>>2787386thanks anon
In which case I'm asking for feedback on this specific image generation:
https://files.catbox.moe/hf0avt.pngThis is with using the Novel AI VAE and Model, from the webui, and only using one txt2img generation.
In this generations, the things that are off to me are:
- The lower background is not continuous; the rail breaks off behind the character and continues at a different spot.
- It's not obvious with this seed, but with the same prompt I easily get fucked up hands, and occasionally fucked up anatomy.
- The stomach has a mild deformation through it.
Other than adding tons of negative prompts (which there are a ton of), I am not sure what prompts (or other settings to configure) could fix these issues.
