‘How To’ investigate cabin noise using an iPad

[classcchap]

This ‘How To’ is about using SpectrumViewPlus on an Apple iPad to investigate noise.

It’s prompted by a thread in ‘Technical’ about TCN (Tyre Cavity Noise). Rather than give a full-on instructional it would be better, I thought, to just introduce app and see where it goes, so ..

Here are a couple of examples from that thread showing the TCN in my W205 C Class:

The first is a Fast Fourier Transform (FFT) ‘snapshot’ of the TCN - shocking isn’t it?

The second is a Spectrogram showing the full half-minute or so test from which the first was taken.

This is my ‘test’ setup I use:

The iPad is secured to a couple of foam blocks on the front seat with the microphone forward and unobstructed - the TCN comes into the cabin through the footwell.

If you’re interested I suggest you download the free version of SpectrumView, have a play, then get back and we can have a chat. You will need the full SpectrumViewPlus to do proper testing - I think I paid £8.

Finally here are the 2 record settings I use:

None of the other settings really matter.

 

[Petrol Pete]

How about " WHY investigate cabin noise with an I-pad ?" If you don't like how your car sounds sell it and get one you like the sound of. FFS

 

[Dippo]

Your chart says the noise level is 18 dB. According to this chart, that's just above the level of calm breathing or leaves rustling (15 dB).

What is Noise - Noise Levels - noiselimiters.co.uk

 

[Mactech]

It's 18dB above the other noises in the car. Check the first few lines of your link!

 

[Deleted96908]

This ‘How To’ is about using SpectrumViewPlus on an Apple iPad to investigate noise.

It’s prompted by a thread in ‘Technical’ about TCN (Tyre Cavity Noise). Rather than give a full-on instructional it would be better, I thought, to just introduce app and see where it goes, so ..

Here are a couple of examples from that thread showing the TCN in my W205 C Class:

The first is a Fast Fourier Transform (FFT) ‘snapshot’ of the TCN - shocking isn’t it?

The second is a Spectrogram showing the full half-minute or so test from which the first was taken.

This is my ‘test’ setup I use:

The iPad is secured to a couple of foam blocks on the front seat with the microphone forward and unobstructed - the TCN comes into the cabin through the footwell.

If you’re interested I suggest you download the free version of SpectrumView, have a play, then get back and we can have a chat. You will need the full SpectrumViewPlus to do proper testing - I think I paid £8.

Finally here are the 2 record settings I use:

None of the other settings really matter.

I'm into pro sound as part of my job/business - so can never get too much of stuff like this.

Thinking aloud. Plug in a directional mic to your iPad, a shotgun or similar, and point it at the various surfaces in the car. You might find some sympathetic reflectors bouncing the high frequencies around the car - most noticeable I should imagine from the surfaces close to where you sit. Not quite sure what you'd do about it, other than maybe sticking some leopard print all over the place... Rude Merc anyone?

Also thinking you could record a typical sample, invert it, and pair your phone/iPad with something like this:

Bose Soundlink II

I reckon you'd hear quite an improvement.

Thanks for posting.

 

[classcchap]

Thinking aloud. Plug in a directional mic to your iPad, a shotgun or similar, and point it at the various surfaces in the car. You might find some sympathetic reflectors bouncing the high frequencies around the car - most noticeable I should imagine from the surfaces close to where you sit.

Thanks for posting.

Good thinking, my purpose here though is to keep it simple - it’s a very powerful tool just using internal mic! But by all means explore - here for instance is a white noise sweep I did early on to check for cabin resonances:

All good fun - and thanks for your thoughts.

 

[grober]

Since most TCN seems to be confined to the 190-250 Hz band is it possible to configure the detection/display to show this frenqency band in greater detail?
Another interesting reference cited the following?
The acoustic cavity inside the tire-wheel assembly is known to contribute to audible noise in the passenger compartment of some vehicles. Traditionally, the source of this noise issue has been viewed as only response from the acoustic resonance of the air cavity inside the tire-wheel assembly. This paper demonstrates that coupling between the cavity resonance and a structural resonance of the wheel can amplify the response and result in higher noise levels in the passenger compartment. Laboratory measurements show that vehicle interior noise generated when coupling occurs between the cavity resonance and a structural resonance of the wheel is significantly higher than when the coupling is eliminated. This phenomenon is demonstrated by comparing vehicle noise levels which result from two different wheel designs: one that has a structural resonance very near the acoustic resonance and one that does not.
Unfortunately it doesn't specify what the difference in cabin sound levels is.
Raises the prospect of an alloy wheel with a structural resonant frequency which re-enforces any tyre fitted's TCN?

