dyim@athena.mit.edu (Derrick H. Yim) (11/13/90)
Hello everyone out there in Net-Land! Has anybody who has used Soundedit (by Farallon) have any guesses at how the people who wrote it implemented the "Tempo" feature? Choosing "Tempo" from the "Effects" (?) menu allows the user to either double or halve (approximately, as far as I can tell) the playback rate withoutaffecting the pitch. I'd eventually like to implement such a feature in hardware, but would like to first try it out in software on either a Soundedit file or on a 'snd' resource. If someone could send me some sample code (Mac code preferred, of course), an algorythm, or a reference (book, paper, journal, etc.), I'd be eternally greatful. Thanks. Derrick Yim dyim@ahtena.mit.edu P.S. please reply by e-mail if possible, as I need to do this ASAP.
dyim@athena.mit.edu (Derrick H. Yim) (11/13/90)
After receiving some e-mailed responses, I realized that I phrased my question in somewhat of a misleading manner. When I said I wanted to "double or halve the playback rate", I wasn't referring to the sampled rate (e.g. 20KHz, etc), but rather the playback tempo, speed; e.g. I want a way to make someone sound like they can talk twice as fast without sounding like Alvin the Chipmunk. So with that, here's my new and improved post, which hopefully is a little bit clearer: Hello everyone out there in Net-Land! Has anybody who has used Soundedit (by Farallon) have any guesses at how the people who wrote it implemented the "Tempo" feature? Choosing "Tempo" from the "Effects" (?) menu allows the user to either double or halve (approximately, as far as I can tell) the tempo at wich the sound is played back without affecting the pitch of the sound. For instance, if you recorded yourself saying "Mary had a little lamb", and it took you 2.5 seconds to say it, you could then alter the waveform so that it would sound like you uttered the same phrase in 1.25 seconds, but your voice would still be recognizable (the pitch of your voice would be unchanged) I'd eventually like to implement such a feature in hardware, but would like to first try it out in software on either a Soundedit file or on a 'snd' resource. If someone could send me some sample code (Mac code preferred, of course), an algorythm, or a reference (to a book, paper, journal, etc.), I'd be eternally greatful. Thanks. Derrick Yim dyim@ahtena.mit.edu P.S. please reply by e-mail if possible, as I need to do this ASAP. Well, hope that's better. Keep those responses comin' and thanks a bunch to those of you who responded already.