[Return]

Report a post

Preview
>>14257
I think my prompts are rather straight forward. I still use a tag-like approach.
This is the initial prompt for my last work.

3girls, furry, brick wall, (blindfold:1.5), arms behind back, kneeling, tears, crying, fear, afraid, panic, closed mouth, sonic style, full body, <lora:Acme Beam Gun_Illustrious:0.8>hand gun, pov, muzzle blast fire, shooting at forehead ADDCOMM
amy rose, red dress, face down on floor, lying on floor, lying on stomach, dead, lifeless, unconscious, tremble, collapsed ADDCOL
blaze the cat, dress, head exploding, blood splatter ADDCOL
rouge the bat, black bodysuit, cleavage

I used a regional prompter with 3 columns. As checkpoint I used novaFurry_Illustrious. It already has many sonic characters trained. So I don't have to use to many extra Loras, which often leads to strange behavior.

This created the first picture in this collage. I still had to add and fix some things.
Post number No.14381
Board Artificial Intelligence
Optional. Describe what's wrong with it.