Peripheral vision

Peripheral vision is the part of human vision that takes place outside the center of gaze, in the region of the visual field away from the point the eye is fixating. It is also called indirect vision. Peripheral vision resolves far less detail than central vision, but it covers most of the visual field, is comparatively sensitive to flicker and motion, and underlies a broad range of visual tasks.

In virtual reality (VR) and augmented reality (AR), the properties of peripheral vision drive several design decisions. A head-mounted display covers only part of the natural visual field, so the edges of a user's view are often black or empty. The low acuity of the periphery is the basis of foveated rendering, which reduces rendering quality away from the center of gaze. Motion in the periphery contributes to vection and cybersickness, which is why many VR applications darken or restrict the edges of the view during artificial locomotion. Researchers have also built head-worn displays that place low-resolution lights or simple cues in the wearer's periphery.

Reviewed 6 October 2026. Every claim checked against the cited papers, review correction, NEI and Meta developer pages and UploadVR report, with paper authors, venues and DOIs confirmed via Crossref. About review dates.

Definition and extent

There is no sharp anatomical border between central and peripheral vision, and the terms are used in different ways in different fields. In their review of peripheral vision, Hans Strasburger, Ingo Rentschler and Martin Jüttner note that the transition from the fovea outward is smooth. Citing Wandell (1995), they give the foveola a diameter of about 1 degree of visual angle and the fovea a diameter of 5.2 degrees. The parafovea (about 5 to 9 degrees in diameter) and the perifovea (about 9 to 17 degrees) surround it, and together these regions make up the macula, about 17 degrees across.

In perimetry, the clinical measurement of the visual field, the central visual field is the area within 30 degrees of fixation, and the peripheral field is everything beyond it. Strasburger points out that in this sense the periphery is about seven times larger in area than the central field. Vision researchers often use a much smaller boundary. The 2011 review uses "peripheral vision" for anything more than 2 degrees from fixation, following the observation by Korte (1923) that differences in form recognition already appear a few degrees from the center; Korte used the term "indirect vision" for this reason.

The outer limit of the visual field is larger than many textbooks state. The 2011 review originally gave the lateral extent as about 90 degrees from fixation. In a correction published in the Journal of Vision in April 2024, the authors called that figure "wrong" and, citing Traquair's classic text, gave the lateral extent of each eye as around 107 degrees, so that the horizontal extent of the whole field is about 214 degrees. They named the widespread use of cupola perimeters, together with a clinical definition of the visual field as what perimetry records, as a likely reason for the error, and noted that the 90 degree border is the one seen in standard perimetric maps. In a 2020 paper Strasburger tabulated published values for the temporal extent, which range from 90 to 115 degrees, and noted that an extent of more than 90 degrees had long been known. Purkinje reported in 1825 a temporal angle of 100 degrees (115 degrees with a dilated pupil), 80 degrees downward, and 60 degrees upward and nasally. Alexander Friedrich von Hueck in 1840 reported 110 degrees outward, 70 inward, 95 downward and 85 upward, and wrote that when looking into the distance a person overlooks 220 degrees of the horizon.

Physiology

The retina has two classes of photoreceptor, cones and rods, and their distribution explains much of the difference between central and peripheral vision. In a 1990 study of eight whole-mounted retinas from seven donors aged 27 to 44, Curcio and colleagues counted an average of 4.6 million cones and 92 million rods per retina. Peak cone density at the center of the fovea averaged 199,000 cones per square millimeter, varied widely between individuals, and fell steeply with eccentricity, dropping by an order of magnitude 1 mm from the foveal center. The center of the fovea contains no rods: the rod-free zone had an average horizontal diameter of 0.350 mm (1.25 degrees). The highest rod densities lie in an elliptical ring at about the eccentricity of the optic disk, and rod density then declines slowly toward the far periphery.

Visual acuity falls as eccentricity increases. The 2011 review concludes that variation in spatial scale is the major contributor to differences in performance across the visual field, and that this variation is well described by an inverse linear function, although the parameters vary widely between visual tasks. Brian Guenter and colleagues at Microsoft Research used a psychophysical model in which the minimum detectable angular size grows linearly with eccentricity; from a user study they estimated a slope of 1.32 to 1.65 arcminutes per degree of eccentricity.

