A drummer taps a snare drum very lightly, barely audible. Then hits with full force, producing a loud sharp crack.
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 drummer
- a snare drum
- a drumstick
Event: a drummer strikes a snare drum
Audio presence
- a snare drum
Sound: a snare drum hit
Key standards
Visual physical commonsense (V-PC)
- The drumstick visibly strikes the snare drum twice, first as a barely-there tap and then with clearly full force.
Audio physical commonsense (A-PC)
- The full-force hit is noticeably louder than the light tap.
- Each strike has the characteristic snare drum impact timbre.
Cross-modal physical commonsense (AV-PC)
- Each drum hit sound aligns in time with a visible stick impact on the snare drum.
- The louder hit corresponds to the full-force stick strike.
- The quieter tap corresponds to the light stick tap.