From the course: Learning ComfyUI for Stable Diffusion

Unlock this course with a free trial

Join today to access over 24,900 courses taught by industry experts.

Tuning ControlNet parameters

Tuning ControlNet parameters - Stable Diffusion Tutorial

From the course: Learning ComfyUI for Stable Diffusion

Start my 1-month free trial Buy for my team

Tuning ControlNet parameters

“

- [Instructor] In the last movie, we applied control net conditioning to mix a prompt with a simple line drawing to direct the composition. The result that we got with the default parameters for the apply control net node are not really satisfactory because I'm getting a very precise rendition of my bad drawing of a hummingbird. I'm not an ornithologist, and I barely even know what a hummingbird looks like, so my drawing isn't anatomically correct. But with those default control net parameters, I get a version of a hummingbird that follows the contours of my scribble drawing, but doesn't really respect the actual reality of a hummingbird in terms of its overall proportions. For example, the wings are way too thick here. They need to be slender and more elegant. We can help this along by changing up those apply control net node parameters. Get in close on those. The strength is obvious enough. That's how much power we're going to impart from the external image. Start percent and end…

Contents