tweak the processing parameters of the microphone array (recommended for advanced users only!)

Using the UMA-8 with Siri[Top]

Plug the supplied USB cable into the micro-USB port on the UMA-8, and plug the other end into a spare USB port on your Mac. Open System Preferences and then Sound, and select the miniDSP micArray as the input source:

To enable Siri to respond to hands-free voice commands ("Hey Siri") on the Mac, you will need to a. set up a keyboard shortcut and b. map the keyboard shortcut to a voice command. This is described in this article on MacWorld:

• How to make Siri activate when you say 'Hey Siri' to your Mac with macOS Sierra

And that's it! You should now be able to say "Hey Siri" from anywhere in the room to activate it. Check out the section Apple's help page for more details on Siri.

Further experimentation[Top]

By default, the UMA-8 is configured for automatic direction detection. You can alter the behavior with the switches on the board as follows:

• SW1: enable and disable microphone beam-forming.
• SW3: manually select direction of beam-forming.
• SW4: enable and disable automatic direction detection.

Note that the LEDs around the edge of the board (on the underside, if the switches are on the top) will indicate the direction of beam-forming.

After some experience with the setup, you may wish to tweak the audio processing parameters of the UMA-8. Start the miniDSP micArray configuration program and click the Connect button to change parameters. For example, to turn on the automatic gain control and increase mic signal levels, try settings like this:

Wrapping up[Top]

That's it for this app note! Have fun, and please let us know about your UMA-8 and Siri experience in our forum.

The miniDSP UMA-8 microphone array, with onboard direction detection, echo cancellation, and noise reduction, has a wide variety of applications. In this app note we'll run through its use with Watson, the "intelligent virtual assistant" from IBM to build a ChatBot. For those of you wondering what is a chatbot? It is a service, powered by rules that you interact with via a chat interface.

Why use the UMA-8?[Top]

By default, a chatbot requires a microphone to capture audio and one could say you could use a random 2 dollar plugin mic. This is fine if you are sitting at the computer, but for use of these programs as a "far field" hands-free assistant a more sophisticated microphone is needed. The UMA-8 has:

• Beam-forming running across an array of 7 microphones to improve voice detection and eliminate extraneous noises.

• Echo cancellation and noise reduction to reduce the effects of non-voice sounds (like music playing) and noise (traffic, kitchen noises etc).

The UMA-8 is "plug and play" – you do not have to configure anything to make it work with RPi. If you wish, however, you can use the miniDSP plugin to tweak the processing parameters of the microphone array (recommended for advanced users only!)

What you need:

• 1 x Raspberry Pi 2 or 3 along with a USB power supply for your Pi
• 1 x Keyboard
• 1 x HDMI Display
• An internet connection
• 1 x pair of Headphones (if your HDMI display doesn’t have inbuilt speakers)
• 1 x miniDSP UMA-8 USB Microphone Array
• An IBM BlueMix Account (see below details for registration)

As for skill-set required to put this demo together, you'll need:

• Some basic experience with the Raspberry Pi platform
• Some basic Bash skills (cd, ls, etc.)
• Basic Nano text editor skills (opening and saving)
• An eager mind, ready to learn how to make a Watson chat-bot!

Connecting Everything Up

First up, let’s build our bot! Take your Raspberry Pi and connect up the keyboard, HDMI display and headphones (if you need them). Next, you’ll need to plug in the UMA-8 USB Microphone Array, simply plug it into a USB port! That’s all there is to it as the UMA-8 is plug&play with RPi.

Preparing the Pi

Before powering on the Pi, you’ll need to download Raspbian Jessie LITE from the Raspberry Pi Foundation. Ensure you get the LITE version which doesn’t include a desktop. Burn the downloaded image onto a micro-SD card. Plug the micro-SD card into the Pi and plug in the power. You will see your monitor or TV come up with the a login prompt.

To log in, use the default Raspberry Pi login details.

Username: pi

Password: raspberry

If you have an Ethernet (wired) internet connection available, you're all set. Otherwise, you’ll need to setup the Wifi. Follow this guide to configure the Wifi on your Pi.

