1.4
Nothing changed, just fixed the video combine nodes to work again.
after i finish my other flows i am working on will update this to be better designed
1.3
added sam/segs mask option
added watermark removal option
added more knobs and switches.
many small changes
included my upscale/extas workflows from my advanced video flow they work fine here.
https://github.com/guoyww/animatediff/ my animate diff lora is from the release of v3 i think. I use the motion module and lora from here.
my lcm lora was from here https://huggingface.co/latent-consistency/lcm-lora-sdv1-5
Both these loras are loaded in the model area in the named lora loaders, ensure you load both of them. LCM is not optional, animate diff lora can be bypassed if you want.
strength of 0.6 lcm lora
strength 0.5 - 1.0 animate diff lora
I use V3 animate diff
https://github.com/guoyww/animatediff/
v1.2
added simple masking
added a few extra control options for upscale.
added background control if you create masks
added inpainting option if you created masks.
v1.1
updated/fixed ip adatpor nodes
added open pose gen
added a few switches to turn on/off things easier.
added controls for open pose.
Simple video to video
This was made for all the people who wanted to use my sparse control workflow to process 500+ frames or wanted to process all frames, no sparse. That flow can't handle it due to the masks and control nets and upscales... Sparse controls work best with sparse controls. If you want to process everything. This is how you do it.
This workflow has no masks, no cutting, no batching/unbatching, no enhancing no frame skipping, It's just a straight-up video changer that changes one video to another. Because no heavy loading is done, you can process 200+ frames in 1 animate diff batch, really I don't know the limit.
All previews here can still be run through my enhancer in the sparse control flow. I only included the base generations previews from this workflow. They can be further enhanced.
Replacing backgrounds in this is very hard, but possible with correct inputs, prompts and settings.
Gl and have fun
Description
FAQ
Comments (4)
I keep getting the following error:
When loading the graph, the following node types were not found:
SelfAttentionGuidance
Nodes that have failed to load will show as red on the graph.
I'm not sure if it's a module that's missing, nothing comes up when I use installed missing modules. It must be a part of a package it's not detecting. I did try to update comyfui as well.
"ModuleNotFoundError: No Module named 'comfy.options' ? how do I fix this? I'm thinking the MagicAnimate changed something drastic. Now ComfyUI won't open. Thank You
I keep getting this error on the first KSampler: RuntimeError: shape '[48, -1, 40, 40]' is invalid for input of size 204800
I really like this WF but it is not working correctly. First I have issues with nodes that were supposedly installed that do not show as installed in the flow. being the IPAdapterapply node and the metadata nodes. I think it IPAdapter is causing issues with the K samplers as neither is working and fail to execute. I tried cloning and "fix Node" but not able to resolve. Any suggestions?
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.