Since CivitAI is going down THE DRAIN
I won't be updating this workflow here anymore, and I won't post any more models from now on.
I hope to find interest from people to keep updating this. Thanks everyone!
Keep following this workflow development and more on my new Discord server: https://discord.gg/EwKE8KBDqD
Get new releases from GitHub: https://github.com/diodiogod/Comfy-Inpainting-Works
Now there is a "simple/compact" version that I will update together with the normal one.
Read "about this model" on the right panel please!! I recommend reading v6.0, v5.0, v5.1 as well.
This is a proper inpainting with alimama control-net + Flux Fill + Flux Tools (Depth an Canny) support from V7 and above.
I try my best to avoid hiding the workflow flow, while balancing keeping it clean and user-friendly.
This workflow will make sure your image is composited in the end to prevent VAE degradation AND from v5 will make sure your image is a multiple of 8 (so it doesn't get resized by ksampler) at the same time it avoids resizing at all (unless you want to) since resizing will degrade the original image as well.
For inpainting that leaves a face and changes the whole background and character, Alimama works way better. Alimama also works better for LoRas. Alimama is terrible at outpainting and Flux Fill is amazing. You should test both. My comfyUI OOM or crashes when switching from one to the other IF my system memory page file is not large enough. With a pagefile of 50GB I can freely switch from Flux Fill to normal Flux Dev now.
It does use a lot of custom nodes to make life easier and workflow cleaner. Use manager to install the missing ones.
Supports loading of an external mask (need to match the image size) or doing the inpainting within the right-click "Open MaskEditor" on the image itself.
Supports FILL -the normal fill, like Auto1111- (use FastBypasser - black node)
Supports area inpaiting x full image inpaiting (use FastBypasser - black node)
Supports variable denoise (Change denoise on the CONTROL ROOM)
Supports Negative (use FastBypasser - black node)
Proper composite so no VAE degradation
Support for Flux Tools (Depth an Canny) with Alimama
I added LoRa support and Daemon Detailer support (both are optional)
Description
I decided to ad Daemon Detailer as well. Sometimes it's better, but you'll have the option to use it or not.
I changed the way the image saver retrieves the prompt to a simpler method, should work now for everyone.
Added a guidance control as well.
Some other small changes like you can DISABLE the aura mask if you want. Read the notes on the green group.
FAQ
Comments (11)
Missing Node Types
When loading the graph, the following node types were not found
StringListToCombo
I can't even find it on Google -)
Oh. its the nodes that changes strings (the name of the scheduler, checkpoint, sampler) to a type that the ksampler accepts. Those accept "combo" but don't accept a string. So You need to convert them. "StringListToCombo" if from Logic Utils. I use them a lot: https://github.com/aria1th/ComfyUI-LogicUtils
Use the manager to install missing nodes, it's easier.
It's the first hit on my google search "StringListToCombo comfyui". Please tell me if it worked for you!
Strange. I had this node. Had to reinstall it. Thanks.
@eloidor Someone else also asked me about is as well. It's probably not just you who had a problem with it. Hope it is working now!
you need to review your workflow there is numerous problems, such as mask if you set that to "false" wich is already off, the workflow try to use it anyway and it don't work if you use only one image as a single input.... the lora stack too get some problems.
You see, it needs a mask. With the false toggle It will use the main (the middle one) loaded image mask, in which case you as the user need to make the mask by right clicking on the loaded image and opening the mask editor. You need a mask to tell where the inpainting should happen. Without that it won't work. Try that and get back to me please.
What is the problem with the LoRas Stack? can you show me a print?
@diogod it's what i done mate (look at my profile i made many inpainter models and workflows) i did already a mask on the main image (the middle one) and it return into errors on errors. and it's always the mask in cause, and the switcher is on "off" for "mask use". Boolean = off.
@EKKIVOK Well I can't really help you if you don't show me the error. I don't get this error, if I did I would not publish it. Can you show me a print, maybe? The mask nodes (the green group) needs to retrieve a mask, in the end it will distribute this mask to the other nodes. Is the mask being processed on the upper group?
On the current version I DID forget about the flux guidance. It works without it, but I'll update the worflow soon to add it.
My "very fainted aura mask group nodes" were not behaving as it should. I'll update the worflow once again.
I'll also move the ksampler "denoise" to the control room to make the tinkering options more centralized.