hey so I’m trying to think of a way to generate structures for music pieces.
the idea is to divide a certain length into different parts that make up a structure.
I want to control the complexity (weirdness) of the structure with a percentage.
The percentage should determine the length and order of divisions, if their is a lot of change in rhythm and if the colours are harmonious or more chaotic
so low percentage should result in something like the first rectangle, a high percentage something like the second.
afterwards the parts get generated as types(chorus, verse, outro,…) depending on their length and a musical harmony (root, minor, major,…) depending on the parts color, the previous parts color and the next parts color.
any sense of a direction to think in is appreciated
ok I’m thinking to use Perlin noise and take data at various zoom levels and layer them on top of each other. So a high percentage of weirdness would just have more iteration of scale/layers then a low… let’s hope that works for starters.
each line can represent an iteration of the structure. so I’ll try to composite them to form 1 structure. the first lines of value’s should count more than the last ones.
it’s still not quite what i want. I’m looking for something else than an average i guess.
no one did something like this before? or knows of some method thats normally used for this kind of thing?
I don’t seem to find much on the net, or maybe i don’t quite know the terminology to find it…