Wan2.2 i2v 14B Generates Black Mask and Light Fringe in Videos
TrangTranCiu · reddit · 2026-08-13
A user encountered visual artifacts while running the Wan2.2 i2v workflow with the 14B GGUF model and the lightx2v-i2v-14b-480p LoRA. Specifically, a black mask and a bright outline (light fringe) appear around the subject throughout the generated video.
The user has spent the past week trying to troubleshoot the issue by modifying files like nodeswan.py and modelbase.py with ChatGPT's assistance, but the problem persists. They are seeking help from the community to resolve the error.
More from Multimodal
- LTX-2.5 Open-Source World Model: 10s 4K Video in 6.8s from a Single Image — dr_cintas · 2026-08-13
- LTX2.5 Tested on RTX 5090: Generates 60-Second Video in 5 Minutes — tostane · 2026-08-13
- CapCut AI Edit: Local Corrections for Seamless Video Continuity — AIwithGhotai · 2026-08-13
- Indie Dev Tackles Real-Time Virtual Try-On: Stuck on Datasets and Model Quality — Old_Manufacturer7905 · 2026-08-13
- Flora Launches Custom Actions: Generate Custom Image/Video Tools via Natural Language — round · 2026-08-13
- MiniMax H3 Video Model Generates 'Peaky Blinders' Clip — JustARedditUser33 · 2026-08-13