Downloading and Installing NodeJS

Once you’ve hacked your way into the pi, it’s time to start installing the brains of our bot. First up we need to install NodeJS. We can easily download node from our inbuilt package manager. . Run these commands to get NodeJS downloaded and installed:

curl -sL https://deb.nodesource.com/setup_6.x | sudo -E bash -
sudo apt-get install -y nodejs

Getting the Bot’s Brains Connected Up

Next up, we need to install the application that NodeJS will run to allow our bot to hear (via the UMA-8) and speak (via the speakers in your display, or headphones). First download the repository which contains the code:

git clone https://github.com/minidsp/uma8-watson-chatbot mic-array

cd mic-array

Now, let’s install all of the extra bits and bobs this application requires to run:

sudo apt-get install libasound2-dev
npm install

Once the installation is complete, you’ll be ready to jump into IBM Watson

Configuring IBM Watson

Now that our Pi is completely ready to go, let’s put it aside for a second, we need to configure our bot’s brain so he knows how we’d like him to respond to our questions! If you haven’t already, sign up for IBM BlueMix. Once you’re logged in, navigate to the BlueMix Services Dashboard.

Click the big “Create Watson service” button in the middle of the screen.

Next, you’ll be presented with a catalogue of all the service IBM BlueMix has to offer. Let’s skip down to the “Watson” section in the sidebar.

We now need to add all 3 of these Watson services, one at a time:

• Conversation
• Speech to Text
• Text to Speech

First up, let’s install the Conversation service. Click the tile (outlined in red above). Once the page has loaded, we can leave the default settings and choose the “Create” button in the bottom left corner.

Lastly, we need to get the “Service credentials”. There details are used by our app will to connect to Watson.

Click the “View credentials” button to expand the tile and then copy the lines which contain the “username” and “password” into a text document on your computer. We’ll need these later, so make sure that they’re clearly labelled as “Conversation Creds” or similar.

Once you’ve got those detailed stored on your computer, we can move on to adding the other two services. Repeat this process and copy the credentials for the other 2 services too. NOTE: You will need all 3 pairs of credentials in order for the app to work.

Programming our Bot with Watson Conversation

Now that your BlueMix account is setup and you have all of the required services created, it’s time to write the logic for your chat bot. IBM have written an excellent blog to explain the basics and get you started. Check it out from this link.

Once you have created a bot you’re happy with, you’ll need to note down it’s workspace ID so you can tell your Pi which bot to use. To get your workspace ID, navigate to the conversation homepage by clicking the “Watson Conversation” link in the top-left of the bot configuration page.

Next, select the three dots in the top-right corner of the workspace tile you’d like to do and choose “View details”

Finally, copy down the “Workspace ID”. There’s a small, handy button next to the workspace ID which allows you to copy it to your clipboard.

Telling our Pi how to talk with Watson

Now that we have finished the process of retrieving our Watson credentials, we need to let our bot know to login to Watson. Jump back in front of your Pi, and ensure you’re in the mic-array directory which contains the app. We need to edit the “settings.js” file to include these details. Open the file with this command:

sudo nano settings.js

Now, replace each of the settings outlined in this file, with the settings you recorded earlier. You will also notice the workspace ID setting under the “Conversation”. Type in the workspace ID which we saved earlier here.

You’ll also see a setting to include a name for your bot. By default, his name is Duncan and he will respond when ever you address him by his name. You can change these names if you like!

Once your app is configured, it’s ready to run!

node app.js

Further experimentation[Top]

By default, the UMA-8 is configured for automatic direction detection. You can alter the behavior with the switches on the board as follows:

• SW1: enable and disable microphone beam-forming.
• SW3: manually select direction of beam-forming.
• SW4: enable and disable automatic direction detection.

Note that the LEDs around the edge of the board (on the underside, if the switches are on the top) will indicate the direction of beam-forming.

After some experience with the setup, you may wish to tweak the audio processing parameters of the UMA-8. Start the miniDSP micArray configuration program and click the Connect button to change parameters. For example, to turn on the automatic gain control and increase mic signal levels, try settings like this:

Wrapping up[Top]

That's it for this app note! Have fun, and please let us know about your UMA-8 and Watson Chatbot experience in our forum.

The miniDSP UMA-8 microphone array, with onboard direction detection, echo cancellation, and noise reduction, has a wide variety of applications. In this app note we'll run through its use with Amazon's Alexa "intelligent personal assistant" and a Raspberry Pi.

This app note is based heavily on the article published by Amazon for their Alexa sample application.

