ComfyUI Node
Latent Reaction-Diffusion
A ComfyUI node in Latent/Noise with 12 inputs and 1 output.
Latent Reaction-Diffusion
- latent
- mask
- LATENT
◄seed0►
◄iterations200►
◄feed_rate0.036►
◄kill_rate0.065►
◄diffusion_u0.16►
◄diffusion_v0.08►
◄time_step1.00►
◄strength0.75►
◄channel_modeshared►
◄temporal_modelocked►
CategoryLatent/Noise
Inputs (12)
| Name | Type | Default | Description |
|---|---|---|---|
| latent | LATENT | Latent that will receive reaction-diffusion patterns. | |
| seed | INT | 00–18446744073709550000 | Seed for the initial chemical concentrations. |
| iterations | INT | 2001–2000 | Number of Gray-Scott simulation steps. |
| feed_rate | FLOAT | 0.0360–0.1 | Feed rate (F) controlling how quickly chemical U is replenished. |
| kill_rate | FLOAT | 0.0650–0.1 | Kill rate (K) regulating removal of chemical V. |
| diffusion_u | FLOAT | 0.160–1 | Diffusion rate for chemical U. |
| diffusion_v | FLOAT | 0.080–1 | Diffusion rate for chemical V. |
| time_step | FLOAT | 1.000.01–5 | Simulation time step used during integration. |
| strength | FLOAT | 0.750–5 | Scale of the normalized pattern relative to the latent's standard deviation. |
| channel_mode | COMBO | shared | Reuse one simulation for all channels or rerun per channel. |
| temporal_mode | COMBO | locked | locked reuses the same pattern for every frame; animated reruns the simulation per frame. |
| maskopt | MASK | Optional mask (often image-sized) to limit the pattern injection to masked areas. The mask is resized to latent resolution (bicubic when downscaling). |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| LATENT | LATENT | — |