Tesla Model S Audio Quick FR Measurements

[gags11]

So I decided to do quick measurements of the audio in my 2020 Tesla Model S.

I have never been impressed with car audios. The best I had was in my late Panamera with Burmester sound. It could play very loud without distortion, but still, something was missing and it did seem a bit colored.

When I first listened to Model S audio, I thought it was not impressive and really bright. Thankfully when taking the treble down to at least -3, things changed.

Since then, I have been listening to this system, and have been very impressed. Songs that I know very well sound really good and right to me: timbre, voices, etc. Today, I did a few quick measurements, and it did confirm that a lot of work has gone to make the audio sound right, especially given all the reflections in the car.

I have placed the mic in my driver seat and measured from a few spots. FR changes a bit as aspected, but overall looks impressive. Distortions are good too. Again, the treble needs to come down, and some may prefer to dial down the bass, but not me.

 

[johnp98]

Awesome! I would be curious with a flat tone control what it looks like (more just curious what the vanilla setting is, and also how large of a Q the tone control uses).

Thanks for measuring though! As I have been quite interested, but don't have a Tesla yet.

 

[gags11]

Awesome! I would be curious with a flat tone control what it looks like (more just curious what the vanilla setting is, and also how large of a Q the tone control uses).

Thanks for measuring though! As I have been quite interested, but don't have a Tesla yet.

I will try to do a measurement with all flat tone controls tomorrow.

BTW: Model 3 and the new Model S systems supposed to be better, but I cannot attest to that.

 

[ppataki]

Thank you @gags11 , really cool!
Would you mind doing a sweep measurement at listening position please? That way we would not only have the FR and the distortion graphs but also the time domain related ones (like impulse response, step response, wavelet, etc)
Would be awesome to see those too
Many thanks

 

[Chrise36]

So I decided to do quick measurements of the audio in my 2020 Tesla Model S.

I have never been impressed with car audios. The best I had was in my late Panamera with Burmester sound. It could play very loud without distortion, but still, something was missing and it did seem a bit colored.

When I first listened to Model S audio, I thought it was not impressive and really bright. Thankfully when taking the treble down to at least -3, things changed.

Since then, I have been listening to this system, and have been very impressed. Songs that I know very well sound really good and right to me: timbre, voices, etc. Today, I did a few quick measurements, and it did confirm that a lot of work has gone to make the audio sound right, especially given all the reflections in the car.

I have placed the mic in my driver seat and measured from a few spots. FR changes a bit as aspected, but overall looks impressive. Distortions are good too. Again, the treble needs to come down, and some may prefer to dial down the bass, but not me.


View attachment 156472


View attachment 156473

View attachment 156474

View attachment 156475

View attachment 156476

Try to measure it with you sitting in the car it changed the FR in my car also take an MMM measurement see if there are any changes compared to the static measurement. Also i tried to compensate when driving with an open window in one of my DSP settings.

 

[thewas]

Would also recommend using the MMM technique in car interiors as this way you get a more exact and less rippled mid/highs FR which is also closer to what we perceive.

Funnily I had started FR measurements of my cars (using MMM) long before starting measuring my home systems, now more than 20 years ago.

 

[BenB]

I'd say your second graph (in purple) is pretty much what you want in a car. Under 200 Hz you have to fight the masking effect of background road noise, so you need that elevation. It looks like you might actually benefit from a bit of boost in the 80-200 Hz region (2-3 dB maybe), and a bit of a cut near 1 kHz. That way your average measurement would more closely resemble the purple measurement.

 

[gags11]

Thank you @gags11 , really cool!
Would you mind doing a sweep measurement at listening position please? That way we would not only have the FR and the distortion graphs but also the time domain related ones (like impulse response, step response, wavelet, etc)
Would be awesome to see those too
Many thanks

Those plots were sweeps. I have never really made sense of the impulse response or how to correctly plot them. But here is an impulse response from one of the measurements.

 

[gags11]

Would also recommend using the MMM technique in car interiors as this way you get a more exact and less rippled mid/highs FR which is also closer to what we perceive.

Funnily I had started FR measurements of my cars (using MMM) long before starting measuring my home systems, now more than 20 years ago.

I was not aware of the MMM technique. I watched a few videos and read about it. Makes a lot of sense.

Here is a measurement using the MMM averaging from the listening position. I’m sitting in the driver seat, back reclined and moving the mic at the position where my head would be normally.

Also, all the tone controls are dialed to flat, all at 0.

 

[ppataki]

Those plots were sweeps. I have never really made sense of the impulse response or how to correctly plot them. But here is an impulse response from one of the measurements.

View attachment 156581

If you share the mdat file I might be able to give you some hints

 

[gags11]

Looks like there is something happening around 1-1.5KHz region, possible resonances?

 

[gags11]

If you share the mdat file I might be able to give you some hints

I shared the mdat file with you

 

[Chrise36]

I would try to correct the range from 300hz and down the rest seems ok if you dont like bright sound what do you think as it is now?

 

[jtwrace]

I don't know anyone else on this forum that has more car audio experience than @hardisj

 

[gags11]

I would try to correct the range from 300hz and down the rest seems ok if you dont like bright sound what do you think as it is now?

I think the highs are a bit too much for me with all the tone controls dialed flat. I prefer at least -1.5-3db down for the treble. When I did I a single sweep measurement with controls all flat, the treble region seemed to tilt upwards. You do not see that with the MMM technique. Not sure why.

Edit: here is a single sweep from relative driver’s head position with tone controls flat and -3db down for the treble.

 

[Chrise36]

I think the highs are a bit too much for me with all the tone controls dialed flat. I prefer at least -1.5-3db down for the treble. When I did I a single sweep measurement with controls all flat, the treble region seemed to tilt upwards. You do not see that with the MMM technique. Not sure why.

Edit: here is a single sweep from relative driver’s head position with tone controls flat and -3db down for the treble.

View attachment 156599


View attachment 156600

Try to do more measurements with the MMM to gain confidence with it. You can do left channel first eq to taste then follow with the right it seems you need a bit of eq in the 1+ khz range and the bass region if it is too bright low the treble a couple of db.

 

[gags11]

Try to do more measurements with the MMM to gain confidence with it. You can do left channel first eq to taste then follow with the right it seems you need a bit of eq in the 1+ khz range and the bass region if it is too bright low the treble a couple of db.

I will do more measurements. It seems using white noise instead of pink noise may be more representative. The above MMM measurement was done using pink noise, which does not really show how bright the system sounds when the tone is set to flat.

 

[AudioJester]

Dummy question here, but how are you getting REW to output to the car? Is it a simple USB connection?

 

[gags11]

Dummy question here, but how are you getting REW to output to the car? Is it a simple USB connection?

MacBook Pro is sending audio output via Bluetooth to the Tesla

 

[gags11]

This has been a learning experience. I finally did the MMM technique with white noise around the driver head position, which matches nicely to the sweep measurements, and show the treble boost when tone controls were flat.

1. Tone controls flat
2. Treble adjusted / cut (see pic below)
3. Treble adjusted and immersive audio high (my preferred setting)