Spatial content is not ordinary video with a depth effect added at the end. It must be authored for a viewing zone, an optical volume and a real room.
Why must content start with the endpoint?
A spatial scene has no universal master. The same asset can read as convincing on one optical architecture and unstable on another because the viewing zone, focal plane, depth range, panel resolution, enclosure and tracking model are different.
Begin with an endpoint profile: physical dimensions, orientation, expected viewing distance, viewing cone, number of observers, ambient light, optical center, safe depth range, frame rate, codec and interaction method. Only then choose the camera, scene and motion.
What should occupy the scene?
Start with one dominant subject. A person, handbag, shoe, mechanism, artifact or character gives the eye a stable reason to interpret depth. Conventional commercials often fail because every element competes for attention and the background behaves like another room rather than the interior of a display.
Remove scenic information that does not help the spatial read. Use the chamber as a stage, not as a window for arbitrary footage.
How do scale and grounding create credibility?
Choose scale deliberately. Human content should feel anatomically plausible for the display format. Product content should be large enough to reveal form without losing category cues. Keep the subject fully contained unless the optical effect explicitly supports crossing the display plane.
Ground contact is critical. Feet, wheels, bases and product supports need a coherent relationship to the chamber floor. A restrained contact shadow, reflection or occlusion cue can anchor the subject; an incorrect one makes it float. The interior floor and background should share one controlled color system so the depth cue comes from form and light rather than a pasted-on stage.
How much depth should you use?
More parallax is not automatically more convincing. Excessive disparity can create discomfort, double images or unstable edges. Define a depth budget around the display's comfortable focal region, then reserve the strongest separation for the subject feature that matters.
Keep fine text and critical interface elements close to the display plane. Sony's official Spatial Reality Display guidance specifically warns that two-dimensional UI can flicker or appear to pop out excessively when left/right views diverge.
What motion works?
Motion should reveal form or intent. Slow rotation exposes a product silhouette. A small weight shift makes a person feel live. A controlled exploded view explains a mechanism. Avoid continuous camera pushes, abrupt zooms and cuts that change the apparent size of the physical display. The chamber should remain stable while the subject moves within it.
For loops, hide the seam in a natural pause or cyclic action. Let the first frame establish the subject quickly; public viewers may arrive at any moment.
How should people be captured?
Use a camera position and focal length that preserve natural proportions. Keep hands and feet inside the safe frame unless the system is designed for edge crossing. Match eyeline to the intended observer height. Separate hair and clothing from the background without producing a hard cutout edge.
For live presence, consistency matters more than cinematic complexity: stable exposure, predictable framing, low latency, clean segmentation and intelligible audio.
How should products and objects be presented?
Choose materials that the display can reproduce reliably. Highly reflective, transparent and very dark objects may require special lighting and edge treatment. Use rotation only when it teaches the viewer something. Layer annotations outside the hero object or bring them forward briefly, then return to an uncluttered view.
Exploded views, material transitions and cutaways are valuable because they reveal information that a physical sample cannot show easily.
How do you validate a spatial scene?
- 01
Check containment
No unintended clipping, frame collision or detached extremities.
- 02
Walk the viewing zone
Inspect center, edges, near and far distances for crosstalk, distortion and lost detail.
- 03
Test the room
Evaluate real ambient light, reflections, sightlines and surrounding visual noise.
- 04
Watch the whole loop
Confirm stable scale, no accidental zoom and no glitch at the seam.
- 05
Observe a new viewer
Can someone understand the subject and intended action without explanation?
What should the final content package contain?
Deliver the approved master, endpoint-specific adaptation, poster frame, safe-area reference, duration, audio policy, rights metadata and fallback scene together. Record which physical profile and software version passed validation. Spatial content becomes operational only when another person can launch the correct version with confidence.