opaqueice wrote:
> The simplest method is to record the S/PDIF output of the SB using
> (say) a computer soundcard, and then compare the resulting WAV file
> with the FLAC file decoded to WAV on the server, or with the original
> WAV file the FLAC was encoded from. 

This is actually not a very simple thing to implement properly.

First, you need to record on a better ADC than you are using to feed it, 
or you can't tell which is causing any errors. With good cables 
(Grinning, ducking and running)

Second, it is unlikely that you will have the gain on the input ADC 
exactly match the gain of the sampled DAC. If the gain is off by a small 
fraction, you will find that every byte is different, so it will appear 
to be 100% different.

It is trivial to verify that FLAC will decompress a file to exactly the 
same data as the source.

If I was going to prove that the bits -> DAC to ADC to bits chain was 
identical, I'd probably use a variation of a zero knowledge proof.
Make up a sequence of pure tones, say a second long at 20hz, 30hz, 40hz, 
60hz, etc. play it and record it, run each section through an FFT and 
make sure that the frequencies are exact, and that the amplitudes are 
"highly correlated". Once that works, try some simple mixed samples, say 
a 40 hz + 300 hZ + 700 hz

But this is getting pretty serious, and most reviews are not close to 
this serious.

I'm also not sure I agree with a comment up thread that if it wasn't 
right, it would sound like random noise. I could easily imagine a 
hardware or software bug that caused all frequencies to be, say 1% low, 
so A 440 would be 435 hz. I would bet that most folks won't not even 
hear that it was off.
_______________________________________________
audiophiles mailing list
[email protected]
http://lists.slimdevices.com/lists/listinfo/audiophiles

Reply via email to