A book is placed gently on a desk with a barely audible tap. Then the same book is slammed down hard, producing a loud bang.
Seedance 2.0ByteDance · proprietary
Kling 3.0 OmniKuaishou · proprietary
Veo 3.1Google · proprietary
LTX-2.3Lightricks · open
Ovi 1.1character.ai · open
JavisDiT++JavisDiT · open
MagiHumanGAIR-NLP · open
Per-prompt rubric
Basic standards
Visual presence
- a book
- a desk
Event: a book is placed or slammed on a desk
Audio presence
- a book
- a desk
Sound: a book on desk impact
Key standards
Visual physical commonsense (V-PC)
- The book visibly contacts the desk twice, first as a gentle placement and then as a forceful slam, with the book itself meeting the desk on each impact.
Audio physical commonsense (A-PC)
- The second impact is noticeably louder than the first impact.
- Each impact sound is a real book-on-desk thud.
Cross-modal physical commonsense (AV-PC)
- Each impact sound aligns in time with the visible contact of the book on the desk.
- The louder impact corresponds to the forceful book slam.
- The quieter impact corresponds to the gentle book placement.