Hey guys, how's it going? I'm back (for now) with a new post after months.
I've been testing ComfyUI with the Anima models, and I have to say I love how the images turn out with that model and using ComfyUI. So, using what little I know about this Stable Diffusion interface, I created a fairly simple workflow that works quite well for me to generate high-quality images.
This workflow, as I mentioned before, is very simple, and that's because it's designed for people like me who haven't used ComfyUI much and find all those nodes and connections overwhelming. This simple workflow could help make generating images in ComfyUI feel less like a medieval crusade (laughs).
To make the workflow work optimally, you'll need to install the ComfyUI custom nodes, as some components of the workflow require them:
https://github.com/ltdrdata/ComfyUI-Impact-Pack
https://github.com/kijai/ComfyUI-KJNodes
https://github.com/ltdrdata/ComfyUI-Impact-Subpack
https://github.com/yolain/ComfyUI-Easy-UseI hope this resource is as useful to you as it was to me.
For the 1.5 version of this workflow, you should install this custom node below if you want to use the signature apply function.
https://github.com/vergil-sparda0701/comfyui-apply-signatureDescription
Well, here's a slightly more updated version of my workflow for creating images with Anima models in ComfyUI. I've added a few extra nodes to improve image quality, as well as a custom node to add a signature to the images in case you (like me) want to add a custom watermark to your images.
Here's the custom node for the watermark application
https://github.com/vergil-sparda0701/comfyui-apply-signatureFAQ
Comments (2)
I love it, it's really easy to understand and 0 errors. TYSM. I have a really noob question tho, what is the prupose of having 2 ksamplers next to the other? I'm afraid to touch those lol.
According to my research while updating this workflow, the second sampler is for further refining the image, so it comes out with fewer errors, and from what I've noticed, it really works since the images come out with less distortion.