 

[sturgeon]

Is it worth trying tyre pressures slightly above and below the recommended, to change the shape of the cavity? There might be a sweet spot.

Recording with the AGC off and at a higher sample rate would be interesting.

 

[clk320x]

How about " WHY investigate cabin noise with an I-pad ?" If you don't like how your car sounds sell it and get one you like the sound of. FFS

WOW

 

[classcchap]

Since most TCN seems to be confined to the 190-250 Hz band is it possible to configure the detection/display to show this frenqency band in greater detail?

Unfortunately no, not with SpectrumView. This is actually why I specify the lowest sample rate available: 8000Hz. But it’s a good point.

 

[classcchap]

Is it worth trying tyre pressures slightly above and below the recommended, to change the shape of the cavity? There might be a sweet spot.

It’s certainly worth a try, but in theory no, it’s only affected by temperature, but then, my tyre pressures are same all round, one test I saw 2 peaks - when I reset pressures it went back to 1 peak. So yes, I think a closer look is indicated - perhaps Benz might look into it?

Recording with the AGC off and at a higher sample rate would be interesting.

.

Yes, and I’d encourage anyone to experiment - give it a try. I found AGC ON was better because it saves having to boost it at display time - I found that tricky and error prone. In the end, of course, we need a consistent procedure if we mean to compare.

 

[MercedesDriver]

This is the recording I’ve made with the phone placed on a passenger seat and a mic faceing forward as advised and am heppy with the findings. All it has recorded is the bonnet distance diesel engine noise and a few pound coins rattling from my front ashtray. Really please as there is no TCN recorded.
Next step will be to record kids while playing as a proof of a reason why I became a grumpy old bugger .

 

[classcchap]

... is it possible to configure the detection/display to show this frenqency band in greater detail?

Ah..correction. I didn’t spot ‘display’ so, yes you can zoom in at display time - but it doesn’t improve resolution so in reality it doesn’t actually help. I know what you mean though.

 

[classcchap]

View attachment 74510 View attachment 74509

This is the recording I’ve made with the phone placed on a passenger seat and a mic faceing forward as advised and am heppy with the findings.

Whey... not so fast. I’d want to know more about that red line at the bottom. I’ll get back in the morning with display settings that will show the low-band

Top marks for giving it a go though - I like that.

 

[MercedesDriver]

Please don’t tell me if something is wrong with my Merc.

 

[grober]

OP----- HAVE YOU READ THESE THREADS?
W205/ S205 Wind Noise?

Acoustic Glass - First impressions - MBWorld.org Forums

 

[classcchap]

This is the recording I’ve made with the phone placed on a passenger seat and a mic faceing forward as advised .... Really please as there is no TCN recorded.

Your recording looks good - but now you need to set up the display, then fine adjust for FFT ‘Snapshot’ or ‘full Spectrogram.

So, this will set-up the display:

Then this sets up a few other switches:

Having set the display you can now set Analysis for Spectrogram:

You should see something generally like this:

All you need to do here is set the vertical scale so it’s 230 in the middle and 30 at the bottom, using normal iPad squeezes and such.

It’s easier to do than write about - hope it goes well, please let me know.

Next I’ll post the FFT settings which is much easier since you’ve no made the basic settings.
The I’ll cover some general points on doing the test run.

I do hope this is going to wrk.

 

[classcchap]

OP----- HAVE YOU READ THESE THREADS?

No, I’ve not read the threads but I do know of the 205 wind-noise
issues - can’t say I have that problem though. It’s tempting to think their hearing TCN but TCN really does sound nothing like wind. The vendor offered me ‘sound-proof’ windows but I knew enough to realise that would make things worse - so I declined.

I’m going to say here that the W205 is, in my view, the best buy in its class by far and I’m having great difficulty finding a replacement. Unfortunately it’s also hobbled by a very bad design fault - but I’m sure going to miss an otherwise excellent car and I look forward to it’s replacement.