ComfyUI-Pixaroma

Pads an image with a solid colour so an outpainting model can fill the new area in, then optionally scales the result down to a megapixel limit and reports the final size.

Mid grey is the default fill. Any colour works, but a strongly coloured fill can tint the whole generated image, because a model trained to replace it learns the colour as well as the shape. Grey is neutral, so it has no hue to bleed.

To ratio grows the image to a target shape and the anchor decides which side the new space appears on. By side lets you set an exact number of pixels per edge. The megapixel limit is optional: with it off the image keeps its padded size.

The width and height outputs report the FINAL size, so they can feed an empty latent directly.

ComfyUI Node: Outpaint Pixaroma

Category

Inputs

Outputs

Extension: ComfyUI-Pixaroma

More nodes in ComfyUI-Pixaroma

Other 👑 Pixaroma/✂️ Resize & Crop nodes

Run ComfyUI workflows without the setup