What does Signal Processing mean? May it be an audio or video signal, a reading from a sensor, we typically speak of signal processing from the moment we're trying to process something out of this data. So far so good?

Three categories of signal processing exist: Analog, Discrete and Digital. In our case, we focus on Digital Audio Signal Processing since it allows greater flexibility, stability and better noise performance. We first sample and quantize the analog signal or in other words convert the signal to the digital domain. From there, we use a DSP (Digital Signal Processor) to process the digital signal (0's and 1's). What's a DSP? It is a processor optimized for mathematical operations which are very common in signal processing.

What are the applications and why we need these DSPs?  In the case of audio systems, level control, filtering, equalization, mixing, limiting or compressing are just few examples of signal processing applications. Don't worry about what they should be used for as we'll get back to it in the other FAQ posts.

Alright, so let's summarize! Digital Audio Signal Processing is an efficient way to easily modify the signal with the least impact on the audio quality. Digital Audio Processors save time, headaches and cost of installation with unparalleled flexibility compared to analog crossovers.

Digital crossovers are active crossovers made with a set of filters in a digital signal processors. A low pass, a bandpass and a high pass filter is all you need to build an active three-way box. However, designers still need to be careful with amplitude and phase response of the system at the cut-off frequency of each filter.

With our MiniDSP kits, there will be no need for some fancy calculations or other complex  equations.  The resulting frequency reponse is plotted for you on a graph to make sure you have full control over your crossover. Thanks to the our wide range of filters, you can easily modify your settings with a couple of clicks and reconfigure your system accordingly.

Put simply, audio distortion is to some extend present everywhere around us and our best hope is to minimize it such that our brain doesn't notice it.  When it comes to audio and loudspeakers, let's make clear that there isn't such thing as a "distortion free" loudspeaker.

So where does the distortion comes from?

Many places within the audio chain may be responsible, and most likely the loudspeaker will be a strong contributor. Because of mechanical and acoustic limitations, loudspeaker drivers simply cannot reproduce a wide range of frequencies without distortion.

Crossovers to the rescue!

Coming to the rescue are crossovers, a set of filters used to separate frequency range for a specific loudspeaker driver, e.g. a subwoofer optimized for a  frequency range from say 20 to 150Hz will need a crossover to limit the bandwidth of the signal being fed to the cabinet.

Passive versus Active?

Passive crossovers apply to electronic filters that do not require external power supplies (passive electronic). In their basic essence,  coils and capacitors perform the task of filtering the audio signal after amplification and separating the frequencies to the correct driver. They are the most common crossovers out there, as they are low cost to manufacture and handy since you only need a single amplifier channel per loudspeaker. The real downside is their lack of efficiency and how the components will affect the flatness of the frequency reponse. This disadvantage causes considerable grief to loudspeaker designers, battling to improve low to medium overall acoustic performance.

Active crossovers, on the other hand, require an external power supply and will separate the signal before amplification. They, therefore, require one amplifier channel per loudspeaker network. (e.g. Low/Mid/High will require 3 amp channels). No more impedance loading interaction from the passive networks, what you design is what you get! Time alignment couldn't be easier and with cost of amplification coming down, active crossovers are winning hands down over passive crossovers. Try one our miniDSP kits to hear the difference!

For more information about digital crossovers, follow this link.

 First off all, you're not the first one with this question, so don't feel bad about asking about it!

The first thing to clarify is the difference between "Active filters", "Active speakers" and "Digital Filters" before we get any of you confused:
- Active filters (as opposed to passive) is made up of active electronic components (Op Amp) to build some filters. Beware as this term is sometimes mixed up with "Digital filters"
- Active speaker is a general term to describe a multiway speaker which has multiple channels of amplification and an active or digital filter to control crossovers.
- Digital filters would typical use a "Digital Signal Processor" (DSP) which will perform filters using a specialized processor, optimized for running the math equations that's required to process the audio signal.  

So hopefully a short and sweet definition which will clarify a couple of terms but in the end, is only an invitation to have you read a bit more online or on our website basic concepts of digital crossovers and digital signal processing.

