HOME | DD

Description
This piece was a test to make an orc warrior. I'm not entirely pleased with this one even after dozens of tweaks and a lot of Gimp-work, but it was good enough to hold onto. She was to be a character in a pathfinder game that never happened.

I've gotten into AI Art with 2023. Well, I make my art in a combo of Stable Diffusion and Gimp. Going back and forth between the two until I get it right. This is an earlier work where I knew a lot less than I do now. However, my general process is to spin out a few dozen images in Stable Diffusion, take some of them into Gimp and mess around cropping things together, removing bits, drawing in a few elements, and so on, then take it back to Stable Diffusion's Img2Img tool that makes new art based on a specific image you give it. I then cycle back and forth between that and changes manually made in Gimp until I have the piece I want.

Throughout this process I liberally change 'ai prompts' and 'training models', and more between each render so the final result is very much a personally crafted piece.

People dismiss AI Art right now, but I feel that once people understand it more, and learn to work with it in combination with other tools, it will get the respect it deserves, just like Photography once had to do.

Description
This piece was a test to make an orc warrior. I'm not entirely pleased with this one even after dozens of tweaks and a lot of Gimp-work, but it was good enough to hold onto. She was to be a character in a pathfinder game that never happened.

I've gotten into AI Art with 2023. Well, I make my art in a combo of Stable Diffusion and Gimp. Going back and forth between the two until I get it right. This is an earlier work where I knew a lot less than I do now. However, my general process is to spin out a few dozen images in Stable Diffusion, take some of them into Gimp and mess around cropping things together, removing bits, drawing in a few elements, and so on, then take it back to Stable Diffusion's Img2Img tool that makes new art based on a specific image you give it. I then cycle back and forth between that and changes manually made in Gimp until I have the piece I want.

Throughout this process I liberally change 'ai prompts' and 'training models', and more between each render so the final result is very much a personally crafted piece.

People dismiss AI Art right now, but I feel that once people understand it more, and learn to work with it in combination with other tools, it will get the respect it deserves, just like Photography once had to do.