Text-to-Image
Diffusers
Safetensors
StableDiffusionPipeline
stable-diffusion
Inference Endpoints

Images with gradients are coming out blocky.

#19
by Explorographer - opened

File Render info:

Prompt:A waterfall in a phantasmagorical forest of twisted trees and boulders, vines, overgrowth, autumn, 8k, Highly detailed, cinematic light, ultra realistic, perfect composition
Negative prompt: ((((ugly)))), (((duplicate))), ((morbid)), ((mutilated)), [out of frame], extra fingers, mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), ((ugly)), blurry, ((bad anatomy)), (((bad proportions))), ((extra limbs)), cloned face, (((disfigured))). out of frame, ugly, extra limbs, (bad anatomy), gross proportions, (malformed limbs), ((missing arms)), ((missing legs)), (((extra arms))), (((extra legs))), mutated hands, (fused fingers), (too many fingers), (((long neck)))
Steps: 150, Sampler: Euler a, CFG scale: 7.5, Seed: 2724423895, Size: 1216x768, Model hash: 2c02b20a

tmpnvep8_ls.png
tmp1k2qdwqm.png

Every water or cloud gradient I render comes out fragmented and blocky.

Thought it wasn't happening with Euler but it is. All samplers are producing this blocky behavior.

More examples. Doesn't matter the sampler or settings. If it has Volumetric fog, Volumetric light or cinematic light in the prompt this is happening.

download.png
image.png
image.png

These are all on checkpoint 768-v-ema.ckpt [2c02b20a]

Sign up or log in to comment