Partito has migrated to Python and is now an open source GitHub repository!Partito is a system for bit-level encoding of harmonic combinations, implementing polyphony as a kind of ternary computing. It is an intuitively navigable and computationally efficient representation of 12-note harmony.
Partito models a neural-network-like system at byte scale, making predictive computing possible through byte partitions. The minimization of activation-entropy in the network itself makes harmony clear.
Operating at transistor level, the system should be a useful core element both in dynamic music creation and in the categorization and analysis of audio and notated music through digital media.
An assortment of automatically generated examples can be found here on SoundCloud.
Paritito uses bit permutations to define harmonic paths and possibilities at the very core of the computer: complete chordal and modal analysis can be reduced to 32 bits/slice, and decompressed without loss. The system builds a geometrically complete bit-space in which a user can move without limit.
To the right and below are images of the engine at work in 1) Unity (showing the geometry of the space) and 2a) and 2b) Pure Data, with descriptions of inputs and outputs.
Partita is currently operable in C and will be exported to Python for use with Google Magenta and other sound/music analysis platforms.
Please send inquiries via the contact form.