Hi guys, I'm over halfway in the process of turning my book into a visual novel and i'm testing out the viability of using ai generated images for the visuals of my visual novel, so far i've found the results quite pleasing to make base images and post-edit them in photoshop for different time-zones. (creating a daytime image and turning it into nighttime myself in photoshop and adding rainy day effects, etc) i found this workaround for not being able to reproduce exact versions of the same room, texture wise because of reflections being interpreted as actual elements ( kept turning sinlight on paintings into a mirror for instance :,) ) to be working out quite nicely. hihi

I run comfyui locally to generate the characters
using flux as base model for realism,
faceid for reproducing the same character as much as possible,
openpose for setting the poses,
and ip adapterv2 for the clothes!

Sophia

🤖 NightCafe

🚀 Features

🔒 Secure Payments