🛠️ Purpose & Design Philosophy
This workflow is a high-fidelity environment built for Illustrious. It prioritizes stability and professional texture over generation speed. It follows an "all-in-one" philosophy: configure your prompts, hit queue, and let the workflow handle the multi-stage refinement from start to finish.
Not for Speed: This is a heavy-duty refinement tool. If you want 2-second previews, use a basic workflow.
Personal Use: Built for my specific production needs. It is shared as-is for those who want a "set-and-forget" pipeline for Illustrious. Adjust the settings and models to fit your needs. The defaults on v19+ will be what I used for the sample images.
All-in-One Logic: The workflow handles generation, detailing, and upscaling in one continuous pass.
If you're needing something with more features: Try checking out Silly All-in-one, Multiple Characters With Regional Loras which is on civitai. I have no plans to implement additional features.
⚠️ Disclaimer & Compatibility
Install at Your Own Risk: Custom nodes can break your environment. I am not responsible for troubleshooting your specific installation.
ComfyUI: Built and tested on the non-app version. Desktop app users may face additional hurdles.
The "Your Version" Factor: Your node versions and environment are 99.9% likely to differ from mine.
Nodes 2.0: I do not recommend using Nodes 2.0. It creates unpredictable UI behavior; I will not provide support for issues involving this feature.
🤝 Support & Boundaries
I have not and never will monetize my workflow. CivitAI is the only platform I am maintaining this workflow on. If you find a version of my workflow posted and/or monetized anywhere and use it, then whoever posted it is responsible for providing you any support. This is the main reason why I don't monetize this: I don't want to be or feel obligated to provide support.
No DMs: DMs are disabled due to repeat spam. Please check the Discussions tab below; most questions have already been answered.
Modifications: You are free to hack this workflow apart. However, you are responsible for fixing it if it breaks.
Custom Requests: I do not make private workflows. If you need a custom solution, post a Bounty on CivitAI. There are many talented creators ready to help you for a fee.
Description
v11i changes:
Dropped KSampler Upscale flow
These seemed to do little to nothing.
USDU will remain for upscaling.
Concat conditioning will be bypassed by default now.
Added a toggle for this.
Custom Nodes used in this workflow as of v11i:
ComfyUI-Impact-Subpack (the Ultralytics Provider node was moved to here)
ComfyUI-ppm probably won't be detected using ComfyUI-Manager since I am only using it for its schedulers. You'll have to search for ppm in the Custom Nodes Manager or download/install it manually if you want to use AYS or Beta1_1 schedulers. ComfyUI-Manager is a nice tool, but it isn't perfect.
SAM Model used (This is needed for all the detailers):
sam_vit_b - download/install to models/sams folder
Ultralytics/Detection Models used:
Watermark Detection: nsfw_watermarks_s_yolov8 - download/install to models/ultralytics/bbox
Face:
Anzhc Face seg 1024 v2 y8n - download/install to models/ultralytics/segm
Eyes:
Eyeful_v2_Individual - download/install to models/ultralytics/bbox
Hands:
hand_yolov8s - download/install to models/ultralytics/bbox
FAQ
Comments (14)
V2!
I just tried using it, but I still don't understand what the non-magnifying nodes are for.
@JoyDopamine you mean the No Upscale USDU? It's just a 2nd pass to denoise the image some more.
@Gladas Got it, this is the photo washing feature.
Basically, If you tried to upscale only on the 1st USDU and it was doing some crazy stuff because your denoise was too high:
You can do a lower denoise on the 1st USDU and then on the 2nd pass you use the same or lower denoise to help refine/clean-up the image.
@JoyDopamine not sure on the term, but sounds right I think?
@Gladas Thank you for the answers and workflow!It feels very comfortable after testing.
Is there a reason why you're not using KJNodes' get and set nodes? It can make workflows easier to manage.
Yes. It's because they make it a real pain to update the workflow (for me). I absolutely will not use them ever again.
Suggestion: Using the 2x AnimeSharpV4 upscaler and then downscaling slightly (like 0.8x Lanczos) often gives sharper results. The issue with 4x-RealESRGAN-anime-6b is the oversized output, which requires heavy downscaling and can cause quality loss and jagged lines.
Thanks, I have tried many upscalers. I'm just using the classic 6b because it's fast and works good enough for me. I've already used the upscaler you mentioned along with many others.
Your advice is appreciated. Maybe you can write up an article and share it with the community. I'll even link it in the suggested resources if it's legit.
The structure and node organization in this is superb, like seriously, thats very neat, lot of effort went into making it so clean, "node porn" worthy, clearly made for personal use, with love.
Asides of that it also produces nice results from getgo with plug and play approach, works really fast avg 40s on 5090 considering the amount of stuff it is doing its very impressive.
Makes the laziest monkey promt into very impressive gen. Very low failure rates: almost no malformations or ridiculous artefacts when detailers go rogue. And to reiterate im talking about -> default bakes untouched by user, with some assembly involved this as might be only flow you ever need for majority off illustrious bakes.
No dependency custom node hell, even with Blackwell card and portable comfy on windows, probably smoothest heavy weight workflow load i had when plugging workflows from this website.
To me, 1 in every triple digit of workflows i load end up being useful/nice/productive, this is one of them. And first one i bothered to review in civitai.
Only thing that kinda bothered me was that there was some hoop jumping involved with finding the same ultralytics that author used and deployed with. I did found all of Anzhcs_YOLOs so thats oke, its quality work.
Still links to it probably should be included in description imo.
Hey thanks for the review and tip!
I think I posted a link to anzhc's hugging face, but it's probably buried in the description. Will take a look to make sure.
Yeah, it's under the commonly asked questions section. I'll probably just add a note on the workflow itself in the next version. I get asked about this a lot.