Pick two auditory channels for two variables and hear them together on random data, with a few levels each, before you bind a real data lane to them. Each cell of the matrix is a pair; click it or press Enter to hear it.
For hearing, not for ranking. A pair that sounds clear here is not proven to work on your data. The page states no listening-test results. Headphones help with pan and fine loudness steps. Sound starts only when you press Play or pick a pair.
The matrix keeps eight channels. If the kept pair uses the channel you swap out, the new one takes its place, so you hear the same pair with one channel changed.
The drum lane is a first look at a percussion layer: one drum on sixteen steps under the pair, with one variable, V4, on one of its channels at four levels.
V3 is a third variable over time. It does not sit in the matrix: it changes how many events sound (density) or how fast they come (pace), not how each note sounds.
Temporal pattern: the share of events whose level follows a slow curve over the loop instead of chance, so you can test whether a pattern comes through.
Paste a pair copied from this page, or open a saved .json file, then press Play. The JSON carries the seed, so the same JSON gives the same loop.