On Mon, Apr 27, 2009 at 07:12:32PM +0200, Lennart Poettering wrote: > > > Please try to reproduce this, then do "ls" in pacmd and paste it > > > somehere. This should show if PA does digital amplification for any of > > > the active streams. > > > > It's always reproducable, here is the output: > > > > http://caiaq.de/download/tmp/pacmd-distortion.txt > > > > Let me know if I can provide more details. > > Hmm, that output suggests everything is in order. Both the sink and > the stream are fixed to 0dB so PA shouldn't touch the data at all, > just pass it through to the device. > > Hmm, what makes me wonder though is the 's24be' sample format. There > might be something wrong with the conversion code for that. Yep, that was my first guess, too. > s24be is 24 bit packed big endian. Is that really the native sampling > format of the device? Yes, it is. There is no sample modification done in the driver at all to keep it as simple as possible. Hence I can exclude that as the culprit :) Daniel
