gogins
commented
3 years ago - [x] Rework HarmonyIFS code to use prime form, inversion, pitch-class transposition, voicing permutation, arrangement permutation, and loudness permutation as the basis of the score function space. Too complex -- implemented with just prime form, inversion, PC inversion, and voicing permutation. Instruments are voices and duration is nominal.
- [x] See if I can get this to make some decent music. Seems to be working, that's enough for now.
Work on PITV and HarmonyIFS reveals issues in underlying concepts.
The basic idea has always been to use a fixed point as a score, where the recurrence generating the fixed point produces harmony as a function of time (harmony does not fold or rotate in time), as well as other musical features.
What I have done in HarmonyIFS is use fractal interpolation in time, prime-form, inversion, transposition, key number, key velocity, channel space where the actual fractal interpolation occurs only for prime-form, inversion, and transposition as a function of time.
I need to experiment with adding various other subgroups to this recurrence, and to reconsider other means besides fractal interpolation of constraining harmony to be a function of time, and to link harmony with actual notes.
This is both an enhancement and, perhaps, a bug.