Walt and Jessie music video with h3 vsafastvideo: 8 steps, 1.4mp, 1 hour on a 4090
AndrewJumpen · reddit · 2026-09-02
A Redditor generated a Breaking Bad-style music video of Walt and Jessie using the h3 vsafastvideo model with no reference images, with convincing results aside from Jessie's face occasionally drifting into Walter's. Workflow details: 8 sampling steps, 1.4mp resolution, 1 hour on a single RTX 4090, using the ComfyUI-H3-Project-Suite plugin for continuous music clips, with the song generated in Suno (free tier).
More from Multimodal
- Google's $300 credit broke Gemini nodes in ComfyUI, so one dev built a Vertex AI workaround — Cute-Appointment6874 · 2026-09-02
- Blogger turns blog logo into Pixar x Grok Bot animation with Fable 5.1 — iannuttall · 2026-09-02
- 4 Favorite Midjourney V6/V7 Style sref Codes, Ready to Copy — michaelrabone · 2026-09-02
- Fable 5.1 vs Kimi K3: cinematic quality against double the speed at half the price — iamfakhrealam · 2026-09-02
- WorldGen hands-on: one scene image becomes movable 3D assets — JaynitMakwana · 2026-09-02
- Stress-testing Midjourney SREF codes: only the styles that survive get published — tisch_eins · 2026-09-02