50
50
Guenter Villnow
50
50
Guenter Villnow
50
50
50
50

This app note is part of a series on car audio sound. The top level app note and list of all parts is:

Contents

Polarity and sensitivity check [Top]

After the installation of speakers, DSP, amps, cabling, your mics and analysis system, the first thing to do is check the sensitivity and polarity of the speakers.

To check the sensitivity, put a single microphone in the middle of the car. Measure every speaker with the same output (e.g., 2.83V) and compare them in pairs that make sense i.e., door woofer left vs door woofer right; tweeter left vs tweeter right, etc. The frequency responses and levels should be similar. To do this, measurement sweeps must be used. This way you can rule out problems in your installation like acoustical short circuits or faulty wiring.

Also check if the sensitivity of the speakers with each other (woofer vs midrange vs tweeter) matches and adapt accordingly.

To check for the right polarity, polarity testers can be used. Note that existing filters or problems in your playback chain (which may occur when you are using an OEM head unit to feed into your DSP) can cause problems with polarity (for more details, see https://www.audiofrog.com/how-to-test-for-polarity/).

I personally like to use a simple 1.5V battery to check if everything is alright. Prior to connecting the speaker wiring to the amplifier, connect (+) of the battery to (+) of the speaker line and (-) to ground of the speaker and hear if it is triggering as expected. Observe if the cone of the speaker goes in the right direction. Simple, but effective and there are few false negatives/positives from this testing method.

Speaker protection through output level and crossover [Top]

Limitation of output power

Speakers can break if used in the wrong frequency band and/or given too much power.

This happens either through mechanical or thermal stress. To be safe from the get-go and during the calibration phase (where overload mistakes can easily happen), speaker protection should be addressed as early as possible during the tuning.

Here is where the data sheet of your speakers come in handy. Check for information about the frequency band in which your speakers can be used. Measure the output voltage of the amplifier and calculate the power delivery to your speaker. An RMS multimeter can be used when your system is turned up to max and all other signal processing (EQ, crossover etc.) is turned off. Note that every speaker is different!

For example, you have connected a midrange speaker that is capable of handling 20 Watts to the Harmony DSP 8x12. Since the amp can drive up to 40 Watts, the output must be set to -6 dB FS to halve the output power.

If your DSP has a limiter or compressor (you can turn a compressor into a limiter when the ratio is turned up to 20:1 or higher), you can add a second line of defense for the speaker through voltage limiting.

Crossover

The crossovers are set up in the Output section tab in the C-DSP tuning tool through "Xover".

In the car it is helpful to have a look at the overlap band. The overlap band can be found through individual measurements of the speaker drivers with the same voltage (when they have the same impedance) and then compared in an overlay view. Shift the overlap band to a point where all speakers are in their comfort zone.

It is best practice to start with 24dB/oct Linkwitz-Riley (LR) crossovers. When both speakers are crossed, their cooperation is checked with both speakers active in the measurement. This example shows the result before applying time delay between the speakers:

For optimum performance, the integrity of the crossover must be checked again when timing is corrected.

Timing and phase correction with delays [Top]

In the car cabin every speaker is a different distance from the listener(s). Time delays must be applied to improve the overall impulse response, quality of the imaging and improve crossover efficiency.

The idea here is to find the speaker with the biggest physical distance to the listener and delay every other speaker accordingly to the same distance. The goal is to even out the arrival time of all sound events from every speaker, even though the speakers are in different locations. This will massively improve the overall impulse response of the system and imaging quality in general.

To find out how much speakers must be delayed, you can do this by measuring the distances from the speakers with tape measure, or you can use the data from your sensitivity measurements.

In this picture you can see how this would apply in a real-world scenario:

The delays are applied on the "Outputs 1-6" and "Outputs 7-12" tab. Note that you can select a time-based or distance-based settings.

A more in-depth tutorial regarding time alignment with REW and UMIKs can be found here: https://www.minidsp.com/applications/auto-eq-with-rew/speaker-time-alignment.

After delays and crossovers are done, the summation function between speaker pairs should be checked. Make sure that both speaker channels from the comparison pair have been calibrated in their crossover and delay settings are playing properly. When the crossover band is measuring flat, timing and crossover frequencies are fine. If the summation band dips or comb filters are visible, crossover frequencies and/or delays must be adjusted.

A very good explanation and tutorial on crossovers can be found here: https://www.audiofrog.com/crossovers-how-they-work-and-how-to-choose-them/.

Routing [Top]

Routing is probably one of the easier tasks in the tuning process. Refer to the miniDSP User Manual for your unit for information on how to use the routing matrix.

Routing is often realized in a trivial way: left signal to the left-hand side of the car and right signal to the right-hand side of the car. However, routing can be an important tool for creating proper soundstaging.

The drastic placement of the speakers in the car often creates extreme acoustic imaging. Try to distribute the signals from your source over several speakers. The miniDSP C-DSP 8x12, C-DSP 8x12 DL and Harmony DSP 8x12 offer that superb possibility. This creates a more immersive experience and lets the speakers work more like a "team" than a bunch of single speakers each given an individual signal.

