MOSSImageControl - ComfyUI Built-in Node Documentation

Adds the color region and depth maps of a MOSS-Image color/depth control adapter to conditioning. Connect the positive and negative conditionings, a VAE, and one or both maps: the maps are encoded to latents, joined into a color/depth control signal, and attached to the conditioning for the part of the sampling you choose. Use it together with a MOSS-Image control adapter loaded through a LoRA loader.

Inputs

Parameter	Description	Data Type	Required	Range
`positive`	Positive conditioning to attach the control signal to.	CONDITIONING	Yes	N/A
`negative`	Negative conditioning to attach the control signal to.	CONDITIONING	Yes	N/A
`vae`	VAE used to encode the maps into latents.	VAE	Yes	N/A
`color_map`	One flat color per region, at the output size. The colors are only region IDs.	IMAGE	No	N/A
`depth_map`	Depth map at the output size, near is white.	IMAGE	No	N/A
`strength`	Scales the maps’ contribution to the image tokens. 0.0 is the same as no maps, 1.0 is the trained strength (default: 1.0).	FLOAT	Yes	0.0 to 10.0 (step 0.01)
`start_percent`	Sampling percent at which the control starts (default: 0.0).	FLOAT	Yes	0.0 to 1.0 (step 0.001)
`end_percent`	Sampling percent at which the control ends. Ending early keeps the layout and depth while the rest of the steps loosen the shapes (default: 1.0).	FLOAT	Yes	0.0 to 1.0 (step 0.001)

Note: Connect at least one of color_map and depth_map; without either, the node returns the conditionings unchanged. A missing map is replaced by zeros, so color-only and depth-only setups both work. When the two maps differ in size, the smaller one is scaled up to the size of the other. This node is marked as experimental in ComfyUI.

Outputs

Output Name	Description	Data Type
`positive`	The positive conditioning with the control signal attached.	CONDITIONING
`negative`	The negative conditioning with the control signal attached.	CONDITIONING

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): c4a40c76f26237408e47cb9f683c13bccc3507bd775bec1b89919947a308c115

​Inputs

​Outputs

Inputs

Outputs