A person claps inside a grand cathedral and the clap rings on for several seconds. Then the same person claps in a small closet, and the sound is tight and dry.
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 person
- a cathedral
- a closet
Event: a person claps in a cathedral and in a closet
Audio presence
- a person
Sound: a hand clap
Key standards
Visual physical commonsense (V-PC)
- A single person performs a single hand clap inside a cathedral.
- The same person performs a single hand clap inside a small closet or wardrobe.
Audio physical commonsense (A-PC)
- The cathedral clap carries a long spacious reverberation tail lasting several seconds.
- The closet clap is tight and dry with rapid decay and no spacious reverberation.
Cross-modal physical commonsense (AV-PC)
- Each clap sound aligns in time with its corresponding visible hand-clap action.
- The reverberation character of each clap matches the visible space in which the clap occurs.