Credit & reuse status
Visual published by xyn.sony.net, shown for identification of the record. Publisher/source credit is inherited from the existing project asset ledger; individual creator and production permission require review. REFERENCE ONLY — NOT CLEARED FOR PRODUCTION. Full frame retained. Publisher image.
What the record documents
Sony's developer platform for the Spatial Reality Display keeps a dated history of its own software. Its release notes record 'SDK 1.0.0 (2020/10/31)' as the 'initial release', the earliest date this entry can confirm from Sony's own documentation. That marks the software platform's first release rather than a marketing unveiling, and it is treated here as exactly that narrower fact. The device itself, sold under the model name ELF-SR1, is built to show content rendered on an attached PC using Unity or Unreal Engine, per the same developer site, current as retrieved on 16 September 2026 rather than tied to one dated announcement.
What creates the image
The display does not show a fixed stereo pair. Sony's technical description states that 'high-speed vision sensors and eye-recognition technology' built into the unit 'detects the position of each of the left and right eyes in real-time, not only horizontally and vertically, but also in depth', and that the attached PC then generates 'light source video...from 3DCG data' matched to that exact eye position. A layer of 'micro-optical lenses' fixed across the panel then 'delivers stereoscopic images to the left and right eyes'. The system recalculates a fresh stereo pair for wherever the tracked eyes currently are, rather than broadcasting one fixed set of views the way an untracked lenticular panel does; the company's own white paper describes the PC as able to 'render literally hundreds of views, each precisely calculated for the user's exact position' as the head moves.
The single-viewer limit
That white paper, headed 'White Paper(ELF-SR1 only)', states plainly that the display supports 'one' user, within a stated working range of '12–30 inches (30 cm–75 cm)' from the screen. This is the direct trade for tracked precision: a system built around one pair of tracked eyes has no defined behaviour for a second viewer, who would either see an image calculated for someone else's eye position or nothing coherent, a point the documentation reviewed here does not itself describe. Outside the stated distance band, tracking accuracy is not specified in the pages consulted.
Questions to bring to a demo
- What happens to the image, precisely, for a second person standing beside the tracked viewer?
- How does tracking behave at the edges of the 30–75 cm range rather than in the middle of it?
- Does the eye-tracking sensor fail visibly, or degrade gradually, under different ambient lighting?
Sony's own documentation is unusually specific about the viewer limit, which is worth noting: many spatial displays leave that number for a buyer to discover in person, whereas this platform states it as a specification.
Sources & provenance
Preserved from the existing project research collection. These records were not independently re-researched as part of the design update. Supplier, historical and technical claims remain subject to editorial review; no hands-on test is implied.
Source publication: undated · Retrieved 2026-09-16
Describes the eye-recognition sensor and micro-optical lens layer that together generate the stereoscopic image.
Source publication: undated · Retrieved 2026-09-16
States the display supports one viewer within a 12-to-30-inch working distance and renders hundreds of position-matched views.
Source publication: undated · Retrieved 2026-09-16
Dates the developer SDK's initial release to 31 October 2020, the earliest confirmed date in Sony's own documentation.
Record: 0009-bf-023 · Rewrite revision: 2026-09-16.1 · Actual site publication: pending. Historical event: 2020-10-31.