ComfyUI Detailer/ADetailer Workflow
If you're having ANY issues with this workflow, please include the error log from your console when writing a comment here on this page. Chances are 99.98% i can help you with your issue - but i need the error log. And me not having to ask for it first will save you potentially hours. Chances are also high the solution to your problem is already documented in the FAQ here.
You can now support me on my Ko-Fi if you have too much money (✿◡‿◡)
Requirements for each version are listed below or can be found inside a Note in the Workflow itself.
Because of the many connections among the nodes, I highly recommend turning off the link visibility by clicking the "Toggle Link visibility" (Eye icon) in the bottom right of ComfyUI.
Description
I originally came from A1111 WebUI to ComfyUI and, honestly, a lot of things felt way more complicated than they needed to be. I couldn’t find workflows that were both visually pleasing and showed me all the important options without requiring a deep understanding of every little thing in ComfyUI.
Over a few months I learned the ins and outs of Comfy and built my own very barebones, simple workflow (v1) with one main goal: streamline the process and make it visually understandable. I released it here thinking it might help others who also want to make the jump from A1111 to Comfy.
Over the last year I’ve kept adding more and more features as people started using the workflow and requesting things. The workflow has grown a lot, but I’m still trying to keep the same ease-of-use that this whole journey started with. The main goal is still the same: give you a lot of powerful options in a layout that’s as clean, readable, and as user-friendly as ComfyUI will allow.
At this point I feel like I have a pretty solid understanding of how Comfy works, and I’ve even created my own custom nodes to add missing functionality so this workflow could match the ideas I had for it. I try to hide as much of the technical complexity as possible — but if you’re ever curious or confused about anything, please feel free to ask!
Thanks to all of your suggestions, the workflow now includes features like:
Single-image and batch generation
Automatic detailers for specific body parts
Upscaling
v-pred models
LoRAs
ControlNet
IPAdapter
Hires fix / refiner
Manual inpainting
Thank you to every single person who uses this workflow, donates Buzz, or shares images on this page.
I really appreciate you taking those extra steps to support and promote the project. ♥️
Requirements
v4.4 and above require ComfyUI version 0.3.51 or above and also need the frontend to be AT LEAST 1.24.3 or later.
v5.1 - Full List
v5.0 - Full List
v4.4 - Additions to v4.3 (IMG2IMG Only)
Otherwise same as v4-4.3 below
v4.2-4.3 - Additions to v4.1
Otherwise same as v4.1 (incl. v4) below
v4.1 - Additions to v4 (IMG2IMG Only)
Otherwise same as v4 below
v4 - Full List
v3 - v3.2 - Full List
v2.2 - Additions to v2
Otherwise same Custom-Nodes as v2 but you can remove Comfyui-ergouzi-Nodes
v2 - Full List
v1 - Full List
How to use
Since all of the different versions work differently, you should check the "How to use" Node inside of the Workflow itself.
I promise that once you read the explanation of the workflow itself, it'll click and it will be a simple plug and play experience.
It's the simplest I could've made it coming from someone who's only started using ComfyUI 4-5 months ago and had been exclusively an A1111WebUI user before.
When were what functionalitys added?
Starting from v3, ControlNet is included.
Starting from v4, IPAdapter is included.
Starting from v4.3, HiRes Fix and Dynamic prompts(wildcards) is included.
Starting from v4.4, Refiner is included.
Starting from v5.0, Manual Inpainting is included.
Any errors during execution?
If you're running into any errors during the execution of the workflow, please check the FAQ of my Guide first. The guide is written for the IMG2IMG Workflow but when issues arise that people run into frequently i'll add the solutions and what's hapenning to that FAQ section.
If you can't find the problem you're running into there - feel free to write me a comment on the model page so that i can help you and other people might benefit from it as well.
Feedback
I'd love to see your feedback or opinion on the workflow.
This is the first workflow I have ever created myself from scratch and I'd love to hear what you think of it.
If you want to do me a huge favor, you can post your results on this Model page.
I'll make sure to send some buzz your way!
Description
you have to install a new node for this version, check description for requirements.
