Replies: 1 comment
-
AnimateDiff is already designed to maintain consistency within a batch. https://github.com/Kosinkadink/ComfyUI-AnimateDiff-Evolved AFAIK, There is no way to achieve higher consistency than this in AnimateDiff. |
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.
Uh oh!
There was an error while loading. Please reload this page.
-
Hi there
I was thinking: what about creating a new very simple node that would take from the AnimateDiff just the section that actually cares for the consistency between the frames so that when you do a video KSampler does not start each new frame as separate new one but rather continues from the previous frame - would that be even possible?
Or even better: could that be implemented as a simple check button (boolean) in KSampler GUI that when TRUE it would do the frame to frame consistency as described above?
Cos with such node + ControlNet openpose we would not need to use old AnimateDiff with all its side effects anymore to create a video!
And also similarly when creating a one still image it would actually make possible to have totally consistent persons, clothes and backgrounds between the images!
Beta Was this translation helpful? Give feedback.
All reactions