<div>Hmmm.. I have no success until now :(</div><div><br></div>Using play -e raw -b 32 -c 1 -r 441000 --endian little dat.raw <br>I am facing with the following error:<br><br><div>play FAIL sox: --encoding: `raw' is not one of: signed-integer, unsigned-integer, floating-point, ms-adpcm, ima-adpcm, oki-adpcm, gsm-full-rate, u-law, mu-law, a-law.<br>
</div><div><br></div><div>That's strange. I am sure the data are in little-endian format (since the storage CPU is Intel), each sample is 32-bit long integer, I have 32768 samples but have no idea about the real sampling rate.</div>
<div>I can try with several speeds, but the question about the "encoding" still unclear for me. And this is the same using sox instead.</div><div><br><br>So, I am stuck here</div><br><br><div class="gmail_quote">
2010/12/11 Rob <span dir="ltr"><<a href="mailto:lau@kudla.org">lau@kudla.org</a>></span><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div class="im">On Saturday 11 December 2010 04:42, ciclo esano wrote:<br>
> Moreover the information about bit rate, channel and everything<br>
> concerning audio file is very difficult to apply here, since the raw<br>
> data are coming from a digitizer after several downsampling processes<br>
> to have all the signals in the audio range (in fact they arise from the<br>
> high frequency of nuclear spins, hundreds of MHz)<br>
<br>
</div>Oh, that sounds brilliant. Well, if they're being downsampled already,<br>
surely you know what frequency to choose, but if not, you could just start<br>
with the common ones (22050, 32000, 44100, 48000), see how they sound and<br>
tweak as needed. I'd assume it'd be one channel and 32-bit samples based<br>
on what you said, but you can experiment with that too.<br>
<font color="#888888"><br>
Rob<br>
</font></blockquote></div><br>