Home Creators Posts Import Register Favorites Logout

Content

It's really hard to get white antlers. Generally she was a harder one to get some level of consistency.

I use NovelAI and if you want to make your own Bai here's what I used. Of course adding NSFW would make some naughtier images. Also you can remove photo realistic to get a flatter anime style. Or you can add 'shiny skin, shiny hair' and get more of an ecchi anime style.

Bai

tall female, fur clothes, blue and white dress, gold designs, {{white}} antlers, skin tight, masterpiece, highres, photo realistic, ice crystals, white hair, cold, sitting on throne, arrogant, looking at viewer, proud, intricate details, earrings with rubies, necklace, rubies, thighup, mature woman,  cinematic lighting, whispy clothing, sheer clothing, perfect lips, gold eyes, snake eyes

Negatives: ugly, duplicate, morbid, mutilated, out of frame, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, ugly, blurry, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, out of frame, ugly, extra limbs, bad anatomy, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, mutated hands, fused fingers, too many fingers, long neck, fat, low quality, jpeg artifacts, signature, watermark, error, text, lowres, deformed, legs, 2girls

Files

Comments

Brad

Like 1,2,3

Yawn

1,2, and 3 are my fav

Jamie R

1,2 and 5 are my favourite Followed by 8. But the prompts are good to know for the AI artwork. I wonder if they'd work on Midjourney? Friends and I have a trial to that for some D&D character profiles

Bruce_Sentar

They probably won't work as well on Midjourney. It and Stablediffusion seem to be trained a lot on paintings. A lot of dead painters who's work is in the public domain. NovelAI and Waifu Diffusion(a stable diffusion mod) are trained on Danbuuro/Gelbuuro. Repositories that are largely anime and fan art.

Yitzhak Brill

Having had my own experience fiddling about with AI image generators, I know how much hard work it can be to just get a picture to turn out right. Eventually the AI can get it, but so many images that are churned out in the process are abstract or pure nightmare fuel. It takes a lot of prompts and refinement for the AI to generate an image that looks good, so I’m glad to see you’ve been able to have some success with it! Plus, it’s a quicker way to get potential character images while you wait on commissions from actual artists. I don’t think anyone of us here will complain about seeing more pictures of our favorite ladies either 😁