Temporal sensitivity behaves differently. Exner proposed in 1875 that the periphery is specialized for temporal sensitivity, and Porter observed in 1902 that the critical flicker frequency (the rate at which a flickering light appears steady) rises with retinal eccentricity. Hartmann, Lachenmayr and Brettel (1979) measured a pronounced increase in critical flicker frequency from the fovea out to about 30 to 60 degrees eccentricity, followed by a decrease toward the far periphery, and Tyler (1987) found an increase up to 60 degrees using stimuli scaled to cone density. Strasburger and colleagues give the everyday example of a 50 Hz television screen that looks steady in direct view but flickers when seen out of the corner of the eye.

Perception in the periphery

Peripheral vision is not simply a blurred version of central vision. The 2011 review quotes Lettvin (1976), who wrote that things away from the gaze do not "go out of focus" but rather "lose the quality of form." A major limit on peripheral recognition is crowding, in which a target that is recognizable on its own becomes hard to identify when other patterns are nearby. Bouma (1970) suggested the rule of thumb that, for letters, crowding sets in when the free space between the flanking letters and the target is less than about half the target's eccentricity. The review adds that Bouma's original rule is well defined and gives better fits than the way it is usually cited.

Ruth Rosenholtz, in a 2016 review, describes peripheral vision as having a significant loss of information but still underlying a broad range of visual tasks, from peripheral recognition and visual search to change blindness and getting the gist of a scene.

History of research

According to Strasburger and colleagues, the first quantitative measurements of indirect vision were made by Hueck in 1840, close to the fovea. The first extensive study was published in 1857 by the physiologist Hermann Aubert and the ophthalmologist Carl Friedrich Richard Foerster in Breslau. Their perimeter presented stimuli up to 60 degrees from fixation and used a brief electric arc flash to prevent eye movements. Later work in the 19th and 20th centuries mapped acuity across the visual field, described the photoreceptor layout of the retina, and developed the ideas of cortical magnification and crowding. The review notes that ophthalmology, optometry, psychology and the engineering sciences each developed their own research traditions on peripheral vision and worked independently of each other for a long time.

Applications in VR and AR

Field of view

The field of view of a headset decides how much of the peripheral visual field it can reach. Writing in 2016, Xiao and Benko noted that the human binocular field exceeds 180 degrees horizontally, while VR headsets such as the Oculus Rift covered around 90 degrees and AR devices such as the Lumus DK-40 glasses and the then-upcoming Microsoft HoloLens covered around 40 degrees. They argued that this leaves the user's periphery pitch black in VR or empty of virtual content in AR, which limits immersion and situational awareness.

A wider field of view is not purely beneficial. In a 2002 study by Lin, Duh, Parker, Abi-Rached and Furness, 10 subjects viewed a virtual environment presented in a driving simulator at fields of view of 60, 100, 140 and 180 degrees. Both presence scores and Simulator Sickness Questionnaire scores rose with field of view and approached a plateau beyond 140 degrees.

Foveated rendering

Because acuity drops away from the center of gaze, a renderer can spend less work on the periphery without a visible loss of quality. Guenter and colleagues demonstrated this in 2012 with a gaze-tracked desktop system that rendered three nested layers at decreasing sampling rates around the gaze point, accelerating graphics by a factor of 5 to 6 on a 1920x1080 display. In a 2016 NVIDIA study, Anjul Patney and colleagues found that simply blurring the periphery reduces contrast and produces a sense of tunnel vision; with a post-process contrast enhancement, subjects tolerated up to twice the blur radius before noticing a difference. Their renderer reduced the number of shades by up to 70 percent and allowed coarse shading up to 30 degrees closer to the fovea than Guenter's method.

Commercial headsets use two forms of the technique. Fixed foveated rendering, as described in Meta's documentation for its Quest headsets, renders the edges of each frame at a lower resolution than the center without eye tracking; Meta's documentation adds that the headset lenses already blur the edge of the field of view. Eye-tracked foveated rendering moves the full-resolution region with the user's gaze. For the Meta Quest Pro, UploadVR reported Meta's own measurements in October 2022: at default resolution, fixed foveation gave performance savings of 26 to 36 percent and eye-tracked foveation 33 to 45 percent, depending on the foveation level.

