1/3
1 Year of Service
For me, the process usually starts with one or more images from a base model (I generally dislike 'fine tuned' models), then some combination of manual editing, inpainting, outpainting, blur suppression, denoising, and upscaling. And I, almost always, use more than one model alomg the way. I've actually gotten into it with people that want my prompt(s) so that they can produce their own variants of my work... I'm always trying to explain to them that there is no single prompt to generate my images.I think it’s art if you’re trying to incorporate different tools in the making. Different LORAs, checkpoints, prompts engineering, upscaling, in painting, those are just a few tools you use to produce the desire images. Pretty dang hard and complicated if you ask me. So imo if a person have to think and tinker the AI engine to make the art they desire, not just any random, then it’s art. Of course I might be bias…