Stuart Frederich-Smith (SignalFunctionSet) released a VCV Rack module called Gravity in May 2026. At its center is a circular display with a single moving point. Six modes govern its motion: a double pendulum, a spring-bound rocket in a planetary gravity field, a billiards table, a Pac-Man-style maze, a LOGO turtle drawing, a spirograph (Maurer rose). Whatever the mode, the moving point's spatial coordinates are converted in real time into voltage outputs — bipolar X/Y position, radius, angle, six sector CVs, six boundary-ray gates. These voltages can patch into anything: pitch, filter cutoff, modulation depth, envelope speed. The pendulum plays. The maze plays. The billiard ball plays — without knowing it.

The step sequencer paradigm, from the Buchla 248 through every modern Eurorack sequencer, rests on a premise that has never changed: a composer writes a list of pitches, a clock steps through the list, music results. This is already-construct — the sediment that a completed chisel-construct cycle deposits. The modular community has spent decades refining this paradigm: probabilistic sequencers (Turing machines, random shift registers), Euclidean rhythms, Markov chains — all sophisticated, all operating within the same epistemological frame. Even analog chaos modules (Lorenz attractors, electrical feedback loops) get absorbed into this frame; their outputs are understood as "a kind of randomness," one more flavor of uncertainty layered onto the note list.

Gravity's chisel: remove the note list from the definition entirely. It substitutes a different question for "which pitch should play?" — where in space is the moving point of this physical system right now? The double pendulum's chaotic arc, the perturbed orbit, the billiard's collision path, the topology of a Pac-Man maze — each system evolves according to its own dynamics, spatial coordinates become voltages, voltages become sound. The composer did not write the pitches. Physics did. This is not random (physics is deterministic), not designed (you set initial conditions, not outcomes), not AI-generated (no training set, no learned pattern). It is a third state: determined but not commandable geometry.

Where is the remainder? In the spatial trajectories of these physical systems — the chaotic arc the double pendulum traces through air, the reflections a billiard ball carves across a circular table, the LOGO turtle's drawing accumulated over minutes, the closed symmetry of a Maurer rose. These trajectories have always contained musical structure (periodicity, complexity, boundedness, sensitivity to initial conditions) but were never absorbed into the sequencer paradigm. Not because they weren't interesting, but because the note-list paradigm was already complete enough that no one needed to reach for a physical system instead. Gravity makes this remainder audible: the Pac-Man maze's path topology was never music, but it was always a spatial system, and spatial systems were always potentially musical. The 1967 LOGO turtle, the 1980 Pac-Man, billiard mechanics, planetary gravity — all were used for decades in their own contexts without anyone reading their motion as voltage.

Frederich-Smith's broader practice is archaeological. He has previously ported Barry Truax's GSX granular system (1985), Steve Reich's tape phasing, Paul DeMarinis' Pygmy Gamelan, and the IRCAM CHANT project — historical computer music techniques that existed briefly, were submerged by commercial tools, and settled into sediment. He excavates them and makes them playable again inside contemporary modular synthesis. Gravity is not a port of any historical system; it is a new invention from the same structural orientation: physical system trajectory as musical material is the logical extension of an archaeological practice.

Seeing this now matters more than later. The VCV Rack library is hardening into a categorical vocabulary: "sequencer," "chaos module," "random source," "LFO." Gravity sits outside all of them — "chaos" in the modular context means electrical feedback chaos, not physical simulation; "random" implies probability, not deterministic geometry; "sequencer" requires a note list, which Gravity does not have. The creator's own description is "chaos and motion engine" — both words are approximations; "chaos" points at the wrong tradition, "motion" is correct but not yet specific. Once the categorical vocabulary fully sets, the physical simulation approach will be absorbed into the nearest existing label and lose its structural distinctiveness. It hasn't yet. This remainder is still alive.

Han Qin (秦汉) · @hqinjarsy · SAE Judgment & Aesthetics