Before doing any digital signal processing, we need to digitize the analog signals to 0's and 1's using Analog to Digital Converters (ADC) and vice versa with Digital to Analog converter (DAC) at the ouput ends. Two very important product design choices then come to play:

* The sampling frequency, i.e. how often do we want the ADC to take a reading of the analog signal, and turn it into a digital signal?  A simple rule of thumb says: whatever signal we want to digitize, sampling frequency should be at least twice the maximum frequency of the signal (this is also called Nyquist-Shannon theorem). With the human-ear audible spectrum ranging from 20Hz to 20 kHz, the sampling frequency should at least be above 40 kHz. While CD players use 44.1 kHz and DVD use 192kHz, most professional audio products are standardized to 48kHz and 96kHz.

* The bit resolution (also called bit depth) is the second important concept to pay attention to. It basically dictates the number of bits used to represent each sample. Larger the resolution translates to better dynamic range and signal to noise ratio (SNR) performance of the product. That's because the converters will be able to better represent very small fluctuations of the audio signal. Once again a simple formula helps us calculate the minimum bit depth required to achieve a specified dynamic range:

DR = 1.76 + 6.02* Bit depth

e.g. CD players typically use 16 bit depth.  Therefore, the maximum achievable dynamic range of the system, no matter how good the DAC is, would be:  1.76 + 6.02*16 or about 98dB. To achieve better dynamic range, DVD, Blue Ray players and most professional audio products use 24 bit depth.
Note that ultimately this formula yields results at 24 bits that cannot be achieved in the real world due to the IC limitation of about a maximum of 120dB dynamic range.

To summarize,  sampling rate and bit depth of audio converters both matter when trying to achieve higher dynamic range for better quality audio. To make sure you get the best digital audio quality out of our platforms, all our products are standardized to 24bit, 48/96kHz converters. Wait till you hear the difference!

A parametric equalizer is one of the most common filtering block in audio processing.  They are fully configurable for all three parameters:

• Frequency: The filter boost/ cut frequency where the filtering should happen
• Q: A measure of the filter sharpness, i.e. how wide do you want the filter to boost/ cut the signal. A low Q (e.g. 0.5) means a wide effect of the filter looking like a hill. A high Q (e.g. 20) means a sharp effect looking like a sharp peak/dip.
• Boost/ Cut: Controls whether you want to remove or add signal at a specific frequency

A graphic equalizer is basically a bank of filters used to boost or cut at a specific frequency. Typically the boost or cut goes between +/-12dB and with a constant Q. The number of filters depends on the application and is typically expressed as octave separation between filters. E.g. 1/3 octave graphic equalizer means the center frequency of each filter is placed at 1/3 of an octave away from the previous and next filter. 1/3 Octave filters are typically 31 bands while 2/3 octave are only 15 bands.

What are the applications? Graphic equalizers are the perfect tool to modify the frequency response of your system however they use a substantial amount of DSP power (31biquads) which is wasted most of the time. A Parametric Equalizer is more powerful for that reason.

No. MiniDSP kits only require software configuration at the initialization. Once the board configured, the configuration is automatically saved to the board and will be reloaded after any power cycle. The board does not have to be connected to a USB from a PC source anymore. A miniUSB charger will serve the same purpose of providing DC power to the board. We recommend this being said that you make sure to select a good quality charger to keep the noise as low as possible.

Yes. There is only a one time charge for our plug-ins so once purchased, you can use them as you wish.

miniDSP Plug-ins are based on Adobe Air and Flash environment. If your plug-in installation went well but the software is not showing on your screen, your computer is missing some software updates to properly display the software. Simply connect your PC to Internet as the software will automatically search and install them.

All our plug-ins are tested to be fully compatible with Windows (WinXP SP2, Vista, Win 7) and Mac OsX.

For Windows 7, we've however noticed that in some cases, some dialog boxes would not appear (depending on your Win7 settings). To prevent this issue from happening, simply enable the WinXP 2 compatibility mode as described in the following microsoft video.

A very good question requires a very long answer!

