š ļø 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.
ā ļø 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.
š¤ 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
This is just v13a with Epsilon Scaling added as an option.
Several markdown notes have been added regarding how it works and interacts with the workflow thanks to ChatGPT.
I do not use it across the entire workflow, but it is available to use where it is applicable.
This requires your ComfyUI to be on version v0.3.64. Upgrading your comfy install comes with its own risks and I am not responsible if your comfy install breaks. Do so at your own peril.
I want to reiterate: the detailer nodes will have an error if you do not update Impact Pack after updating ComfyUI to v0.3.64
FAQ
Comments (22)
Hi, I like your work. I'd like to suggest the z-tipo-extension extension as an additional feature.
Thanks! Always open to trying new things but..
NGL, I tried messing with z-tipo with another user of my workflow and 100% hated it. That's pretty much what led to the wildcard portion being added.
Hey! Thanks for the workflow, it gives amazing results out of the box, that's amazing!!
I had a question tho, I dont get the wildcards to work?
I enabled the wildcards in the options, i add some prompts including the "__myWildcard__" but it's like it doesnt go through, the combined positive prompts will always be the simple prompt i have
(when i create a new workflow with just the impactwildcardProcessor it works fine)
Anything I'm doing wrong?
Thanks again, love your work :)
Hi, let me test on 13b and I'll see if I connected something wrong. I was pretty tired when finishing it up. I'll get back to you shortly.
On my first try, it didn't add the wildcards, but on the 2nd (and after) it did.
Testing this led me to see that the results are coming out blurry on the 2nd USDU. Time to go back tot he drawing board.
@GladasĀ Haha nice! Thanks for having a look (and so quickly!)
Hope you can iron out those kinks
But yeah again, stellar work!
Regarding 13b:
I am getting some blurrier results with it today on the 2nd USDU.
I plan on working off of 13a and just adding Epsilon Scaling to it and see how it goes.
Maybe deleting the 2nd Upscale before USDU is a factor too.
Tested Epsilon Scaling with 13a.
Inserted the Epsilon Scaling node between Dynamic Thresholding and CFG Zero Star.
Initial testing using my Oshino Shinobu standing water in a red swimsuit image has yielded a comparable result without the blur I saw in 13b. Going to try on a few more images before cleaning things up and releasing as 13a-Epsilon
Hey man!
Is there a reason you are using a "impactWildcardProcessor" instead of a impactWildcardEncode"?
I find it slightly better, as it allows loras inside the wildcards, which can be crazy useful
I just did the swap and it works pretty great
Yeah, too many people had problems understanding how to use it. This was probably one of the top things asked back when I still had it in the workflow. I had many DMs specifically asking how to use LoRAs. I found it easier to just swap to what I have now.
My DMs are free from that question!
@GladasĀ haha makes sense, thanks!
2nd USDU Upscale Model:
I started using Ultramix_Balanced as my second upscale model. This works pretty well if you're running into the 2nd USDU result being over-sharpened. This boils down to preference of course.
Can you add lora manager lora loader?
The node I use: Power Lora Loader is what works with the Image Saver node thanks to someone who made the custom node to connect the two. Otherwise, it would be ImpactWildcardEncode instead leading to someone asking me how to use loras with it everyday. This is also due to the fact that rgthree is already part of my workflow and is where Power Lora Loader comes from.
It should be pretty easy for you to swap in, if you really want to use that node since it is at the start of the workflow.
I'll take a look at the node later, but without looking at it right now, the answer would be "no".
Looks pretty interesting, but definitely will come with its own influx of questions on how to use it.. I'll try it out on the next update. If questions flood in on how to use it, then out it will go.
Actually, nope this is going to be a hard no. The browser extension is the only thing that makes it useful and from what I am reading that function is limited use after the "trial". This will be a hard pass. I'm not keen on adding something to my workflow that requires people to pay to use.
As I mentioned earlier, this should be easy for you to add if you want to use it with my workflow.
In the end, I designed the workflow my personal use. While I appreciate suggestions, this is definitely a hard pass.
@GladasĀ this is a great workflow. I just wanted to chime in about the lora manager stuff. The only thing that is paid is the extension. Lora manager itself is completely free.
I am not asking you to add it to your flow, I just dont want anybody reading your comments to be put off of a great tool incorrectly.
Keep up the great work, I check out each update.
@2018cfhĀ fair enough.
It's only the extension part (which I did say is limited use on the free trial) that made it look good to me when watching the video on it. Maybe I didn't catch the part where they say that on the video (if they did at all). I only noticed it when I had already installed the custom nodes and went to install the extension.
For anyone wanting to add something like Lora Manager on my workflow:
1. Just take the inputs from the Power Lora Loader and connect them to the inputs of Lora Manager.
2. Connect the outputs from Lora Manager to Power Lora Loader's input.
3. Delete Power Lora Loader and the connections to the reroutes will now be connected to Lora Manger.
My reason for choosing not to use it: basically, I felt like I got baited into adding something on to my workflow (intentionally or unintentionally) that could potentially require people to pay for a feature. I want my workflow to be 100% free to use. Sure, the extension isn't part of the workflow when using Lora Manager, but it essentially gatekeeps the only function that made it appealing to me in the first place.
@GladasĀ ty šlike I said, I wasn't knocking you at all. Appreciate the added comment. Ive been using the Lora Manager for a couple months now and ironically the extension is the only part I didnt really need. Lol
@2018cfhĀ it's all good.
It's one of the things I get kind of heated about when it comes to what I do with comfyui and AI in general. It's 100% a hobby for me. I'm just a random degenerate weeb who wants to generate nice images of my waifus.
If I were to take money or involve something that potentially takes money from people who use my workflow, then I will feel obligated to treat it as a job or service, which I do not want to do.
I'm not knocking people who do make money from anything AI related, it's just not for me.
Added some notes to the "Model" details about a Power Lora Loader function that I was not aware of:
It can pull the data from civitai (if it exists) and get the trigger words for you to copy/paste. Nice and free QoL feature.
Thanks again G, love your work!