This repository was archived by the owner on Jun 2, 2025. It is now read-only.
Replies: 2 comments 1 reply
-
I'll see if I can get this in: https://gist.github.com/trygvebw/c71334dd127d537a15e9d59790f7f5e1 It doesn't have all the convoluted stuff for webui and stuff, and already or k sampler, so probably easier to implement. No promises though. I tend to break tensors left and right. :P |
Beta Was this translation helpful? Give feedback.
1 reply
-
To put it mildly I would say that foresight is not the first quality in ComfyUI's discussions out of here... ;) |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
I suggested it in ComfyUI's discussions but it got buried, But then WASasquatch was the only person to actually reply so perhaps you might give it a shot.
Its a method of rather then just comping noise into the image, you work the image through the denoiser somehow to generate the noise, with the idea that for sequences of images would have similar noise patterns, this beats out freezing the noise since that just makes it feel like their is dirt on the lens.
In A1111's webui its a script called img2imgalt.py And my limited python makes it seem reasonable.
I'm kinda sure that with this one extra step on top of the batching that this node pack added you could get Corridor Digital to adopt it, they love them node workflows.
Beta Was this translation helpful? Give feedback.
All reactions