Fixed Scheduler Selector after image-saver node update broke it (have to update the custom node)
Added upscaling factor to control the upscaling instead of leaving it to the upscale model
Added Low VRAM Options for IPAdapter Style, IPAdapter Composition and ControlNet
Optional use of Sage Attention Patch for global speed increase if you have triton installed
ControlNet Max strength reduced to 2 (was 10 before)
Removed automatic append of comma after start quality prompt and positive prompt for consistency between nodes and because ComfyUI broke Node-Groups with inputs like the seperator
Swapped ControlNet Start Percent and End Percent
Updated Manual inside the workflow
FAQ
Comments (30)
Hello, I have a question here. I have confirmed that all nodes have been installed and there are no missing nodes displayed on the manager page. However, the "workflowControlNet Models" node has been consistently missing
@vslinx I have read your reply to GesugaoDVA, which only mentioned workflowEnd quality prompt (Optional) and workflowStart quality prompt (Optional), but did not mention workflowControlNet Models
hey sora_935,
the problem is the same though.
One of the custom nodes that are required are missing. After you install them make sure to refresh the browser windows in which the workflow is open (even if you had closed the workflow, refresh the browser regardless).
You can also find this issue in the Guide linked at the top of the description under the „FAQ“ section.
Hello, I keep getting an error when starting up that the scheduler is not being transferred correctly from the “Scheduler Selector” to the “Sampler.” If I disconnect the selector from the sampler, I can select “karras” in the sampler and the image is created. Does anyone know what I might have done wrong?
Hi vslinx,
It's an insane workflow, but I only get good results with a V-Prep checkpoint. When I deactivate V-Prep and use a normal WAI Illu checkpoint, my images don't have color; they're all gray. I've tried different VAE and checkpoints. Could you please help me? I love your workflow so much!
Hey Otaku2D,
this sounds weird! Are all your custom nodes completely up-to-date? I‘m sadly on vacation until 9th of August so i don‘t have the opportunity to check what the issue could be. It could also be that a new update of a node broke something.
Are you using the newest 4.2 version? I know there was an update with the previous versions where i had to implement a converter for the Sampler, they might have changed something that might make this necessary for the scheduler too. Sadly i can only check this on Friday :/
Are the images created normally but just monochrome without any colors? i have never heard of a bug like this. Do you maybe have incompatible lora‘s activated or are using ipadapter? I‘m usually only generating on non v-pred models and have not encountered this yet
i also checked the description of the checkpoint you mentioned here: https://civitai.com/models/827184/wai-nsfw-illustrious-sdxl
and it seems that the VAE is already baked in, so you most likely have to bypass the vae selector as it could overbake your images. I recommend trying a different checkpoint to see if that might actually be the issue.
Hi vslinx,
Wow thanks for the quick reply. I will bypass and regenerate the VAE later. The images are also not completely monochrome but rather beige. If I define colours like blue eyes they also come up.
I have deactivated IPadapter, lora and everything else.
I'll get back to you tonight thanks a lot!
Heya Otaku2D,
no problem - always here to help :D
Hmmm this does sound like it‘s either the wrong clip skip or that it‘s loading the VAE twice or not at all.
I‘d recommend trying the following order:
1. Change Clip Skip to 1 in the Sampler Settings Group. Then Generate an image with all other paramaters from your previous tries unchanged. If this works now it‘d mean the checkpoint was trained on Clip Skip 1 (most checkpoints & loras are trained on clip skip 2)
2. Try bypassing the VAE Selector node in the Model preperation Group at the very start. If generating an image now it‘ll mean that your checkpoint already comes with a VAE and overwriting it with a specific one doesn‘t work because they‘re incompatible and it introduces artifacts.
3. If neither of these before work you can try to generate an image with the vae not bypassed and with a different checkpoint like this : https://civitai.com/models/140272?modelVersionId=2010753
If the generation with a different checkpoint does work it‘s most likely connected to the WAI Illustrious checkpoint and something i can look into as soon as i‘m home again. Otherwise if you can provide screenshots of any errors your console might show i could try helping you fix it before then :D Otherwise i‘ll check on saturday once i‘m home if any comfyui or node updates broke anything 🫡
vslinx Okay, I found the problem with the checkpoint. The denoise was set to 0.6, which Wai Illu doesn't like. Denoise at 1 has no problem. I solved the scheduler problem by extracting the scheduler from the sample and using it at the moment. This might help other users who want to use your workflow in the future.
