Replies: 1 comment
-
The best method I've found so far is to first generate txt2img without AnimateDiff. Then send the image+prompt to img2img. Enable AnimateDiff and experiment with denoising between 0.7 and 0.9. Lower looks more like the original image, but seems to have less motion. 0.75-0.80 seems best in my limited tests. ControlNet Canny is better at keeping the look of the original image, but seems to almost eliminate motion. I haven't tested ControlNet settings enough to know if it would work better. |
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.
-
How can I control the first image generated by animatediff? My current problem is that when I generate a picture by texttoimage , I can get the picture that I'm satisfied with. However, when I choose animatediff to generate a dynamic picture, I find that the first frame of the dynamic picture differs significantly from the picture which I got from texttoimage. Is there a better solution?
Beta Was this translation helpful? Give feedback.
All reactions