⚠️ NOTICE ⚠️
This workflow has now morphed into the UniFlex-Workflow 9.0!!!
Description
See workflow description for significant changes.
FAQ
Comments (8)
seems like a lot of those nodes are broken, don't work with the latest comfy ui, or something. I wish I understood the purpose of the nodes so i could change them.
Can you be more specific? Which workflow are you running?...try the core or light version 5. Have you properly installed all of the custom nodes? What does your comfy boot log indicate?
@kaptainkory I must apologize, it was late and I was cranky. I had a moment.
I run my comfyui set up in a docker container on Arch Linux using CachyOS and I manage my nodes with the cm-cli.py commands, so there are very few people who could or should provide me with any technical support.
With that said, the initial issue was with the nighly version of was node suite and kjnodes but that was resolved after some tinkering, but the golden noise node seems unresponsive to git pulls within cm-cli.
Typically I never have an issue with any nodes due to the genius of the author of the docker image that I got (shout out to jimlee2048) but as far as the "purpose of the nodes" its just me being a noob and not going through the workflow.
If you read this thanks a lot for your hard work and please let me know what the advantage of the golden noise node is and if its worth pursuing to get it working in my set up?
Thanks a lot.
@xobythepoet776 hey, no worries. Welcome to the current world of Comfy, where instability rules! HA! I realize the full AIO version of this workflow is a bit much—maybe overwhelming for some—to get up and running, but I have tried to incorporate mostly popular or fairly updated nodes...and make sure it all runs on my standard-ish setup. That said, I think the Golden Noise node will be going away in the 5.1 version release. You can check more about it at the link below, but, yeah, I don't think it does all that much...and there isn't even a Flux-specific weights release yet.
https://www.reddit.com/r/StableDiffusion/comments/1h8islz/golden_noise_for_diffusion_models/
I have to ask, how do I know if the extra samplers are actually working?
For example, in group e3 controls, if I wanted to test the sigma samplers then naturally I would drag the right side of the node, the sigma output from the Flux Setting Node 2, into the left side of some part of the sigma sequence, prior to the left side of the set sigma node.
I guess I am asking where the proper placement of some of these extra samplers and such would be. Thanks again.
You can try something like this:
1) Run the default workflow [image 1]; without the 2nd sampler (03e) enabled.
2) When you get a result you like, drag the image into Comfy to open the saved/embedded workflow.
3) Bypass the global seed node (00 Switchboard group) to fix the seeds.
4) Enable the 2nd sampler (03e).
5) Run again [image 2].
6) Now, "loop in" one of the detailer nodes, let's say the Detailer Daemon to apply as a final fine-detail enhancer: disconnect the noodle from the Set_Sampler 2 node, connect it to the Detailer Daemon node, then bring it back to the Set_Sampler 2 node. If you set the start to 0.5 and end to 1.0 on the Detailer Deamon node, it should just apply to the low sigma processing (which applies to the final, smaller detail rendering).
7) Run again [image 3].
Now, you should have three images to compare. By this example, the results might be kind of subtle (or maybe not?). On the Detailer Daemon node, really bump up the detail_amount and/or exponent until you get noticeable changes...then back it off to taste.
You can experiment with the other available nodes using a similar approach.
.....
Check out XnView for a nice image viewer program that has a Linux version. ;-)
Really I have to give you more buzz because I rushed through your article without reading it entirely and I have to give you respect and credit for your work. This is a really cool set up and you clearly have all the information available. I will definitely have more questions though. Thanks a lot.
I appreciate it! :-)))
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.