Thanks a lot for your help! If you ever need someone to beta test your new workflow, let me know. ;D
thanks for the feedback after testing Otaku2D !
I usually always keep the denoise at 1 so i never faced this issue before :D but i guess that issue was something WAI Illu specific. I might throw this in the FAQ‘s somewhere!
I‘ll check what the other issue is, once i‘m back home on saturday and then publish a fix if it‘s a common issue :P
I‘ll gladly take you up on that offer once i am about to release a new version! haha :D
Hi vslinx,
Very cool, thank you very much :D. I'm looking forward to the new version ;D.
It's really strange with the scheduler, especially since you're using the core version of comfui. Well, you'll find the error.
Until then, have a nice vacation! :D
Hey Otaku2D,
i just came home and updated all my instances and went right to checking :D
I have tried the freshly downloaded workflow on two instances, one with my 2 week old build and one fresh build.
The old build had Comfy Version 0.3.43, Frontend Version 1.23.4 and Manager version 3.33.4,
then i also tried it on the newest version with Comfy Version 0.34.9, Frontend Version 1.24.4 and Manager Version 3.35.
I could not reproduce this issue with the scheduler selector, neither with sgm_uniform which i usually use, nor the karras scheduler that you've mentioned.
Could you update your comfy + the nodes and see if the problem still occurs with a freshly downloaded workflow? It might've been fixed by whatever was causing it, though the image-saver node was updated 2 weeks ago so i doubt that.
If it still occurs could you send me a log or screenshot of the error in your console of your comfy?
The only other alternative is that you may be using the wrong (old) image-saver repo.
The old image-saver custom node was by the user giriss which disappeared all of a sudden and left this old repo behind: https://github.com/giriss/comfy-image-saver
A lot of people still have this old version installed that is not getting any updates anymore.
Alexopus branched off of this and created a new node with (sadly) the same ID, so you won't get an error initially if you have the wrong one installed. The node i've linked in the description and everywhere else is the new one that is getting regular updates here:
https://github.com/alexopus/ComfyUI-Image-Saver
maybe check which one you have installed and switch if you're using the wrong one, otherwise i'm pretty sure the problem only occurs in your installation as i could not replicate it :/
Hi vslinx,
Thanks for testing it. I'll download the new workflow right away and then test it.
I'll write down all the individual versions for you if it still doesn't work.
You really know a lot about ComfyUI, that's really impressive :D
heya Otaku2D,
alright! thank youf or letting me know in advance :D
Well it's a love hate relationship 😂 all i know from comfyui is from running into problems with it, so it's probably not that much! but thank you nevertheless :D
Hey for the scheduler pass lot of user will have this error use SDAnyConverter between the scheduler and the ksampler and it will work simple hack.
Hi iPiKo,
I reinstalled ComfyUI and the workflow works well now. The problem was that my ComfyUI installation was a portable version with many unused custom nodes. I think my nodes were outdated and unsupported. Have you updated your ComfyUI and nodes?
Otaku2D Yes I have latest comfyui the only way to make your workflow and any different workflow that use a global scheduler is to use that trick.
iPiKo that‘s not entirely correct. It‘s most likely caused by outdated nodes/incompatible nodes you have installed. I have multiple setups and even with a global scheduler it works without problems since there‘s a scheduler converter between the scheduler and the detailers. And i‘m running the most recent comfy (0.35.0) and the most recent frontend update. It could also be that you have the wrong image-saver node installed. There is one that is deprecated and hasn‘t been updated in 2 years and therefore does not support most schedulers. Make sure to have the right one installed.
vslinx I had to disable all plugins and manualy delete the leftover for it to work, so my guess is there is some sort of conflict.
iPiKo That's weird :/
I do have the custom_nodes list of Otaku2D, if you want to share yours we can check which one's you have in common and maybe find the reason/the node that is responsible.
I'd assume it's a node that has a conflict with the image-saver node (that's where the scheduler selector comes from).
