7 Reasons Not To Check Your Mixes In The Car!

We’re often advised to check mixes on a car stereo. But how useful is the fabled ‘car test’ in reality?

The idea of checking your mixes on your car’s stereo system must be one of the most often‑repeated music‑production tips. But have you ever stopped to think why it’s suggested — or even whether it’s genuinely useful? Although lots of music is consumed in cars (and looks set to be for years to come), and the change of listening environment might helpfully put you in a different frame of mind, cars don’t generally offer a great listening environment for all sorts of reasons. In fact, I’ll go as far as to say that you cannot really trust what you hear, particularly in modern cars. And what’s the point in doing a listening test if you don’t know how to interpret the results?

In this article, I’ll explain what you need to know about in‑car audio so that you have an idea of if/how to adapt your audio for in‑car listening, and what you should keep in mind if using car audio systems for mix checks.

Nonlinear In-car Audio Systems

The first thing to note is that audio systems in cars are far from linear: almost all have their bass frequencies boosted. This is partly due to unavoidable physics and partly by design. Small mixing rooms have room resonances (‘modes’), where reflected signals make it hard to judge bass notes because they either sum to make a frequency massively over‑pronounced or cancel out to a greater or lesser degree. Obviously, cars are smaller ‘rooms’ than even (most) home studios, so have significant resonances from the deepest bass notes right up to around 350‑400 Hz. In the extremes, there may be up to ±12dB of variation from the average level at a given frequency.

Unlike studio monitors, which are typically designed for a flat frequency response, the overall frequency response of the typical car audio system is deliberately skewed, losing high end by about 1.5dB/octave in many cases (Figure 1). The goal is to be easy on the ears in prolonged driving sessions, and the tilted curve is achieved by acoustic engineers through a combination of the modal bass boost of the car, the contributions of the midrange and tweeter speakers, and the ‘correction’ of their outputs using DSP filtering. But not all troughs and peaks in the frequency response can be compensated for; DSP can do a lot but it cannot fix everything.

Figure 1. The frequency response of a 2022 Audi A6 (in black). The light blue line is the average, while green is the ±3dB border. The HF response falls by 1.5dB/octave. Note the troughs too, especially at 180Hz.

When driving, the noise produced by the car through its engine, exhaust, tyres and flowing wind noise is also significant, and it doesn’t contain much high‑frequency content. So the perceived balance when travelling is different from that when the car is stationary, and the same can be said when stationary of when an engine is idling or the engine completely switched off. If your mix sounds bloated, bass heavy and even a bit dull in the car, then, especially if that’s the case when the car is stationary, this could well be due to the frequency response of the car. It’s very important that you consider this before you rush to reorganise your mix’s low end after listening in your car!

Figure 2. The purple line is an average of 10 tested in‑car audio systems. The average is not linear, tilted, and bass boosted. Green is for comparison -- a very linear Neumann KH310 studio monitor. Dark blue is driving noise at 100kph of a 2022 Audi A6; the noise inside the car is mainly in the low frequencies.

Is Your Car’s Amp Lying?

In the last decade or two, cars have become rolling computers. Customer demand and even more so legal requirements make the integration of equipment like auto emergency braking, modems for emergency calls, speed limitation and energy consumption reporting mandatory. And, as with every other aspect of motoring, computers now play a key role in managing the in‑car audio experience, with DSP systems similar to those found in digital audio equipment such as mixing consoles and audio interfaces.

Why is so much calculation power needed? Largely, it’s used to integrate the speakers with each other in the ‘cabin’ through delay, equalisation, crossover filters, gain and routing. Also, the high number of additional audio applications such as phone‑call integration, engine sound generation (in electric cars), and both sat‑nav and warning‑sound playback must be served. There are also some complex audio functions which affect audio playback directly: loudness alignment, road‑noise adaptation and air conditioning fan compensation, for example. These all either manipulate the overall volume of the audio system or boost/cut energy in a certain frequency band.

What can we take away from this for our listening tests? Well, if your mix sounds thin, variable or over‑compressed in the car, depending on the driving situation, then this could well be because of adaptive processing by the entertainment system’s DSP.

Figure 3. DSP Concepts’ AudioWeaver is a visual programming environment for DSPs. The user interface is similar to some DAWs, with virtual wires being plugged into modules and virtual speakers.

Every Car Sounds Different

Depending on the market segment and body style of the car, sound systems are designed differently — they have to cater for the circumstances they are confronted with. Also, every car maker also has its own ‘sound philosophy’, leading to very different choices in DSP processing techniques, user options and speaker selection and positioning. So there is no one standard design.

In some cabins, space is at a premium, and this leads not only to small speakers but sometimes even the omission of essential (sub)woofers. Cost is also a factor and good speakers are expensive. At one end of the price spectrum, high‑end sound systems may have up to 39 speakers (eg. the current Mercedes Benz S‑Class), while lower‑end models (commercial vehicles and cheap cars) will have to be content with four or even two speakers.

