Marbles separates random rhythm (t) from random voltage (X) as two independently controllable stochastic layers

Marbles splits generative randomness into two halves. The left half (outputs t1, t2, t3) makes random gates/triggers — the rhythm layer — by jittering a master clock and deriving two contrasting gate streams from it. The right half (outputs X1, X2, X3) makes random control voltages — the pitch/modulation layer — clocked in sync with the t streams. The two halves are synchronized but parameterized independently, so a performer can randomize rhythm alone, pitch alone, or both. This separation captures a general generative-patch principle: temporal randomness and value randomness are far more musical and controllable when you can dial them apart rather than as one blob of noise.

Examples

Patch t1 and t3 to envelope/drum gates and X1-X3 as V/Oct to oscillators. Locking DEJA VU on X while leaving t free gives a fixed melody over an evolving rhythm; the reverse gives a steady groove wandering in pitch.

Assessment

State what each half of Marbles controls, then describe the patch and settings that randomize rhythm while pitch repeats, and the settings that do the opposite.