ControlNet settings:
Preprocessor: none Model: MLSD.
Control mode: Balanced or My prompt is more important.
Resolution for txt2img: 512x768
You can use it with OpenPose (as a second ControlNet unit. In this case use 0.8 weight and "My prompt is more important" mode for MLSD) or generate background and character separately then use img2img to merge them together (this method works much better than 2 ControlNet units)
Description
FAQ
Comments (4)
Could you elaborate further on the "use img2img to merge them together" portion? I've been trying to figure out how to combine a background and character just as you're saying but can't get the 2 controlnets method to work!
When I mentioned 'use img2img' I meant generating the background and the character separately on a white background and then removing the white background from the character image using graphic software. Afterward, merged the character with the background (still in graphic software) and use img2img to create smooth image.
@BlazzzX4 Ah gotcha, I thought that's what you may have meant but wanted to make sure I wasn't missing out on a new workflow. Thank you for the reply!
@Bruce_Genner theres got a to be a way to seperate the background and then merge it back with a script ... someone had to have made that already right?