Reply to thread

I've never actually tried DreamShaper before, but NeverEndingDream was made by the same person, so I assume they're rather similar. I've usually done text prompts then sometimes doing image to image with those results. Lying down, lying on floor/ground, eyes closed, sad usually gets a good result. Sad is actually one of the core tags as their faces get downright goofy without it.

Its also worth noting that NED was actually the easiest to produce unconscious/death poses so I'd imagine DreamShaper has similar ease for it.

Fighting Stance is a tag I've used a lot too. Dynamic pose sometimes for the harder to budge models. I've figured out the Kenshi model a bit more and I'm confident in thinking now that it generates fighting poses the easiest.

Also, on design, I just do what I do for non-fetish purposes and generate characters based on vague prompts, taking designs I like, and start tagging more to replicate them consistently. That example is probably the most complicated one I tried and its only because it came off of CetusMix, which loves doing intricate details without being asked to.