# Wan 2.2 video generation looks distorted/blurry in ComfyUI — need help troubleshooting — retained record

> Retained reddit record from reddit:comfyui.

Canonical: https://highsignal.app/data/records/ecb4675c9c9822c6

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:comfyui)  published 2026-10-12 22:09 ingested 2026-10-13 02:42

[original source ↗](https://www.reddit.com/r/comfyui/comments/1x4fhu4/wan_22_video_generation_looks_distortedblurry_in/)

## retained text



```
Hey everyone! I’m new to using comfy AI and making my first workflow, i’m having trouble getting decent video results with Wan 2.2 in ComfyUI Desktop and could use some help figuring out what’s wrong.  
**My PC specs:**  
GPU: NVIDIA RTX 5070 Ti (16GB VRAM)  
CPU: AMD Ryzen 7 7800X3D  
RAM: 32GB DDR5-6000  
PSU: MSI 800W Gold  
**My setup:**  
ComfyUI Desktop  
Wan 2.2 image-to-video workflow  
Two Wan 2.2 14B FP8 models (high-noise and low-noise)  
Two KSampler Advanced nodes  
832 × 480 resolution  
81 frames at 16 FPS  
VAE Decode connected to Create Video  
**The problems I’m experiencing:**  
My generated videos look distorted, blurry, or have strange colors instead of coherent movement.  
My system RAM reaches around 92% during generation.  
My GPU reaches 100% utilization, while VRAM usage has been around 76% in one recent run.  
I’m trying to figure out whether the issue is my workflow connections, sampler settings, model/VAE configuration, RAM usage, or something else.  
**What I’ve tried:** I’ve installed the ComfyUI\_Dynamic-RAMCache custom node, but I’m still seeing high RAM usage during generation.  
I’m relatively new to ComfyUI, so I’d really appreciate help identifying what to check first. I’ll attach screenshots of my full workflow and system specs, along with an example of the output.  
Thanks!
```



document fetched 2026-10-13 00:42 · canonical https://reddit.com/r/comfyui/comments/1x4fhu4/wan_22_video_generation_looks_distortedblurry_in
