Color Trace is a style used in the pre-production stage of japanese animation. The tag itself already exist in Anima/Noob/Illustrious, but the results are very heavy with the pencil coloring. I trained this lora using color traced images with more of a digital flat coloring style.
To get the results you see in the images above, these are the settings I used.
Visit My Site For Models I Can't Post On CivitAI: https://qtgurl.com/
Description
FAQ
Comments (2)
I feel like the anima one is better than illustrious. Dp you think its because of your prompts or is it because of the differences between the models?
It's because of the difference in models. illustrious (aka SDXL) is a really old and limited model. Due to SDXLs small 4ch VAE, images never come out accurate. Due to it's ancient text encoder, prompts are never fully followed. Anima on the other hand can produce detailed images with proper color due to it's 16ch VAE, higher resolution images by default and understands natural language much better than SDXL. Lastly, the lora training replicates style extremely better. I can reproduce an image that looks 95% like the training material because of how well style is learned through Anima. There's really no reason to keep using SDXL if your goal is to create the best looking anime images.