Update: Comfyui Workflow has been officially released!
This model is an inpaint model for realistic images.
-This model allows you to naturally modify parts of the generated images by prompting.Improve the quality of your inpainting with the rich knowledge of pony models and the flexibility of prompts to meet the needs of more creative purposes.
-V4 version
V4 was created by adding NoobAI to merge list and then remerging the illustrious models to restore realism.Non-remerging model also uploaded as ALT version.
The difference between the two is the internal ratio of NoobAI. The normal version is slightly better in skin tones, etc. In most cases, there's not much difference, but in N-situations or complex compositions, the ALT version will produce better output.
How to use:
Sampler
V3 and V4 = DPM++ 2M SDE (karras for comfyui) or DPM++ 2S a
V1 and V2 = Euler a
Soft inpainting check
Mask bluer 4~8
CFG Scale 4~7
Denoising strength 0.9~1
steps 20~50
-Soft inpainting settingchange "Transition contrast boost" to 1 (default 4)change "Difference threshold" to 0.1~0.25 (default 0.5)
Don't use "Apply color correction to img2img results to match original colors".
Refiner:
Using normal model as refiner (eg. pony realism) at 0.65~0.95 (recommend 0.8) can produce even more great results. I will pick out models that can be used as a refiner, so please try it out.
Good Refiner model list
Real Wives Pony v_1.1 [believe me, best skin texture ever. (need neg fat)]
https://civarchive.com/models/709934?modelVersionId=794078PornMaster-Pro Realism-IL-V3 [prompt variety, danbooru tag]
https://civarchive.com/models/1045588?modelVersionId=2099739RedCraft Illustrious3 Relustion [skin tone, prompt variety, liquid]
https://civarchive.com/models/958009?modelVersionId=1517097ULTRA v9 [soft, dark atmosphire]
https://civarchive.com/models/228525?modelVersionId=2098608Ponyvinculum Exotic_Mix [clear, soft, lighting, animal, nature]
https://civarchive.com/models/1424028?modelVersionId=1809768
Tips:
Only masked is best for quality
Start with Masked content "original", if not worked change to "latent noise" or "fill".
Use image size with height + width>2000~2500 (Recommended size for SDXL).
or use resized by scale around recommended size (aspect-ratio-helper).
If you mask a small part, it is a good idea to mask the top and bottom of the main mask so that the model can see wide area
Your image works as quality prompt, You don't need a lot of prompts, keep them to a minimum and let the model do work.
dog, in beach-Use this model only for AI-generated images.
-Prohibits uses that harm any person by using this model on real-world images.
Description
Changed the base model to ILLUSTRIOUS
Support for sampler "DPM++ 2M SDE"
Improved maximum size limit without corrupt.
Increased overall output image quality.
Adjusted v2's overly sharp output to slightly soft
[Cons]
Decreased response to certain prompts.
Reduced accuracy of hands.
FAQ
Comments (20)
Holy shit! I still use your V2 whenever I need to do pony inpainting, this is amazing!
Thank you, if you set it up correctly it should be better than v2 in many ways.
it's not a "pony" anymore but I'll keep the name to avoid confusion
Thanks God finally there's a good inpaint model for illustrious!
by the way, do you have any ckpt recommendation for inpainting anime picture?
https://civitai.com/models/1376234/noobai-inpainting-controlnet
I found a control net model, and I tried to make one myself, but it wasn't good, so I threw it in the trash.
habibing ahh... that one. Already knew it, but forgot that exist XD
does not understand how to use it, so i just skip using it.
And btw I'm in the middle of using yours, and it's kinda funny, in a good way.
You will understand later
Thank you. I was once scolded by civitai, so please avoid sensitive content as much as possible lol.
habibing Sheesh. Well at least my review are not realistic, so worry not XD
habibing oh i can't post images with rating R and above in yours huh XD
well that's an awkward post of mine hahaha
Oh wait, don't post nsfw stuff, this model has been deleted before.
habibing ahh my bad then
nvm,keep clean for everyone.
Do Loras work with this?
what would you say the advantages the IL model has over the Pony version?
Pony model has a hazy look as if the paint has run out, but IL has a larger limit on the amount of drawing, and is cleaner. In terms of prompts, pony is often superior.
I tried your model and it works great for inpaint, but i have no clue on how to do outpaint like your example, can you tell me how?
In this example, I prepared an image in using a webui-extension called photopia, which can do things like Photoshop. First, load your image into photopia, select crop from the menu on the left and expand the box in the direction you want to outpaint. If the image size is too large, You can change image_size with Alt+Ctrl+I. Then click "send to img2img" button at the bottom (unfortunately, you cannot send it directly to the inpaint tab). Right-click on the image in the img2img tab to copy it and paste it into the inpaint tab. Your image is now ready.
Step 2. Mask all of the expanded areas, and also mask the borders of the original image a little. and also mask any areas you want to inpaint. For outpainting, use "latent noise", with "denoising strength" set to 0.9-1 (0.95 recommended). "Whole pictures" are probably best, but "only masked" is fine too, to widen the reference area, if it's a person, it's a good idea to mask the area above their head with a dot.
In short, prepare a canvas you want to outpaint, paste an image onto it, and mask it. The extended part can be white, gray, or black.