In this 2 part tutorial I walk you through my process of generating characters using Stable Diffusion and Automatic1111. In Part 1 we use the A.I tools along with Photoshop and Inpainting to tweak the design and prep the character for animation. In Part 2 we take those elements into After Effects to rig, animate and finalise the composition.
I’ve taken advantage of the ProtoThing 200 model which I sourced from Hugging Face. Hugging Face is a great website with hundereds of models created by the A.I Art community. Just a heads up, if you’re considering using a local installation of Automatic1111 you’ll need a graphics card with at least 4GB of VRAM, the more you have the better as it can be quite taxing on your machine. Unforunately Mac users will have to wait as well as it is currently not supported.
I’ve had no issues running the software but have a fairly decent system due to the nature of my work. I’m running it on Windows 11, i9-12900K 3200 Mhz 16 Core Processor, 64GB RAM with a RTX 3090TI GPU.
There are a couple of things I need to iron out with the workflow. I still don’t like a lot of the A.I masking tools I’ve tried, just doesn’t seem to grab the fine details all that well which is why I’m still more likely to use the pen tool unless the rotowork can be slightly scrappy.
I’m really happy with how this one turned out, very nearly feels like a splash screen you could find in a videogame. If you’ve found it useful or have your own workflow ideas be sure to get in touch in the comments!










