Recording up to 192kHz?

The maximum theoretical audio frequency is half the sampling  frequency. Nyquist–Shannon sampling theorem - Wikipedia, the free encyclopedia

Thus, if the hardware were perfect, the max audio frequency you could record would be 96 kHz.

But, as we all know, hardware is never perfect.

I did a quick test by feeding white noise at a 192 kHz sample rate into the Line In of the Wolfson card and recorded it.
Using Audacity's "Plot Spectrum" I got a chart showing the audio frequency response of what was recorded:

I actually did not think it would go so high.

I have seen (and heard) several high-end Hi-Fi systems that roll off sharply when they get above 20 kHz (yes, I know that it can be done on purpose.)

What are the frequencies you're interested in?

--

Ragnar

Hi Ragnar,

Thanks for your very detailed reply. That was really helpful.

I'm interest in recording bats which emit calls from around 12-160 kHz. If I could record calls up to 96kHz I would be happy. Do you have any experience using different ultra sonic mics with the Wolfson card.

Alas, I have no experience with such microphones. My interest in high-frequency audio is in analyzing the output from FM stereo transmitters. I'm happy if I can record up to 77 kHz.

The card starts to struggle above 94 kHz, the sensitivity curve wobbles up and down quite bit up there. Anyway, up to 94 kHz should be doable if the rest of your gear is up to it. If you record at 24 bit depth, you should be able to compensate in software for the reduced sensitivity at the higher frequencies.

--

Ragnar

Thanks!

Ragnar,

I am looking to record or at least transmit over Jack an MPX signal from an FM transmitter.  My application will be the analysis of RDS signals living at 56kHz.

Have you had any success in doing this with a Pi?

Hi Aaron!

This is from a WAV file I recorded on the Pi. The curve leans a bit to the right, but all in all it works better than I anticipated.

I haven't tried to push 192 kHz audio through the network. I have done 48kHz between two Pis with Jack, though.

--

Ragnar

Ragnar, that looks really good.  My main concern was that there might some issue with a LPF being fitted or indeed the -3dB point being far to low to capture useful data at 57kHz.

I'm kind of interested as to what the peak around 65kHz is though?  Is it some kind of SCA carrier?

That 65 kHz peak is why I'm looking at that transmitter in the first place. It shouldn't be there.

It has other troubles too...

--

Ragnar

Forgot to mention, applying the Wolfson's built-in high-pass filter to block DC offset, as discussed here, does wonders.

--

Ragnar

I'd spotted that, makes for interesting reading.

I'll hopefully get my hands on one of these and do some testing (my applicable is mainly to get an MPX signal to my control room from the TX to use with RDS Spy).

Good luck with your TX.  It does look a bit poorly.