[Question] What are recommended parameters for transforming existing 1920x1080 footage?

hotshotco / Hotshot-XL

✨ Hotshot-XL: State-of-the-art AI text-to-GIF model trained to work alongside Stable Diffusion XL

https://hotshot.co

Apache License 2.0

982 stars 77 forks source link

[Question] What are recommended parameters for transforming existing 1920x1080 footage? #22

Closed vedantroy closed 8 months ago

vedantroy commented 8 months ago

Hi! I'm trying to transform some existing footage that has resolution 1920x1080 and I'm wondering what are the correct parameters to set during inference. I.e., I know there's width/height, crop top/left, and target width/target height.

My intuition is to

set target_width/target_height to 672 x 384 (most closely matches 1920x1080)
set og_width/og_height to 1920x1080
Resize the footage to 672x384 before feeding it into the model

Does this sound correct?

aakashs commented 8 months ago

For usage with ControlNet @ inference? Yes that is correct!