Chemical Field · V concentration (click to seed)

FIELD ∇² · 128×128

Chemistry Parameters

Feed Rate (F) 0.055

Kill Rate (k) 0.062

Du (diffU) 0.210

Dv (diffV) 0.105

Speed 8×

Gain 1.0

Audio Output

hold the field / keyboard — or latch — to play

Mode

Sound Shaping

Filter Cutoff 8000 Hz

Filter Reso 1.0

Filter Type LP

Distortion 0.00

Bit Crush 16 bit

Rate Reduce ÷1

Scan Row 64

Stereo Width 0.00

Output Gain 1.0

LFO → Chemistry Modulation

LFO Rate 0.50 Hz

LFO Depth 0.000

LFO Target F

LFO Shape Sine

Amplitude Envelope (triggered on field click)

Attack 10 ms

Decay 200 ms

Sustain 1.00

Release 300 ms

Reverb & Delay

── Reverb

Wet 0.00

Decay 2.0 s

Pre-delay 0 ms

Damp 4000 Hz

── Delay

Wet 0.00

Time 300 ms

Feedback 0.30

Ping-Pong Off

Gray-Scott Equations
∂U/∂t = Du·∇²U − U·V² + F·(1 − U)
∂V/∂t = Dv·∇²V + U·V² − (F + k)·V

U, V are chemical concentrations at each grid cell. The Laplacian ∇² couples each cell to its four neighbours. F (feed) and k (kill) define the attractor regime. The V-field row is read out as an audio waveform at each simulation frame.

Parameter Regimes → Timbral Zones

Tuning F and k navigates between fundamentally different dynamical attractors — each producing a distinct sonic character without ever defining a waveform directly.

Spots / Mitosis

Low F, moderate k. Periodic stable spots. Produces tonal, harmonic-rich output with slow AM.

Stripes / Labyrinths

Moderate F and k. Travelling wave fronts. Output is rhythmic, band-limited noise with clear fundamental.

Worms / Chaos

High F, low k. Unstable filaments that collapse and respawn. Produces dense, inharmonic spectral chaos.

Silence / Extinction

Kill rate exceeds feed. V decays to zero. Analogous to oscillator death — a built-in fade-to-silence.

FPGA Hardware Architecture

Each cell in the 128×128 grid maps to a DSP slice performing 3 multiply-accumulate ops per sample tick. The grid updates in parallel — one full 128×128 field per audio sample at 44.1 kHz requires ~2M MACs/sec, well within a Xilinx Zynq-7020.

F, k, Du, Dv map to 12-bit DAC outputs from physical knobs. CV inputs modulate diffusion rates per-row, enabling spatial timbre control analogous to a vocoder but driven by reaction-diffusion topology. The audio output is a decimated row-scan of the V-field passed through a single-pole anti-aliasing filter.