Image Paste Crop

This is the back half of a crop-process-paste pattern. You cut a region out with something like Image Crop Location - or one of WAS's Mask Crop Region nodes - did something to it (upscaled it, ran a second sampling pass, retouched it by hand), and now you need that patch back where it came from, at its original size, blended in rather than just stamped on top with a visible seam. That's this node's entire job: given the crop_data the original crop node produced and your (possibly modified) cropped image, it composites the patch back onto the source canvas.

How it differs from Image Paste Face Crop

WAS ships a second, similarly-named node - Image Paste Face Crop - and the README is explicit that they use different blending algorithms. Paste Face Crop's blending is tuned specifically for face-restoration-style pastes, aiming for the kind of seamless result you'd want from something like GFPGAN or CodeFormer. Image Paste Crop uses a different, more general-purpose blend, which the README frames as "desired in certain instances" - meaning it's the one to reach for when you're pasting back something that isn't a face, or when Face Crop's specific blending isn't giving you the result you want for your particular content.

How it works

Feed it your (possibly edited) cropped image, the crop_data that came from the matching crop node, and it composites onto the original canvas at the exact position and size the crop_data records. The output is the full-canvas image with the patch blended in.

The one thing to get right: crop_data is positional metadata tied to the exact crop that produced it. It's not just "an image at some size" - it encodes where that specific crop sat in the original image. Mixing crop_data from a different crop, or a different image, will paste your patch in the wrong place. If you resize the cropped patch before feeding it back (say, after upscaling it for a detail pass), keep that resize logic consistent with what the paste node expects, since it needs to know the scale changed to place things correctly.

Installing it

Part of the full WAS Node Suite pack:

• ComfyUI Manager - search WAS Node Suite, install, restart.
• Manual - cd ComfyUI/custom_nodes && git clone https://github.com/WASasquatch/was-node-suite-comfyui/, then pip install -r requirements.txt from inside that folder, then restart ComfyUI.

Common issues

Misplaced pastes almost always trace back to crop_data mismatch - double check it's coming from the same crop, on the same image, in the same run. If the seam is visible despite that, try Image Paste Face Crop instead if your content happens to be a face; its blending is specifically tuned for that case and may simply suit your image better.

And the pack-wide caveat: WAS Node Suite has been unmaintained since December 2023, and "Import Failed" errors after a ComfyUI update are a real, recurring complaint tied to older pinned dependencies. Reinstalling requirements.txt inside your ComfyUI venv is the usual fix if the whole suite stops loading.