Figure 4. Speakers in every corner: the 16‑channel Bang & Olufsen audio system used in an Audi e‑tron.Photo: Audi AG

So if your mix sounds different in two cars, it is almost certainly because the cars’ sound systems are different. If you aren’t already accustomed to the sound of a particular system, then, you have to take what you perceive it to be ‘telling you’ with a pinch of salt.

User Settings

How often have you changed settings on your phone or computer without really understanding them in the first place, and were later surprised by unexpected behaviour? Similarly, every car sound system has options which can be set by the user. A simple shelving filter and balance/fader controls are standard, but per‑seat optimisation and feature retrieval for 3D speakers and surround sound can also be found at the higher end of the market. (Volvo even offer the user the ability to add reverb from the Gothenburg Concert Hall to the playback!) If someone has tweaked these options to personal taste (or just because they were fiddling with the controls!), this could be a cause of poor sound quality. So if your mix sounds weird in the car, you should always ask yourself whether the sound system settings might be causing the problem. You can try listening to some other material, of course, to see if you hear similar problems. But if in doubt stick with the factory settings.

Cars Are Loud!

Although in recent years the noise pollution of engines and exhausts has been massively reduced, background noise remains a real issue for in‑car audio. While this could help you notice sounds which ‘jump out’ above the average level, if your mix contains lots of layers and dynamic passages these can easily get lost in the noise floor. In other words, because details can be masked, you need to be very careful about judging levels and dynamic range.

Electric cars are heavier, which means there’s greater displacement of the rubber, and the tyre acts like a toroidal speaker membrane.

By the way, one might imagine that this isn’t such an issue for electric cars since their engines are not so noisy. But actually the average electric car is no quieter than its combustion‑engine counterpart because the biggest source of noise in most cars, under normal driving conditions, is the tyres. Electric cars are heavier, which means there’s greater displacement of the rubber, and the tyre acts like a toroidal speaker membrane.

Seat Positioning: Every Little Counts

Figure 5. Individualisation is an important theme in modern cars, with the seat position lighting, climate — and of course the sound — being adjustable to each occupant’s taste.Photo: Mercedes‑Benz AGThe above‑mentioned characteristics of small‑room acoustics also affect the quality in the mid and high frequencies, and lead to fluctuations in every seat. Even the slightest movement of your head can make a difference to the frequency response at your ears.

Some auto‑makers optimise the sound performance for the driver’s seat, and leave the other seats out of their optimisation efforts. For sports cars, for example, it makes no sense to think about sound in the back seats, since they’re rarely (more probably never) used. For executive sedans/saloons, on the other hand, the sound in the back seat is much more likely to be important, and will therefore probably be accounted for in the development phase. And as I mentioned above, some cars’ sound systems feature DSP options for the user to tweak the sound for specific seat positions.

So if your mix shows problems only on a certain seat (like the back seats or front passenger seat), then this can be due to the seat‑to‑seat variations. Once again, then, before you rush to make judgements about your mix, make sure you know what it is you’re hearing!

Rub, Buzz & Distortion

Trim pieces in cars are, in many cases, fixed with clips to the chassis. It’s not a cost‑cutting measure but good engineering: conventional screws don’t hold up well because vibrations loosen them over time; that’s not a problem with clips. A downside of the clips, though, is that they can vibrate in their holders, and this can lead to audible rub and buzz sounds.

If your music excites the resonance frequency of a trim piece like the dashboard or a metal piece, then, you’ll hear the vibrations as harmonic distortion. It’s often not that elegant‑sounding either, because their harmonic structure and phase relations are not generally pleasing to the ear! So, if you’re performing the car mix test and you hear distortion and unpleasant buzzing noises, then this could well be due to trim pieces which rattle or have resonance issues.

Conclusion?

So, with all this in mind, is it really a good idea to check your mix in the car? I’m afraid I must answer in the same way my lawyer always answers my questions: it depends! I would not advise anyone to use the car as a must‑not‑be‑questioned, absolute reference, but as a place for fun listening and enjoyment. Some audio‑quality evaluations, though, can be performed with sufficient experience and critical listening. If you’ve spent time in your car and know it well, having listened to lots of familiar material in it, then you’ll know the frequency limitations, noise, functions, and goals of the audio system intimately. In that case, then yes, it can be a good idea to use the car’s sound system as a secondary reference. (Really, you should be able to rely on an accurate monitoring system in your studio, and the main benefit of the car is that it can put you in a different frame of mind, which helps you to notice when something might warrant closer attention back in the studio.)

If, on the other hand, you have just hopped into the car of a friend for a quick listen and think you might want to change your mix because you feel it sounds muffled overall and the background vocals sound piercing on the back seat, I’d suggest that it’s not a good idea — you can’t rely on that car check as a reference, because it would be misleading.