>>2833968i'm not sure if i get what you mean by more solid. like the overall quality? if so that's likely because certain combinations of lora/mix/style tags makes 1.5 try to cram as much detail into the image as possible. plus the 3dness of backgrounds from saya lora works really well with 2.5d models. v3 can't do realism on the level of 1.5 so without loras there probably isn't a way to do similar stuff on it alone
