First-person egocentric view. A single loudspeaker is visible directly in front of the listener at eye level, playing continuous music in a quiet indoor hallway. The listener slowly turns their head and body exactly 180 degrees, so the speaker is now clearly behind them. Instead of shifting from front to back, the sound remains perceived in the same direction as before the turn.
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 loudspeaker
- an indoor hallway
Event: a first-person listener turns 180 degrees in front of a loudspeaker in a hallway
Audio presence
- a loudspeaker
Sound: music from a loudspeaker
Key standards
Visual physical commonsense (V-PC)
- A first-person camera initially faces a single loudspeaker at eye level in a quiet indoor hallway.
- The camera then turns exactly 180 degrees in continuous motion so the loudspeaker ends up clearly behind the listener for the rest of the clip.
Audio physical commonsense (A-PC)
- Continuous music plays uninterrupted throughout the clip from the loudspeaker.
Cross-modal physical commonsense (AV-PC)
- While the loudspeaker is visibly in front of the camera, the music is localized in front of the listener in the audio field.
- After the camera has turned 180 degrees with the loudspeaker now visibly behind, the music remains localized in front of the listener rather than shifting to behind.