Skip to content

Encode Masked Media Latent

Encodes a source image plus a binary mask into an InpaintMaskArtifact — the input format that switches Generate Media Latents into inpainting mode.

Category: ModularDiffusion/Encode\Decode

TL;DR

  • Use for inpainting. The resulting artifact carries the mask + masked-image latent + strength together so downstream nodes don't need separate inputs.
  • Connect the latents output to a Generate Media Latents node's input_latent. The Generate node detects the artifact type and routes through the inpaint pipeline automatically.
  • Only works with pipelines that have a registered inpaint pipeline class (e.g. Flux Fill, SDXL Inpaint).

Typical workflow position

Load Image ──┐
             ├─→ [Encode Masked Media Latent] → Generate Media Latents → Decode
Paint Mask ──┘

Node preview

Encode Masked Media Latent

Inputs

Name Type Required Notes
pipeline Pipeline Config Yes Must support inpainting (driver has _inpaint_pipeline_class).
image ImageArtifact / ImageUrlArtifact Yes Source image.
mask ImageArtifact / ImageUrlArtifact Yes Binary mask — white pixels = inpaint region, black = keep original.
strength float (0.0–1.0) No Inpaint denoising strength, default 1.0. Lower values preserve more of the original.

Outputs

Name Type Notes
latents InpaintMaskArtifact Bundle of mask + source/masked-image latents + strength.

Tips & pitfalls

  • Pipeline must support inpainting. If the chosen pipeline has no inpaint variant the node errors on validation. Pick e.g. Flux Fill on the Pipeline Builder.
  • Mask must match the source image's dimensions (it's resampled to the latent grid internally — a wildly mismatched aspect ratio will distort).
  • strength lives on the artifact, not on Generate Media Latents. Adjust it here.

See also