On 2/10/07, jkelly(a)fas.harvard.edu <jkelly(a)fas.harvard.edu> wrote: I'd assume that his concern is due to Vorbis being lossy. Considering that for a very very long time, the gold standard for much speech related research has been 8 or 16khz ulaw compressed audio, I seriously doubt Vorbis is going to impeded much research.. but, I certainly do see why we wouldn't want to just throw information away. Oh.. *sigh* I wish we had a better way of having linked 'source' and 'distribution' forms in commons. For 99.999% of our applications, the lossily compressed file is really what we want to distribute.