In general, it can be said that the more input channels your source has (stereo: 2, 5.1: 6, Dolby Atmos: 12), the more discrete the routing to the speakers can be in order to achieve the artistically intended acoustic imaging.

Routing is setup in the "Routing" tab in the C-DSP tuning tool. With fader and the invert tool, back-fill signals for the rear doors, synthesized center signals and overlapping distributions can be created in the car.

Bass management and mono bass optimization with MSO [Top]

Although car cabins are relatively small in volume (between 2.5 and 4 m3), it has been proven in Praxis that using several sources for low frequencies delivers an overall better result. The load is distributed to several speakers, reducing position dependencies (less change when moving the head in a seat) and improving consistency across all seats. This concept is called "mono bass". Due to less mechanical stress on trim pieces and chassis parts in the vicinity of the subwoofers/woofers, rub/buzz and distortion can be reduced. The overall energy distribution is more even in the cabin.

The optimum result can be found either through lengthy trial and error testing or with algorithmic help. For computer-based help I recommend the tool Multi-Sub Optimizer (https://www.andyc.diy-audio-engineering.org/mso/html/).

An introduction on how to use MSO can be found here: https://www.minidsp.com/applications/subwoofer-tuning/dual-sub-with-mso.

In the C-DSP tuning tool the bass management is calibrated on the "Inputs & Bass Mgt" tab:

You will get the best results if you not only let MSO optimize the HPFs and LPFs, but also use one or two PEQ bands from the "Outputs 1-6" and "Outputs 7-12" tab for MSO to reach optimum results.

Equalization techniques [Top]

Overview

The EQs are setup in the "Outputs 1-6" and Outputs 7-12" tab:

Equalizers are used to even out spectral deficiencies. While the exact use and functions are explained better elsewhere (e.g. https://audiouniversityonline.com/parametric-eq-vs-graphic-eq/), I would like to give some remarks on some caveats and techniques in automotive applications.

Less is sometimes more...

... but sometimes very necessary: although we have a lot of DSP power it does not mean we have to use every EQ band or function.

When approaching filtering, it makes sense to start with the biggest peaks or dips and validate through listening first before adding another filter band. Evening out spectral deficiencies by maxing out all the EQ bands from the get-go and then listening is confusing and difficult even for experienced tuners, since it is hard to evaluate so many changes at the same time. Add filters one by one.

Narrow bands

Narrow bands are often not useful. While it is tempting to even out the smallest peaks and dips, filters with Q > 9 are often not helpful. The spectral resolution of our ear (which has an approximate resolution of a little more than 1/3rd of an octave) is limited and the spatial effect of a narrow band filter can turn negative or create a null when your head moves even slightly from the sweet spot.

Null points

There are some frequencies that simply cannot be EQ'ed: null points. They are in every car and can be maddening from time to time. At these frequencies it simply makes no difference if you boost or cut. Nothing, little or very little is happening in the frequency domain. It can be necessary to restructure the tuning of a channel around a certain null point to keep the desired inner spectral balance of a system intact.

Top-down or bottom-up

What is the best speaker group to start tuning with? You could either start with the tweeters and then go down to midrange, midbass and subwoofer (top down). Or you start with the subwoofer and then go up to midbass, midrange and tweeter (bottom-up).

I recommend starting with the bottom-up approach. This way it is the easier to integrate all the channels, since bass lays the foundation for music. The subwoofer and woofer deliver the most acoustic energy into the cabin and a lot of people use low frequency information to center their acoustical judgement.

Better negative than positive

When reading graphs, it seems to be a natural reflex to add something when there seems to be a lack of level in a certain frequency band. But it is better to think the other way around. Remove what is "too much" instead of adding what is not there.

Digital sources leave only very little headroom, which is why adding gain with EQs is not recommended. The digital path could be overloaded. It is definitely not forbidden, nor does it necessarily make for an instantaneous bad sound quality, but it should definitely be thought about twice. If you do use positive gain filters, it is likely that attenuation is needed in the input stage of the filter.

Allpass, maybe?

Allpasses are probably the most underrated filters and can be a lifesaver in automotive sound. Although they do not affect the spectrum, they affect the phase of the processed signal. This can be helpful on midrange and midbass channels to make staging and integration perfect, when delay in just a certain frequency band is needed.

Auto EQ

REW can make tuning and decision making easy through its EQ application. For a tutorial on how to use REW for automatic EQ, refer to: https://www.minidsp.com/applications/auto-eq-with-rew.

Gainstaging [Top]

During the calibration and tuning process, you should keep an eye on your gain structure and make sure that you are not clipping the DSP system or using unnecessary digital gain to boost analogue noise.

Note that not only boosts in EQ bands, but also cuts can increase the output level and therefore lead to distortion!

If you are not familiar with this issue, I recommend reading this app note: https://www.minidsp.com/applications/dsp-basics/gain-structure-101.

Closing [Top]

That concludes this part! Jump back to the top:


 

Related products - Car Audio DSP platforms