| Subject: | Re: [LCtalk] lanbox lcm triggered by music |
| From: | Jeffrey Burns |
| Date: | Mon, 03 May 2004 09:17:55 +0200 |
Best to you, Fokko,
The idea I gave was something that I never really worked out, since I
mostly work with midi. But at one point, I did some experiments with
finding the fundamental tones of harmonies by speeding up little bits
of sound (1/10 sec.) by a factor of 4 and sending them to fiddle~,
which is the frequency detector in Max. That seemed to work well. The
reason for speeding up the sound is that fiddle~ is deaf to low
frequencies. Separating frequencies in Max with low and high pass
filters is pretty easy. That would produce different dynamic analyses
for low and high tones. (I find the GRM filter best, though it is
expensive on CPU capacity.) Putting these techniques together should
make some kind of a usable machine. A lot depends on the particular
kind of music being analyzed. For example, if the music does not have
strong beats, it might be a good idea to let changes in the
fundamental tone trigger the detection process, rather than the
dynamics. I don't know of examples by other people on this, though it
could be that some of the fft experts on the Max list would have
interesting suggestions.
Jeff Burns
--
http://www.piano-of-light.com
http://www.maxsynagogue.com
- Re: [LCtalk] lanbox lcm triggered by music, (continued)
Re: [LCtalk] lanbox lcm triggered by music, stijn slabbinck
|