@iPiKo I had the same issue, your "hack" solved it. Thnx!!!
@bakker_be fyi if anyone is having the issue of:
"- Return type mismatch between linked nodes: scheduler, received_type(['simple', 'sgm_uniform', 'karras', 'exponential', 'ddim_uniform', 'beta', 'normal', 'linear_quadratic', 'kl_optimal', 'ays', 'ays+', 'ays_30', 'ays_30+', 'gits', 'beta_1_1', 'bong_tangent']) mismatch input_type(['simple', 'sgm_uniform', 'karras', 'exponential', 'ddim_uniform', 'beta', 'normal', 'linear_quadratic', 'kl_optimal', 'ays', 'ays+', 'ays_30', 'ays_30+', 'gits', 'beta_1_1', 'bong_tangent', 'beta57'])"
That means you're having more sampler choices in your sampler that the save node would expect. This is being caused ONLY by the ClownsharkSampler-Custom Node because it is overwriting the default comfy options which is incredibly bad practice for a developer and nothing someone should ever do. So if you're not actively using it i can only recommend deactivating or uninstalling this node. Otherwise there shouldn't be any nodes that should cause issues with a global scheduler. But if using the SDAnyConverter-Method works for you then that's obviously also fine :) Just wanted to leave this here for anyone possibly facing the same issues.
@vslinx I do indeed have those installed. I'll look into why, and see if they can be replaced in the workflows using them.
Thanks for getting back on this!
[SOLVED] The workflows are broken on recent ComfyUI due to frontend changes, according to https://github.com/chrisgoringe/cg-use-everywhere/issues/349
I had to manually update the cg-use-everywhere node. Everything works as expected again.
<ah shit here we go again meme insert>
Thanks xnapx for the headsup! was about to release 4.3 with hires fix, guess i'll implement the changes so that it works first.
vslinx Sorry to be the bringer of bad news. o7 Better confirm for yourself, though, before acting!
What I'm getting on either of your workflows is this monstrosity:
Loop Detected TypeError: Cannot read properties of undefined (reading '0') at http://localhost:8188/extensions/cg-use-everywhere/use_everywhere_utilities.js:204:81 at Array.findIndex (<anonymous>) at find_all_upstream (http://localhost:8188/extensions/cg-use-everywhere/use_everywhere_utilities.js:204:48) at recursive_follow (http://localhost:8188/extensions/cg-use-everywhere/use_everywhere_utilities.js:236:5) at recursive_follow (http://localhost:8188/extensions/cg-use-everywhere/use_everywhere_utilities.js:229:16) at http://localhost:8188/extensions/cg-use-everywhere/use_everywhere_utilities.js:238:17 at Array.forEach (<anonymous>) at recursive_follow (http://localhost:8188/extensions/cg-use-everywhere/use_everywhere_utilities.js:236:45) at node_in_loop (http://localhost:8188/extensions/cg-use-everywhere/use_everywhere_utilities.js:259:18) at GraphAnalyser.analyse_graph (http://localhost:8188/extensions/cg-use-everywhere/use_everywhere_graph_analysis.js:127:17), undefined
it's okay xnapx :D i already had another user report it but he also said that he had some outdated nodes, so i thought it might be an incompatibility issue.
I've tried to replicate it now with use-everything on 7.0.1 and couldn't, at least not with my new 4.3. Can you let me know if you're getting the same error when using this version?
https://huggingface.co/vslinx/ComfyUIDetailerWorkflow-vslinx/blob/main/workflows/TXT2IMG/v4.3/TXT2IMG-ADetailer-v4.3-vslinx.json
(haven't released 4.3 yet on civitai as the site is running into issues rn)
vslinx The issue was ME! Well, it was Comfy because it didn't pull the most recent use-everything, apparently. After manually pulling 7.0 from git, it works like a charm.
Apologies for the scare and thank you so much for your patience (and great workflows! I like the additions I'm seeing in 4.3!)
xnapx aaaaaah! :D
I guess they did some changes for the new subgraphs at 6.3 and then implemented a backwards compatibility fix at 7.0 :D
Sometimes comfy misses these updates and it sucks :') but at least it's solved now! :D glad you like the changes ♥️🫡