> The speex resampler also doesn't match SRC fastest in terms of quality. > I've got some preliminary spectrograms here: > > http://www.mega-nerd.com/tmp/SRC_best_sinc.png > http://www.mega-nerd.com/tmp/SRC_medium_sinc.png > http://www.mega-nerd.com/tmp/SRC_fastest_sinc.png > http://www.mega-nerd.com/tmp/speex_resampled.png > > An explanation of how these plots are generated is here: Problem identified... Turns out it was a really stupid error that caused the window function to be severely truncated. The aliasing lines are now down to about -130 dB (on a similar chirp) and the "noise floor" is around -140 dB (still using single precision, no double yet). Would you mind re-generating the graph for the latest svn version? Thanks, Jean-Marc
