Facesitting challenge
Added 2025-07-26 15:02:10 +0000 UTCAll my images are AI generated. I use NovelAI v4.5 mostly but sometime I switch back v4.0 (I found that v4.0 have more diversity in rendering style while v4.5 is more prompt accurate). I create text myself but I ask ChatGPT for correction since i'm not english native. I also ask for a japanese translation because when there is dialog, I always post the JP variation on Pixiv. I use Paint dot NET for adding these dialogues on images. And finally I upscale (with Real-ESRGAN) then resize (with Paint dot NET) images before publishing them