First of all, so you know that we're all DIYers at miniDSP and like most of you, always on the look out for a better deal! Running a company is a different story this being said. Sure, we could provide 2 or 3 free plug-ins, get them out of the door in little time and simply make what is commonly called as a fixed DSP architecture. Most manufacturers out there chose that path since it is much easier to develop (one time engineering cost) & support (single firmware).

So why aren't you doing it?

At miniDSP, for one, we're not a multi-national company. We're a group a passionate people who believe in building a low cost DSP community. We want to make sure that we support you as much as we can with our engineering department, not a call center with a random guy who has no idea about audio. Our mission isn't only to build a product, it's also to support it and give you the most out of it. Everytime we will release a new plug-in, engineers will spend valuable time redesigning the UI, rebuilding a new audio configuration and carefully testing it to make sure it meets our high standards. Unfortunately, even with our wicked gardening skills, we still haven't figured out a way to get that famous "free money tree" to pay for it... :-( So in the end,  we have to account for a small charge on each plug-in.

What do you get out of it?

Free un-limited upgrades . Once a plug-in purchased, you will get free access to new features & addon we will make to future versions of the plug-in. All our customers will receive a download link once the new plug-ins are released to the market.

Hope this clarifies our mission and how we plan to support growing the miniDSP community. For any comments, send us an email and we'll be glad to answer your questions.

On one side, you have a PC running its task from a CPU, which granted is very powerfull but also meant to perform a whole range of processing tasks. Audio being one of them.

On the other hand, you have a dedicated processor (called Digital Signal Processor) like our miniDSP kit, dedicated to running only one type of operation, audio signal processing.

What's the difference between the two? The main drawback, that even the most powerful PC will struggle to come around is the minimum overall latency.
What's latency? As a general term, it's the overal time it takes for the signal to travel from the input to the output. Taking the case of an analog input, it first needs to be converted to digital by an ADC, then the signal being processed by a processor (CPU or DSP), to finally get converted back to analog by the DAC.

The major issue in a PC environment using a filter solution is that a couple of variables needs to be added to this path. The audio drivers (even with Asio at best 5 to 25ms), the processing by a software running on an OS that may not be a Real Time OS, and finally feeding the signal back to the sound card using the driver again all add up together. The end result is an overall latency potentially streching to the order of 100ms or more. Compare this specification to the miniDSP fixed overall hardware latency (input to outputs, inc ADC/DAC + processing) of about 1.5 to 2ms (depending on the plug-in), it is clearly obvious that dedicated processing is the way to go for minimizing latency by a 100x order.

Hope this explanation makes sense and feel free to send us your comments on the best latency you may have achieved recently using filter software applications. The information listed here was written to the best knowledge of the experience we had over the years. With technology going so fast, who knows what it

Simple answer. In any of your miniDSP kit design, start by listing the number of inputs and outputs which will help you define how many DSP kits you need, then start to worry about the plug-in. So let's say you want a stereo 4 way crossover:

- That's equivalent to 2 inputs by 8 outputs. Two solutions from there on: 1 x miniDSP 2x8/8x8 unit OR 2 x miniDSP 2x4 units (kit or boxed). The advantage of the 2x8/8x8 unit allowing you to control both speakers from the same plug-in. In the case of 2 x miniDSP 2x4 units, one will have to toggle the USB connector from one unit to the other one to be able to flash the unit.

- Then down to the plug-in that is a must for any configuration. If you select the option of 2  miniDSP 2x4 configuration, then the 4way advanced is your best bet. In the case of the 2x8 or 8x8 configuration, the miniDSP 4x10 or 10x10 plug-in will be your choice.

Another typical question: Do you need more than 1 plug-in for a stereo configuration? The answer is no. A single plug-in can be installed on multiple miniDSP kit.

Hope this makes sense.

 For simplicity sake during the manufacturing process, all miniDSP boards are shipped with RCA connector pre-mounted. If you want to build your own box, with your own connectors, the easiest way is to use the correct pins of the expansion connector. These pins will feed the exact same signal to the DSP and allow you to connect your RCA jacks to the board. Another option would be to solder your wires directly on the connector pads. There is a lot of space at the back of the connector and if you're a bit electronic inclined, that's not a hard task to perform. Just make sure you don't create a short else you could indeed damage the board.

