A person sits in a downstairs room. Footsteps from upstairs come through the ceiling clearly and sharply, each step distinct.
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 downstairs room
- a ceiling
Event: a person sits in a downstairs room
Audio presence
- footsteps
Sound: footsteps through a ceiling
Key standards
Visual physical commonsense (V-PC)
- A downstairs room with a visible ceiling and a seated person is rendered.
Audio physical commonsense (A-PC)
- The audible events have the cadence and timbre of distinct footsteps.
Cross-modal physical commonsense (AV-PC)
- The footsteps appear to originate from above the ceiling rather than from within the visible downstairs room.