sexual situations

physical violence

disturbing

male nudity

extremist

hanging

hate symbols

nazi party

revealing clothes

weapon violence

female swimwear or underwear

male swimwear or underwear

partial nudity

white supremacy

adult toys

graphic male nudity

illustrated explicit nudity

nudity

self injury

graphic violence or gore

graphic female nudity

corpses

wide hips

convenient censoring

peeing

oral invitation

emaciated bodies

exposed female nipple

blowjob

female nudity

sexual activity

sexual intent

undressed

male underwear

female swimwear

genitals

female underwear

thick thighs

breasts out

strapless leotard

vore

bestiality

breast out

one breast out

huge breasts

gigantic breasts

huge butt

covered nipples

hair over breasts

no panties

sitting on face

anal

dildo riding

downblouse

oral

porn

futanari

hentai

nude

lingerie

sexy

nsfw

suggestive

bukkake

cumshot

cumdrip

cum in pussy

cum on face

cum on hair

cum on body

cum on tongue

cum on hands

cum in mouth

cum on pussy

eat_cum

fellatio

implied fellatio

after fellatio

triple fellatio

autofellatio

pov fellatio

fucked silly

presenting ass

presenting pussy

spreading legs

spread legs

swimwear

Simply put, this is a ComfyUI workflow to batch image-to-image process rendered scenes from something like a Ren'py game with DAZ3D-rendered graphics to photo-real scenes. It uses the following technique: <ul><li>WAS Nodes to batch load and save images in a directory to another directory. Use either Auto-queue or set the batch number to the amount of images you wish to process. Keep in mind that WAS does not order filenames the same way windows does.</li><li>WD14 Tagger to trigger certain Loras to load when certain things are included in the source image, eg. loading a blowjob-related Lora when WD14 detects the word "fellatio". It also switches to a lower Denoise value when "close-up" is detected.</li><li>Conditioning (Set Mask) to try to make the men and women look more like their respective genders. Using the condition "redhead woman" also kind of works so that you can apply an embedding directly to the area it detects, but this only kind of works.</li><li>3 Controlnets: softedge (HED lines), depth map (Zoe), and mediapipe face to help preserve the scene composition.</li><li>UltimateSDUpscale as the primary I2I sampler - for screen aspect ratio, this seems to work best when I downscale to 1MP so that there's 2 768x768 tiles, then upscale.</li><li>FaceDetailer x 2 - once for the faces, once again for the genitalia.</li></ul>Custom Nodes used: WD14 Tagger, Impact Pack, CLIPSeg, FaceDetailer, Auxillary Controlnets, WAS Nodes, UltimateSDUpscale, maybe some others??UPDATE 1.1: I disabled the 2nd Detailer and one of the controlnets, and also the 3rd conditioning mask. This version does two passes - like a low-res and a high-res (at the full output resolution), which I think yields better results, at the cost of speed.

Created a 2-stage process. First it does a full-image pass before upscaling back to the original resolution for UltimateSDUpscaler. Some groups and nodes are bypassed because I was getting wonky results with the second Detailer set to recognize hands, but that's just because I don't know how to use it properly, I think (it was turning hands into faces).

v1.1

Render To Real v1.0

Render To Real NSFW batcher

MV5BNDY2MGRkMGItOTk4Yi00Y2E5LTk0MWUtMDczYjdjN2MyMzFiXkEyXkFqcGdeQXVyMjUyNDk2ODc@._V1_.jpg