SDXL Best Workflow in ComfyUI

Embark on an exciting journey with me as I unravel the unique workflow I've developed to utilize Stable Diffusion SDXL within ComfyUI. Notably, this workflow has been made possible through a judicious use of CLIPTextEncodeSDXL, CLIPTextEncodeSDXLrefiner, and the WAS suite. Having extensively experimented and honed this process, I'm thrilled to share my expertise and practical insights to expedite your design endeavors.

In this video, I walk you through each step, meticulously breaking down the techniques to aid your understanding and application of these advanced tools. Irrespective of your familiarity with AI-driven design, this comprehensive tutorial promises to add value to your creative journey.

#comfyui #stablediffusion #sdxl #tutorial #workflow

[bgtubber]

Thanks for explaining this. You make some of the best Stable Diffusion and Comfy UI videos on Youtube!

[PawsitivelyTech]

Great results! Details like fingers and facial symmetry came out really nice!

[AlmazovS]

Hi! Good results! What do you use for text concatenation?

[justme]

I got the WAS Suite but when I attempt to use the Text Concatenate node it won't accept any text and the output says ASCII instead of String. Any idea what I'm doing wrong?

Edit: Looks like I can use a text to string node, but still a bit confused why my outputs are stuck to ASCII. Seems an unnecessary extra step.

[Enricii]

Best channel for ComfyUI workflows, please keep it up.
Does it ever happen to you to get always the same face/model when asking for woman photo? Even changing seeds? Any suggestion to avoid this problem?

[atanekoatan]

Now I understand why people stick to Midjourney :-)

[rssgtzk]

Can I ask what exactly CLIP dimension means?

[generalawareness]

I tried this and in 1.0 ComfyUI gave errors for it so it really didn't like it.