Nodes/ComfyUI-ZImagePowerNodes/⚡| VAE Encode (for Soft Inpainting)
ComfyUI Node

⚡| VAE Encode (for Soft Inpainting)

Encodes the input image and its inpainting mask into latent space using a VAE model.

By martin-rizzo·Created 7 months ago·Updated a day ago· 346
⚡| VAE Encode (for Soft Inpainting)
  • pixels
  • vae
  • mask
  • LATENT
output_size
mask_blur_pixels
Category⚡Z-Image

Inputs (5)

NameTypeDefaultDescription
pixelsIMAGEThe input image to encode
vaeVAEThe VAE model to use for encoding
maskMASKMask indicating the area of the image to apply inpainting.
output_sizeCOMBODetermines the size of the output latent space.
mask_blur_pixelsFLOAT0–256Width in pixels for softening the mask edges. Higher values create a smoother transition with larger masked area.

Outputs (1)

NameTypeDescription
LATENTLATENTLatent space representation of the encoded image and mask.