Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Feature Request: Loop Render (using previous AI image as input) #17

Open
enzyme69 opened this issue Oct 24, 2022 · 6 comments
Open

Feature Request: Loop Render (using previous AI image as input) #17

enzyme69 opened this issue Oct 24, 2022 · 6 comments
Assignees
Labels
enhancement New feature or request

Comments

@enzyme69
Copy link

Describe the feature you'd like to see:

Since unlimited "local render" is working, maybe we need also the ability to batch render static or animated frames -- with different random seed or same seed. I would like to render over night usually, since my local render is kind of slow 4-10 minutes per image, I can let it render for 6-8 hours while asleep or when computer idle.

Another request is if we could "loop" render X times the image from last render or image last AI render, sometimes it could produce interesting details.

Thanks!

Additional information

No response

@benrugg benrugg changed the title Batch Render or Loop Render Feature Request: Batch Render or Loop Render Oct 24, 2022
@benrugg
Copy link
Owner

benrugg commented Oct 24, 2022

Yeah, this is a great idea, and it's in my plans. I definitely want to make this happen.

@benrugg benrugg self-assigned this Oct 24, 2022
@benrugg benrugg added the enhancement New feature or request label Oct 24, 2022
@Ainaemaet
Copy link

Was just about to write up a feature request for this. Very happy to see it being worked on. We're blessed to have devs with a positive work-ethic.

Thank you @benrugg !!

@benrugg
Copy link
Owner

benrugg commented Nov 11, 2022

Batch processing is now possible with the animation support in the latest release! You can animate Stable Diffusion settings (including prompt text!) and run as many variations as you want.

I will leave this open, though, because the one thing that isn't available yet is "loop rendering". ie. the ability to use the last AI image as the start for the next one.

@benrugg
Copy link
Owner

benrugg commented Nov 11, 2022

See the Animation Instructions page for more info. (And I'll be posting more instructions there soon)

@benrugg benrugg changed the title Feature Request: Batch Render or Loop Render Feature Request: Loop Render (using previous AI image as input) Nov 29, 2022
@TutajITeraz
Copy link

Hi, I'm trying to achieve something similar, but it seems to me that instead of adding one option, it's best to make Compositing Nodetree work properly. Currently, for some reason, anything after AIRender and before Composite node - is ignored and not saved, even if save before and after is checked. I tried saving it manually because the preview in "Render Result" appears correct.

@benrugg
Copy link
Owner

benrugg commented Dec 23, 2022

@TutajITeraz ah, that's a really good point. It's working that way by design, but tbh I probably didn't think that through well enough. It certainly seems like it should save the final composite, not just what was returned by Stable Diffusion. I'll add this to the list for the future.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

No branches or pull requests

4 participants