• Why use the UMA-8?
• 1. Getting connected
• 2. Setting up
• 3. Running
• Limitations
• Wrapping up

Why use the UMA-8?[Top]

In the Amazon example article for the Raspberry Pi, it's suggested that a cheap USB microphone be used. (The Raspberry Pi does not have an inbuilt microphone.) This is not an optimum solution. Instead:

• The UMA-8 has beam-forming running across the 7 microphones, which improves voice detection.

• The UMA-8 also has echo cancellation and noise reduction, to reduce the effects of non-voice sounds (like music playing) and noise (traffic, kitchen noises etc).

The UMA-8 is "plug and play" – you do not have to configure anything to make it work with the Raspberry Pi and Alexa, just plug it into your Pi and follow the instructions!

1. Getting connected[Top]

Connect your Raspberry Pi to a keyboard, mouse, and an HDMI monitor or TV. For the UMA-8, just plug it into one of the USB ports. (It is powered over USB).

When installing the Alexa application, you will have a choice on whether to output audio from the 3.5mm analog jack or over HDMI. We selected HDMI and connected the Pi directly to the second input of a nanoAVR HD so the audio output from the Pi goes through a home theater system with room EQ. We bet Alexa never sounded so good :)

Before powering on the Pi, download Raspbian from the Raspberry Pi Foundation and burn it to a micro-SD card. Plug the micro-SD card into the Pi and plug in the power. You will see your monitor or TV come up with the Raspbian desktop.

At this point, you may like to explore a little. If you are using a Raspberry Pi 3 with inbuilt Wifi, use the Settings (top right of screen) to join your wireless network. You may also want to change your keyboard layout to U.S., as it defaults to a U.K. layout.

2. Setting Up[Top]

All of the steps are documented in the Amazon article. You will need to create an Amazon developer account and "register" your Raspberry Pi. Then, using the terminal in Raspbian, download and install the sample application.

You will need to edit a text file to enter the credentials Amazon generated for your Pi. When you do so, be aware that for the field "Product ID" you must use the value that Amazon labels "Device Type ID." (If you use "Security Profile ID", it won't work and you will have to start again.)

With the credentials entered, run the installer. Select Yes to the prompts asking if you want to download required software. You will also need to select audio output from the 3.5mm jack or HDMI. (We used HDMI.) The installer may take half an hour or more to run.

3. Running[Top]

To start the Alexa client, follow the steps in the Amazon article. Be sure to follow all steps in the instructions exactly!

You will need to open three terminal windows and execute a command in each of them. After entering the second command, wait until a window pops up asking you to authenticate. Clicking the Yes button will open a browser window – you will need to log into your Amazon developer account and then click "Okay."

The third command starts the "wake word" service, so you can say "Alexa" to wake up the device. You don't have to use this. Instead, you can just click the "Listen" button on the screen.

Once you have completed all steps, just say "Alexa". You should get a beep in response, after which you can ask it any question you like.

By default, the UMA-8 is configured for automatic direction detection. You can alter the behavior with the switches on the board as follows:

• SW1: enables and disables microphone beam-forming.
• SW3: manually select direction of beam-forming (if beam-forming is turned on and automatic detection is turned off).
• SW4: enables and disables automatic detection of microphone direction.

(The LEDs around the edge of the board will indicate the direction of beam-forming.)

After some experience with the setup, you may wish to tweak the audio processing parameters of the UMA-8. To do this, you will need to unplug it from your Raspberry Pi and plug it into your Windows or Mac computer. Start the miniDSP micArray configuration program and click the Connect button to change parameters. For example, to turn on the automatic gain control and increase mic signal levels, try settings like this:

To do more, use the Alexa app on your iOS or Android device to browse for and add "skills."

Limitations[Top]

There are some things to be aware of:

• If you are outside the USA:

  • You will have difficulty getting the companion Alexa app, as it's only available in the US Apple App Store and Google Play Store. There are workarounds, at least for the iOS version, like here.

  • You will not be able to set your location to outside of the USA.

• Some things are not supported with this sample/Raspberry Pi version, like Amazon Music. You can still add skills with the Alexa app.

Wrapping up[Top]

That's it for this app note! Have fun, and please let us know about your UMA-8 and Raspberry Pi/Alexa experience in our forum.