pure data externals & generative audio systems.

Stochastic processes, drift, noise, self-excitation, and chaotic dynamics, models drawn from physics, biology, and probability, and turned toward sound. The lineage runs through Cage and Xenakis, indeterminacy and statistical mass, though our own instruments favour chance with a shape: bounded, structured, audible as process, the edge of the deterministic.

// log
rhope~: two-barrier drift-diffusion rhythm generator

rhope~: two-barrier drift-diffusion rhythm generator

A signal-rate accumulator drifts and diffuses between two bounds. Each crossing fires a commit, resets the walk, and reports which side it landed on, the deliberation time, and whether a timeout forced the decision. Built on the same first-passage-time mathematics used to model deliberation and decision-making.

→ read more
Hawkes processes for self-exciting rhythm

Hawkes processes for self-exciting rhythm

A point process where every event temporarily raises the chance of the next, producing the organic bursts and lulls you don't get from a plain random clock. The basis for the multi-voice coupled-Hawkes performance system and the hawkes~ external.

→ read more
Corpus Explorer: navigating a sound corpus by ear, not by filename

Corpus Explorer: navigating a sound corpus by ear, not by filename

Every grain in a corpus gets a 29-feature vector: MFCCs, spectral contrast, onset strength, spectral slope. PCA and t-SNE fold that down into a space you can fly through. Picking a point runs a weighted k-nearest-neighbour search and triggers overlap-add granular synthesis on whatever is closest.

→ read more