Quoted By:
I have a question!
Well, is not an specific question but a general curiosity,
When we use the stable diffusion to make lewd anime girls, it drinks from the boorus, righ? The entire prompting system is done around the boorus tags,
But when you use stable with hiperrealistic models and such...where does it drink from? Because is not as if the boorus have lot of that, if you want to recreate an scene from lords of the rings or make an AI influencer, where the ai gets references to work with?
I also noticed in prompts the way of prompting for that sort of thing is completely different to anime, like more descriptive, Like "gandalf as female" or "ironman armor in the bottom of the sea rusting" is not tags like with anime.
That's my "question" basically, just curious about how stable works with "realistic" stuff and where it drinks from
