Let’s make a picture of a stylish elven man perched in a tree high above an ancient forest! Here’re the initial txt2img generation parameters, in Stable Diffusion Webui: Prompt: an elven warrior crouching on a huge tree branch, white hair, looking over shoulder, looking at camera, forest far below, towering trees, (moss:0.7), ornate clothes, ancient ...

I worked on a headshot of an android with Stable Diffusion Webui. I wanted it to look quite human but with some features that were a bit off, like panels and plastic-y skin (think Data from Star Trek: The Next Generation). So I started with a regular person. The setting is at night in the ...

I wanted to create a more emotional image with Stable Diffusion Webui and thought of a sad girl sitting at the edge of a dock, with her feet hanging down and a sad expression. Usually it’s hard to get Stable Diffusion to compose both the scene and a figure together in a very specific way. ...

I was experimenting with a more flat, animated style in Stable Diffusion using various models. Clouds are fascinating and storm clouds are even better. Add in some rugged mountains and a white castle, all at sunset, and that sounds like fun! After some tinkering with models and prompts I founds a generation batch that looked ...

I love rugged terrain so I thought I’d try generating a castle built into a towering mountain, going for an angle looking down from high elevation. I figured Stable Diffusion would struggle with the concept of both the castle and the mountain and the angle so I started with generating a number of images of ...