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
First time writing a guide for a workflow, you can find it here.
The workflow explains multiple different scenarios for what you can use it as well as explaining every node and group in depth with some extra information if you want to spend the time reading it!
IMG2IMG Transfer for copying image composition
re-arrangement of nodes
compatibility updates
release took a long time because i couldn't stop downloading images from civitai and danbooru and making them my own with my favorite characters and styles... hope you enjoy it as much as i do
FAQ
Comments (10)
Hey @vslinx!
Dude, this is incredible! I've been using it pretty much without issue for a few days.
However, now I'm running into this weird problem where if I turn on the detailer, it turns that area black. (if i select the hands detailer, the debugger shows that it's finding hands, but they are now black ... just black hands, the preview shows the updated image, but the subject now has black hands).
This is happening for all the detailers, i'm not sure what I am doing wrong, I even reloaded the workflow from scratch to make sure I didn't change a setting or accidentally move something around.
Any idea what could be happening?
Heya @coogins,
glad you've been enjoying the workflow! 🥰
That sounds weird. Do you see just a black 64x64 pixel block in the Debug section or the actual body part that's supposed to be inpainted but completely in black?
For me it sounds like it might be a node issue, do you have all of them completely up to date?
It could also be that your comfyui has something cached weirdly or some connection has popped off, maybe one of the denoisers? Could you save the workflow (in it's broken state) and upload it to something like mediafire so i can download it and check if i can see what the issue is?
Hey TTeslaThunder,
sadly the node publisher of the Image Saver node made a mistake that broke workflows because he had a duplicate name for the same node. He has fixed that in a new version release 3 days ago, but it does not seem to have fixed the already broken update released before that.
I'm currently creating new screenshots for a new version release and have already finished my new version for the workflow with this fix included. You can already download it here if you already want to use the fix:
https://www.mediafire.com/file/yhdqabr5lqeyrjk/ADetailer-v4.2-vslinx.zip/file
vslinx thank you
Nice workflow. If I may suggest a few small improvements, I'd add an optional sage attention patcher node (from kjnodes) as well as replace the upscaler solution you have with a more versatile upscaler node like CR Upscale Image (from comfyroll). Being able to set the rescale factor is absolutely essential when upscaling. I'd also recommend using setter and getter nodes (also from kjnodes), which will significantly clean up the spaghetti - trust me, once you get used to them, you'll never want to not use them in complex workflows. I read in the other comment that you're already working on a next version, so I thought I'd share these little tidbits, while you're already updating the workflow anyway.
hey xnapx,
thanks for the feedback and the recommendations!
The addition of the Upscale factor is defintiely something i'll add, sadly the CR Upscale Image node currently does not accept an upscale model as an attribute input into the node itself. So you can't just connect a "load upscale model" output to the input of that node. Since i want to keep the workflow user-friendly for non-technical users i'll probably open a MR later on the Node itself and see if it gets merged and then implement it, but it's a great addition!
The Sage Attention Patcher node is a good idea too and really uncomplicated to implement, so i'll probably release and implementation for it in the next version already!
Getter & Setter nodes is something i "sadly" won't do. I think it's fine to use it with smaller workflows but i've had some issues with them in the past. Once you have a workflow of the size this one currently is it becomes somewhat hard and unnecessarily convoluted to not just keep track of your nodes but also about variable names. I have a lot of people who customize the workflow for their own needs and adding another layer of abstraction to the workflow would greatly diminish the "easyness" with which you can pick this up and change things. I also had some comfy and kjnodes updates in the past that have moved the getter and setter nodes into the front so you also can't hide them neatly behind others and just add quite a few nodes when toggling the link visibility off would just fix the issue.
That was a lot of blabla but i hope you understand where i'm coming from and why i won't add these nodes sadly :D
But thank you for the feedback in general and i'll 100% add the other 2 suggestions!
vslinx Fair point about get/set nodes. While I personally prefer scrolling through the dropdown list over tracing the spaghetti, I can see the counterpoint of added abstraction - (to which I would respond that clear constant naming mitigates a lot of the argument, but I'm not here to change your mind.) - It's a style choice and I appreciate and understand your explanation of why you opt against that. <3
Side note: I was expecting more skepticism about the SageAttention node in a t2i workflow, since it requires some involved prep to be able to use and also only grants minor benefits in terms of performance. Personally, I will die on the hill that gaining even 0.1 % of VRAM in order to squeeze out the max resolution or be able to use just one more lora is always worth it. (The actual VRAM benefit in a given t2i or i2i workflow is closer to 1%, for full disclosure.)
Keep up the great work!
xnapx glad you understand :D i think it's a preference situation anyways, i don't see much of a benefit for me personally so it's not really a priority ^^ rather spend the time on adding features then :P
Yeah i noticed that the benefit is somewhere between 1-4 seconds per image depending on wether i use img2img or txt2img, but even these few seconds stack up - especially once you start detailing. I'm not really against it since i'll leave it to the user, deliver it by default with the disabled state and will make sure to add information inside the workflow and the guide on what it does and where to find a manual on how to install triton etc.
The benefit might be miniscule but it doesn't cost anything to add ^^Especially since the little extra boost of performance might help people with lower VRAM to use stuff like IPAdapter which is usually hard to use for GPU's under 12GB.
Thanks again for the recommendations! :D
xnapx i'm happy to announce that your recommendations are in the new version :D
I submitted a pull request for comfyroll to change that the rescale factor node also accepts upscale_models as a node input. But sadly after a bit of digging i found out that the person behind the node (Suzie1) has been inactive for over a year and that's why the node has not received any updates over the last year. Seemingly they have pruned all their info's 3 months ago on their personal github too, so i had to swap to another node (wlsh_nodes) that offered what i needed and that i already had some experience with and it works well :)
Thanks again for the recommendations, i thought they were really helpful!