No. MiniDSP is using the USB protocol for control only. No audio can be streamed to the miniDSP Kit. Please refer to miniDSP kit datasheet for more information.

Yes it does. miniDSP kit come in many versions:

• An un-balanced kit version, for connection with an RCA connector
• A Balanced kit version, for connection with a balanced or unbalanced inputs/outputs
• Or even a Balanced + Unbalanced unit such as the miniDSP 2x8 and 8x8 units.

With the miniDSP balanced kit or balanced 2x4 being used in a wide range of applications (balanced/unbalanced connectivity), a clarification is in order on the expected levels depending on the jumper configuration and type of signal. The following application note will assume that you understand what is a balanced/unbalanced signal and understand the difference. If you don't, please take the time to have a look online.

miniDSP balanced kit/2x4 Input/Output circuity

- An operational amplifier turns a balanced signal to a unbalanced signal for processing by the DSP.

- To allow higher input level before saturation, an input padding network attenuates the signal based on the jumper position. The two jumper positions (0.9/2Vrms) refer to the maximum value of a balanced input signal before reaching 0dBFS. Note however that these values are for a balanced signal (it's a balanced board after all.. :-) so please see more about unbalanced inputs below.

- DSP does its processing on the signal in digital domain and outputs an unbalanced signal from the DAC. It's the task of an operational amplifier to turn this unbalanced signal to a balanced signal along with amplifying the signal up to a stronger level (up to 2Vrms).

Hoping we didn't loose you so far in this electronic mumbo-jumbo... In case we did, don't worry about it, all the information you need to know is summarized in the following tables.

Balanced input signal

Note that the miniDSP Balanced 2x4 and Balanced Kit are by default configured for the 2Vrms setting.

Unbalanced input signal

Now when it comes to feeding an unbalanced signal to any balanced input, only one leg of the operational amplifier is driven, i.e. half of the typical of a balanced signal configuration. The 0.9V/2Vrms settings therefore have a different meaning:

• In the 0.9Vrms, the board can actually accept up to 1.8Vrms with an unbalanced signal before reaching saturation (0dBFs).
• In the 2Vrms position, the board can accept up to 4Vrms unbalanced before reaching saturation (0dBFS). 

The below table summarizes it all for you:

Hoping this summarizes it all, feel free to contact us if you have any questions.

One of the strong points of our miniDSP kits is their flexibility. Our products weren't designed for a fixed application, but rather with the ability to give DIYers as many options as we could. For those of you wishing to tweak/modify their miniDSP kit with a custom add on cards for balanced input/ouputs, two options are available:
- OPTION1: You could use a balanced in/out stage and connect them to the unbalanced input. It may not improve the SNR/DR but will at least allow you to interface to our kit at very low cost. (Cost of a transformer - Google balanced to unbalanced for more info on how to)
- OPTION2: You could also use an external ADC/DAC balanced in/out by interfacing them to the I2S bus on the expansion port. There aren't any guideline or template guides on which IC to use, but balanced in/out stages +ADC/DAC are fairly common information on the web. For more information about our I2S connectivity to our board, have a read at our I2S technical note first(http://www.minidsp.com/support/downloads ) to correctly understand how to interface with the board.

In an effort to keep everybody happy, our miniDSP kit comes in two versions (A&B):

• Version A is meant for consumer line level signal (-10dBV) such as a PC source, Portable media players or any other sources with an output level < 0.9Vrms. The intend of this version is to have high input sensitivity for such low driving sources, hence driving the miniDSP close to the input level ADC like to be (-15dB below headroom).

• Version B is meant for CD player line level (2Vrms), preamplifiers or any other source with higher output driving capabilities. If you intend to insert your miniDSP kit in line with such source, you better use this version otherwise overdriving of the input converters will result in saturation, aka poor sound quality.

 Yes, you can indeed bypass the ADC and DAC by using I2S digital audio signals instead. All required signals are available on the expansion connectors of the miniDSP kit. Have a look at our I2S technical note for more details.

