[Chris Cannam] I finally got around to set up an entirely unqualified and unscientific comparison between rubberband and the 'stretch' utility from the pvoc package (which basically is not much more than a dumb and straightforward application of Mark Dolson's original phase vocoder implementation). The comparison uses a 25-second audio fragment containing mostly accordeon, voice and some sparse bass/percussion. Audio is stretched to double duration. This isn't claimed to provide much insight of interest to anyone but myself, but if you find this kind of thing entertaining, text and audio output can be found at http://quitte.de/dsp/rubberband_vs_stretch.html Cheers, Tim