Genetic programming synthesiser for windows?
-
- KVRian
- Topic Starter
- 1273 posts since 9 Jan, 2006
http://www.pawfal.org/Software/fastbreeder/
This looked interesting and was wondering if something like that could or already has been made as a VST.
I'm also just curious what peoples thoughts might be on using a genetic approach for synthesis.
This looked interesting and was wondering if something like that could or already has been made as a VST.
I'm also just curious what peoples thoughts might be on using a genetic approach for synthesis.
- KVRAF
- 9577 posts since 16 Dec, 2002
kyma has something likt this......havent played with it much though
Amazon: why not use an alternative
-
- KVRAF
- 1615 posts since 19 Aug, 2004 from Toronto
I like the idea.
I've been wanting to see/hear more audio oriented approaches to this kind of thing, as I've seen quite a bit of intriguing visually themed work lately.
Much of it coming from people using Processing.
http://www.processing.org/
I've been wanting to see/hear more audio oriented approaches to this kind of thing, as I've seen quite a bit of intriguing visually themed work lately.
Much of it coming from people using Processing.
http://www.processing.org/
drab
-
- KVRist
- 205 posts since 25 Feb, 2005 from Germany
Novel idea, yet I think it would take too much effort to get anything but a hideous sound out of that.
It might be interesting to see genetic algorithms used in hosts, though, for the generation of presets.
In FL Studio, for instance, you currently can use "hybrid" mode to mix presets. This could be done more sophisticated. Eg., first you define a group of presets to start with. Then the host generates a new group out of that, creating new presets by combining the settings of former presets and adding a little randomness here and there. Then you listen to the new presets and delete those which you don't like. Out of the remaining presets, the host generates a new group, and so forth.
This could be done with any synth. To get more desirable results, the user might exclude certain controls from the process as well as define value ranges for the controls to be randomized.
It might be interesting to see genetic algorithms used in hosts, though, for the generation of presets.
In FL Studio, for instance, you currently can use "hybrid" mode to mix presets. This could be done more sophisticated. Eg., first you define a group of presets to start with. Then the host generates a new group out of that, creating new presets by combining the settings of former presets and adding a little randomness here and there. Then you listen to the new presets and delete those which you don't like. Out of the remaining presets, the host generates a new group, and so forth.
This could be done with any synth. To get more desirable results, the user might exclude certain controls from the process as well as define value ranges for the controls to be randomized.