No, only a single plug-in can run at a time. You can however easily upgrade from one to another with a simple click of a mouse.

Unfortunately no. The tools we use to build firmware versions are proprietary licensed software which make it impossible for us to release an IDE. MiniDSP are not meant to be DSP & firmware engineers development tools. Our ready-made plug-in concept is to simplify DSP programming, not make it as complex as any other manufacturer dev tools.

The answer goes as follow: miniDSP was indeed sucessfully implemented in some cars. However, note that car environment is one of the harshest environment to deal with and miniDSP kit unbalanced was not designed for the purpose (e..g balanced input to reject noise, REM..etc) Lack of grounding, noise, dirty power, all the bad elements are in a car. If this is your first install, best to read more about it.

Unfortunately no. Because of the way we control the boards (HID driver), having two software instances running concurrently would require a unique identifier. It could however maybe happen one day if we get enough requests to justify redesigning the software/firmware and once we understand better how our customers are using our kits.

 No. MiniDSP kits along with their plug-ins are a complete plug&play solution designed and engineered for an easy implementation. No need to know coding, no need to understand biquads or building windows plug&play drivers, we already did the programming job for you. Simply enjoy tuning your audio system!

Unfortunately no. MiniDSP kits and plug-ins work side by side to run algorithms on the DSP IC. Plug-ins are actually only the control part of the firmware that is to be loaded on the IC. While they can run offline to test one of your configuration, they won't process any audio on their own.

That's it, you made the big step and purchased your miniDSP kit along with a plug-in. You can check the leadtime before shipment from the webstore product page. It typically ships around 3 to 5 working days (i.e. NOT including week ends and remember that we're based in HK time zone).  Also note that we will enable the plug-in download once your order shipped, combined with the information on how to track your shipment. A single email, a single process simplifying our task. Thanks for your patience!

miniDSP Ltd warrants all our products to be free from defects in materials and workmanship for a period of one year from the invoice date. Our warranty terms do not cover failure of the product due to an external trigger, misuse, servicing or usage out of the recommended installation instructions.

With this said, make sure to carefully read the user manual when customizing the board. Any board damage investigated as failure to follow proper installation may void the warranty. So if in doubt, make sure to ask  before any modifications.

In the event of a board failure, here are the steps to be followed:

• Contact our technical staff using the contact us form and clearly describe the symptoms.
• Before returning any goods back to our Hong Kong office, products first need to receive a return authorization (RA) number. Doing otherwise,  miniDSP Ltd deserves the right to not process your return.
• Once the RA# issued, customers are required to ship goods back to us at the customer's expenses. miniDSP Ltd will return the repaired/replaced product back to you at our cost.
• For any doubts, questions or comments with regards to our warranty terms and conditions, please contact us.

 miniDSP store uses the Paypal gateway as an efficient and secure method to process all payments. With this said, we can still process any international visa cards.

If you want to use your visa card instead of a Paypal account, simply click on "Pay with a credit card" at the last check out step of the order. See below image an example paypal screenshot.

Yes, you definetely can combine multiple items to an order since our webstore is setup such that no additional surchage applies.

miniDSP uses Speedpost Hong Kong for all our shipments leaving from our Hong Kong warehouse. An order typically takes 3-4 "working days" (i.e. not including week ends or holidays) to be prepared and packed by our staff. Once ready to ship, a tracking number will be emailed to you along with the downloads instruction for the plug-in.

Typical shipping leadtimes are 3-4 working days not including Custom clearance in your country. All parcels will be provided with a tracking number for easy online tracking. To the question "Why aren't we using Airmail?", our answer is simple: "Too much troubles for both seller/buyer". Lost parcel, long leadtimes, anxious customers are the typical expectations of airmail shipping. With Speedpost, our team is proud to say that not a single parcel has been lost!

Custom/Import Duties: Customer shall bear all country, provincial, government, state and local sales, use, goods and services, value added, privilege and similar levies/taxes. Please contact your local custom office to learn of your country's import fee structure.