Fixing AI anime background inconsistency by pre-shooting a 360° orbit video with MiniMax H3
Hailuo_AI · x · 2026-09-03
A Japanese animator shares a practical fix for a common AI-anime pain point: character consistency is largely solved with reference images, but backgrounds get redrawn inconsistently on every cut. The method: first use MiniMax H3 to generate a single 360° orbit video of the room — starting from a wide-angle single illustration — treating it as the one canonical reference. Camera placement and cut points were blocked out with a gray clay scene (just boxes and cylinders) in Blender, generated via Claude Code. For each cut, a frame showing the wall the camera faces is extracted from the orbit video and used as the reference image. The author notes the source doesn't have to be an orbit video — any consistent background source you can crop frames from works.
More from Multimodal
- Gemini's 88% video token cut landed on 3.7 Flash, not the 3.8 everyone's talking about — Servola-Journal · 2026-09-03
- Open-source ArtSmoker pipeline turns text prompts into textured, Blender-ready 3D models on your own AWS — niravdd · 2026-09-03
- SonicCaps: 15M-caption audio dataset improves CLAP retrieval and zero-shot classification — serrjoa · 2026-09-03
- Midjourney character sheet + Seedance long video recreate a 70s Giallo thriller — Kyrannio · 2026-09-03
- Catwoman character explorations generated with Midjourney — hewarsaber · 2026-09-03
- Fable 5.1 one-shots a full quad-turbo W16 engine CAD model and animation — repligate · 2026-09-03