Hours of tuning for one 54-second video: keeping a character, 3 pets and a cooker consistent
wnstj1085 · reddit · 2026-11-15
A Redditor shares a 54-second AI-generated video where they spent hours keeping one character, three pets, and a capsule cooker visually consistent across the whole clip. The image shows it was made with Vidu 3.0, with the character's outfit and the pets' appearance holding up remarkably well — a hands-on look at tackling multi-subject consistency, one of the hardest problems in video generation.
More from Multimodal
- Swipe-comic 'Follow the little boat' shows vast mind pausing for a child-sized wonder — tallmetommy · 2026-11-15
- Grok Imagine team layoff sparks debate: will code beat diffusion for video? — IndraVahan · 2026-11-15
- Blogger shares the coolest prompt to test on any AI video model — umesh_ai · 2026-11-15
- Realistic Midjourney v8.2 portrait: full prompt for a candid morning-reading scene — tisch_eins · 2026-11-15
- Racing a 1-Day Credit Expiry: Redditor Ships AI-Animated Short 'Deprived of Your Tribe' Part 1 — Creepy_Phase_7556 · 2026-11-15
- Blender MCP hits 34.5k stars: direct 3D modeling in plain language with any LLM — sidahuj · 2026-11-15