Alright guys, I finally got around to taking some measurements of the B&W system in my 2020 M5 Comp.
Just for some background, I am a longtime car audio sound quality competitor/enthusiast. I have many years of experience with acoustic measurements inside of cars, and with tuning car audio systems to maximize performance.
For the measurements I used a 5 microphone array and SysTune Pro software. The reason for 5 microphones is because the interior of a car is a highly reflective acoustic environment. Reflected sound can cause serious comb filtering anomalies when using a single microphone. Instead, we want to collect a "spatial average" where measurements are taken from a few points inside the car and then averaged together to remove any single-location cancellations that might occur. In other words, inside of a car you might get one reading with a mic in one location, but you move the mic 12 inches left or right and you might get a different reading.
Unfortunately, the B&W audio system does not offer an AUX input. Without an external source input for the system I could only collect magnitude and frequency data. If I could have run my measurement signal from my computer into the B&W system, then I could have also collected timing data which gives us a lot more information on things like delays, phase and speaker polarities.
I set up the microphones around the driver's head location in the car. 5 mics surrounding the headrest area. I was treating this like a tuning session for a single seat location tune which is the more common approach to high-end car audio. When we are talking about the top sound quality cars in the country, most people have set them up to sound best from a single seat location because this is easier and generally produces better results. With a single-seat tune we can use DSP to correct for imperfect speaker install locations by using time delays on some speakers. We delay the speakers that are closer to the listener than the other speakers so that all of the sound arrives at the listener at the same time, despite them sitting to one corner of the car. I have assembled a microphone stand that holds the microphones around the area that the driver's head would occupy while seated in the car. By doing this I am able to get 5 microphone locations recorded simultaneously and the SysTune software will average them together and provide the results.
Since I could not run an audio signal into the car from my computer, I had to rely on a standard pink noise audio file. In this case is was an uncorrelated pink noise from 20 Hz to 20kHz.
You can see the result in the attached image.
There is a general downward slope from the bass region to the upper frequencies. This is actually a good thing because cars are very different than homes or concert halls. In a car, we need for the bass to be significantly louder than the higher frequencies due to the small size of the listening space and the highly reflective nature of the surfaces. A perfectly flat acoustic spectrum inside of a car will not sound good. The bass will sound thin, and the highs will sound overly bright and harsh. So a downward slope like we see here is quite common.
In this case, there is about a 10 decibel reduction in output from 30 Hz to 1000 Hz, and then another 5 dB reduction from 1000 Hz to 16 kHz. This is very similar to how many sound quality competitors will tune their cars, at least as far as tonality is concerned.
There is a sharp dip in the response at about 140 Hz. I attempted to correct this dip with some EQ boost, but the EQ bands offered were 100 and 200. So I tried boosting 100 Hz in the B&W EQ. See the chart with the blue line which represents the measurement of the car afrer boosting 100Hz. Notice how the overall level increased but the sharp dip at 140 Hz is still there? It turns out that this dip in response is being caused by a cancellation. That means that the speakers in the car are fighting with each other at the driver seat in the 140 Hz range, and there is an acoustic cancellation happening in that location. This causes the response to dip and unfortunately this cannot be corrected with EQ. The way to fix this would be to adjust the speaker delays or crossover points. Those are both options we do not have available to us with the B&W. A modern, aftermarket DSP would allow us to correct those things and we could probably fix this issue. But there is no way to fix it with the B&W system.
I attached a picture of the results when I added some boost to the 100 Hz EQ in the B&W DSP. You will see that the overall level in that region was boosted, but the sharp dip in response at 140 Hz still remains. I do not recommend boosting this EQ band because we don't get any benefit and the EQ boost will actually drive the speakers into distortion sooner. Think of it like an imaginary car with a top speed limit set to 150 mph. Imagine that you could drive the engine hard enough to go 200 mph, but the electronic limit is set at 150 regardless. So in this case you would be pushing the engine harder than the car could possibly go. You try to go faster but it just isn't possible. When we have a cancellation like this it's the same thing. No amount of EQ boost is going to fill in that dip, but every click of EQ boost will cause distortion to go up. So we don't want to use it here. There is no benefit, and we might actually make things sound worse.
Higher up in the 300 Hz to 4000 Hz range there is some peakiness in places. For example, there are peaks at 1100 Hz, and 3000 Hz. With a good aftermarket DSP I could use an EQ to bring these peaks down, but I don't think the B&W EQ has enough precision to correct these.
None of the above information touches on the other aspects of sound quality like imaging and soundstage. Since I could not run my computer audio into the car, I was not able to get timing data to analyze speaker arrival times.
When I have more time, I will post some of my listening evaluation of the imaging and soundstage. Spoiler-alert, but these aspects are quite poor in the B&W system from my listening thus far. But I would like to do a formal and objective evaluation of that in the future.
Anyway, that's all I feel like writing about it at the moment.
Cliffs Notes:
1. The overall tonal response is mostly smooth, with a 15 decibel gradual slope from the bass region to the treble. That kind of slope to the response is considered desirable in the car audio world.
2. There is a cancellation at 140 Hz which is being caused by the speaker locations and the car's geometry. This null cannot be fixed with EQ, so boosting EQ in the 100 Hz section of the B&W EQ is not recommended as a solution.
3. Upper frequencies are generally well-balanced with only minor peaks in a few places, like 1100 Hz and 3000 Hz.