VRHow / VR History / How VR displays improved

How VR displays have improved

At a glance: comparable dated examples

The rows use the units each source actually publishes: total pixels for DK1/DK2 and per-eye pixels for later dual-display headsets. A total-panel figure is not directly equivalent to a per-eye figure.

1960s: CRTs made a head-mounted image possible

Ivan Sutherland's 1968 paper describes a headset with two miniature cathode-ray tubes. Prisms magnified each tube's image into a virtual image about 18 inches in front of the eye; the reported view was roughly 40 degrees, and half-silvered mirrors allowed the wearer to see computer imagery and the physical room at once. The tubes drew wire-frame line graphics, not today's textured scenes.

The date needs a small qualification. The Computer History Museum catalogues the Harvard object as “1967 ca.”, while Sutherland's published paper is from 1968. It is safer to call this the 1967–68 Sutherland/Sproull system than to force a single “first” date. Either way, the engineering limitation is clear: CRTs, high voltage and a counterbalanced support arm made the display an experimental apparatus rather than a portable product.

2013–2016: consumer VR moved to flat panels

The consumer revival used small, readily available flat panels behind magnifying lenses. A contemporary Road to VR comparison records the Rift DK1's 1280 × 800 LCD and DK2's 1920 × 1080 Pentile OLED. Those figures are total panel dimensions; dividing the DK2 panel across two eyes gives 960 × 1080 pixels per eye before lens distortion and the rendered image's sampling are considered.

The important DK2 change was not just more pixels. The same comparison describes its OLED panel as low persistence and reports much less smearing than the DK1's LCD. OLED also allowed very dark unlit pixels because it does not need an LCD backlight. This is a useful historical example of why “higher resolution” is an incomplete explanation of a better-looking headset.

2019–2021: LCD caught up through subpixels, persistence and refresh

By 2019, Valve's Index showed how an LCD could be tuned for VR rather than judged by resolution alone. Valve lists dual 1440 × 1600 full-RGB LCDs, 80/90/120/144 Hz modes and an illumination period of 0.330 ms at 144 Hz. It also says the RGB layout supplied 50% more subpixels than OLED and that the higher frame rates improved motion realism and optical comfort.

Optics improved in parallel. Index used dual-element, canted lenses with physical IPD and eye-relief adjustment; Valve says the design delivered a typical-user field of view 20 degrees wider than HTC Vive. These are not panel-resolution changes, but they alter how much of the panel is useful, how uniformly it looks sharp and how much the wearer can see by moving the eyes.

HTC's 2021 VIVE Pro 2 is a later, directly comparable LCD milestone: dual RGB low-persistence LCDs at 2448 × 2448 pixels per eye, with 90 or 120 Hz operation and up to 120 degrees horizontal field of view. A 2022 Scientific Reports study measured the VIVE, VIVE Pro and VIVE Pro 2 and found that both spatial resolution and temporal response mattered; it specifically associates the Pro 2's 120 Hz and 5% duty-cycle pulsed emission with improved spatiotemporal performance.

2020s: density, compact optics and seeing the real world

Meta's 2020 Quest 2 announcement is a clear standalone example: 1832 × 1920 pixels per eye, described by Meta as 50% more pixels than the original Quest, with support for 90 Hz. The jump shows the practical trend toward per-eye figures, but it still does not describe lens clarity, edge-to-edge sharpness or motion behavior on its own.

Apple's announced 2023 Vision Pro pushed panel density in another direction. Apple says its two micro-OLED displays pack 23 million pixels combined, with custom catadioptric lenses, wide color and high dynamic range; it also states that the R1 chip streams new images to the displays within 12 ms. This is a display system improvement, not a claim that every scene is automatically sharper: optics, rendering resolution and motion-to-photon behavior remain part of the result.

Passthrough changes what “display quality” includes in mixed reality. The Sutherland system already combined a see-through optical path with CRT imagery, but modern headsets can reconstruct a camera view. Valve's Index page documents stereo 960 × 960 RGB cameras intended for computer vision and “high-quality stereo pass-through.” That camera specification is not the resolution of its virtual display; it is evidence that the real-world view is a separate imaging pipeline with its own latency, exposure and distortion limits.

What actually improved?

• Spatial detail: examples move from a tiny CRT wire-frame image to thousands of pixels per eye, and eventually micro-OLED density.
• Temporal clarity: low persistence, short illumination windows and higher refresh reduce the blur or judder caused by head and eye motion. The FDA-authored study cited above measured this separately from spatial resolution.
• Optical usability: better lens designs, IPD/eye-relief adjustment, subpixel layouts and wider fields of view make more of the panel useful.
• World awareness: optical see-through and camera passthrough add a second visual path; a high-resolution virtual panel does not guarantee a high-quality passthrough view.

Sources

• Ivan Sutherland, “A Head-mounted Three Dimensional Display” (1968) — original technical paper
• Computer History Museum — The Sword of Damocles catalog record
• Road to VR — Oculus Rift DK2 and DK1 comparison (2014)
• Valve — Index headset display, persistence, optics and camera specifications
• HTC VIVE — VIVE Pro 2 official specifications
• Zhao et al., Scientific Reports — “Spatiotemporal image quality of virtual reality head mounted displays” (2022)
• Meta — Introducing Oculus Quest 2 (September 16, 2020)
• Apple — Introducing Apple Vision Pro (June 5, 2023)