9 October 2026

^{Early Outside In/ See through prototype of the Sony PS VR2 featuring IR LEDs for outside-in tracking, and a see-through view camera on the front and center part of the headset (image: PlayStation.blog)}

 

It is Sunday afternoon, and you have decided to spend an hour somewhere considerably more exciting than your living room. Imagine putting on an eye-tracking virtual-reality headset to play a flight game that supports foveated rendering, checking the cockpit instruments before looking up at a mountain range beyond the windscreen. The landscape seems richly detailed, although you suspect your flying skills may not justify the view.

In a game that supports eye-tracked foveated rendering, however, the graphics system may be doing considerably less work outside the region you are looking at. While you inspect an instrument, it gives that part of the image greater attention; when your eyes move towards the mountains, the high-detail region follows. The aim is to make those changes unobtrusive enough that you remain absorbed in the flight.

There is a good reason for such selective effort. A virtual-reality system produces a view for each eye and updates the images repeatedly as you move, so that the cockpit and landscape continue to appear in the right places. Detailed graphics add to that workload, while the hardware has a finite amount of processing capacity. In a standalone headset, it must also work within a battery budget and a device you can comfortably wear.

Foveated rendering offers a way to spend that capacity where it provides the greatest visual benefit. Its starting point is a feature of your eyes that becomes obvious whenever you try reading something without looking directly at it.

 

How your eyes guide VR eye tracking

 

Focus on one word in this sentence and try making out several words farther along without moving your eyes. You can tell that the text continues, but reading those letters is harder, which is why your gaze keeps shifting as you move through a paragraph.

Our sharpest daylight vision is concentrated at the fovea, a small central region of the retina where cone photoreceptors are densely packed. We move our eyes to bring interesting objects into that region, whether we are reading a sentence or checking an aircraft’s altitude. Peripheral vision remains important for detecting movement and understanding our surroundings, but it does not deliver the same fine detail.

When you look at the cockpit instrument, therefore, the mountains are still part of your view without being available for the same close inspection. Rendering their smallest details at maximum quality can demand graphics work that adds relatively little to what you perceive at that moment.

To take advantage of this, an eye-tracked system needs an estimate of where you are looking. Conventional eye trackers combine cameras, illumination and algorithms, as Tobii explains, to turn measurements of the eyes into gaze information. The rendering system can use that information to decide which part of the image should receive the greatest detail.

Your ordinary eye movements therefore become an input to the game’s graphics, without requiring a button press.

 

How foveated rendering saves graphics work

 

Foveated rendering is a graphics technique that concentrates rendering quality in selected regions of the view while reducing work elsewhere. It takes its name from the fovea, and eye-tracked versions move the priority region with the player’s gaze.

One way to achieve this is to reduce pixel shading outside the high-detail region. Shading helps determine the appearance of surfaces, including their colours and lighting, so reducing the rate at which it is performed in selected areas can lower the cost of producing an image. The display still presents a complete view; fewer shading calculations in a region do not mean that its scenery disappears.

Qualcomm’s technical explanation describes how these efficiencies can give developers room for higher resolution, faster image updates or more complex visuals. The benefit depends on the implementation and workload, and a reduction in rendering effort does not automatically produce an equivalent reduction in the headset’s total power consumption.

The way quality changes across the image is equally important. Developers need to balance the savings against what players might notice, choosing regions and transitions that preserve a convincing view as the eyes move. Sony’s account of PlayStation VR2 rendering technology describes a PlayStation 5 hardware feature that supports foveated rendering, including smooth resolution changes across the display. Eye tracking can help improve quality and performance in games that use it.

There is also a simpler approach, called fixed foveated rendering, which prioritises a predetermined region, usually around the centre of the view. It does not need to measure gaze, although it cannot move that region to match the player’s eye movements. The difference becomes clear when you glance towards an instrument at the edge of the cockpit: an eye-tracked system can follow that glance, while a fixed system keeps its priority region in place.

Rendering approach	Where rendering quality is concentrated	Requires eye tracking?
Similar rendering quality across the view	No gaze-based priority region	No
Fixed foveated rendering	A predetermined region, usually around the centre	No
Eye-tracked foveated rendering	A region following the player’s gaze	Yes

These approaches concern the distribution of rendering quality; games can also use other techniques to manage detail and performance. Which approach a player encounters depends on the headset and supporting software, rather than simply whether a game looks impressive.

 

Neuromorphic sensors that watch for change

 

Following the eyes introduces its own engineering challenge, because the tracker and graphics system must respond quickly enough to keep the effect unobtrusive. As Tobii’s developer guidance explains, tracking delays and visible boundaries between regions can create image artefacts. A glance up from the instruments could expose lower-detail scenery before the high-quality region catches up.

Existing eye trackers already enable this technique, but engineers continue to seek faster and more efficient ways of measuring gaze. The motivation is especially strong in wearable devices, where sensing and processing must share limited power and space with everything else the headset does.

One possibility comes from neuromorphic technology, a broad term for designs inspired by biological nervous systems. Paris-based Prophesee develops event-based vision sensors whose pixels independently report brightness changes that exceed a threshold. Where a conventional camera repeatedly captures complete images, an event sensor supplies a stream of changes, potentially reducing redundant data and responding quickly to movement.

In May 2025, Prophesee and Swedish eye-tracking specialist Tobii announced a collaboration to develop event-based eye tracking for virtual reality, augmented reality and smart eyewear. Their stated aim is a fast, power-efficient solution suited to compact, battery-constrained devices, with Tobii describing the sensors as a complement to its existing camera technologies.

This work concerns how gaze information is collected, while the graphics processor remains responsible for drawing the scene. It offers a route towards more efficient tracking, rather than evidence that gaming headsets already depend on neuromorphic processors. Its practical value will depend on how well the resulting systems perform when integrated into actual devices.

For players, the most successful version of this engineering would be easy to overlook, because its purpose is to preserve the experience while making better use of the hardware.

Back in the Sunday cockpit, you glance towards the mountains before checking your altitude again. If the system is working well, both receive the detail you need when you look at them, leaving you free to enjoy the flight. And, perhaps, to give that approaching ridge a little more attention.