Vection and cybersickness

Visual motion in the periphery is a strong cue for self-motion. Xiao and Benko, summarizing earlier research, write that vection (the illusion of self-motion) is "primarily derived from peripheral visual cues", with the central visual field playing only a small role. When a user moves through a virtual world with a joystick while sitting still, this visual motion conflicts with the vestibular sense and can cause simulator sickness.

Restricting the periphery during virtual movement is therefore a common comfort technique, known as Dynamic FOV Reduction, vignetting or tunnel vision. Ajoy Fernandes and Steven Feiner of Columbia University tested soft-edged circular cutouts that narrowed each eye's view as gamepad speed and turning rate increased. In a two-session study with 30 seated participants wearing an Oculus Rift DK2, the restrictors reduced reported VR sickness without lowering presence, and most participants did not notice them. Meta's developer guidance describes vignettes as darkening or occluding the edges of the screen during movement and warns that the narrower view can feel disorienting or claustrophobic. Meta's locomotion guidance also suggests making scene geometry opaque in peripheral areas, for example the side windows of a vehicle, to reduce optic flow.

Other designs show stable content in the periphery instead of blacking it out. In a 2017 developer blog post now hosted by Meta, Tom Heath described a head-locked static scene revealed at the periphery during locomotion, reasoning that a stable peripheral environment gives the brain evidence that the body is not moving, whereas a black border is ambiguous. Xiao and Benko's sparse peripheral display included a "countervection" mode that showed motion in the periphery opposite to joystick-driven movement; 11 of 14 participants preferred it to having no peripheral display, although the authors described their questionnaire results as inconclusive. In a paper accepted to IEEE VR 2025, Tongyu Nie and colleagues described "peripheral teleportation", which renders the periphery from a pair of rest-frame cameras that follow only the user's physical motion. In a between-subjects study with 90 participants, it reduced discomfort compared with a black field-of-view restrictor and an unrestricted condition.

Peripheral displays

Several research prototypes add a low-resolution display aimed only at the periphery. Xiao and Benko's sparse peripheral displays, presented at CHI 2016, surround a headset's main screen with diffused RGB LEDs. SparseLightVR added 70 LEDs to an Oculus Rift DK2, extending its roughly 84 degree horizontal field of view to about 170 degrees. SparseLightAR added 112 LEDs to a custom see-through headset based on a Samsung Gear VR and a Galaxy S6 phone, with a 62 degree central display and a total field of view of more than 190 degrees. In a search task with 17 participants, average targeting velocity for initially visible targets was 49.2 degrees per second with SparseLightVR, compared with 41.4 with the headset's regular 84 degree view and 21.2 with a simulated 50 degree AR view; all three differences were statistically significant.

For AR and smart glasses, the periphery is also a place for notifications that do not block the center of view. The eye-q prototype by Enrico Costanza and colleagues (MobileHCI 2006) placed LEDs in eyeglasses so that cues appeared in the wearer's periphery; their user study found that the cues could be designed to meet specific levels of visibility and disruption for the wearer. At CHI 2022, Janaka and colleagues tested a circular progress bar shown in paracentral and near-peripheral vision on an optical see-through head-mounted display and found that, compared with textual and linear progress bars, it reduced notification distractions during face-to-face conversations without loss of eye contact and was preferred by users. A 2026 preprint by Yutong Ren, Arnav Reddy and Michael Nebeling, PeriphAR, used peripheral cues to confirm gaze-based selection on a monocular AR display and reported that peripheral vision was more sensitive to color than to shape, with that sensitivity falling quickly at lower contrast.

Visual field testing

Peripheral vision loss is an early sign of some eye diseases. The US National Eye Institute states that glaucoma usually causes a slow loss of vision beginning with side (peripheral) vision. VR headsets are being evaluated as portable perimeters. A 2024 systematic review in the journal Eye identified 64 studies describing 36 headset-based perimetry devices and reported a growing consensus that they perform comparably to, or even better than, standard automated perimetry. A 2025 systematic review in PLOS ONE, limited to adults with glaucoma, included 14 studies of 10 devices, including the Oculus Quest, and concluded that the results were promising but that test-retest repeatability needed more study.

See also

• Human visual system
• Retina
• Field of view
• Foveated rendering
• Fixed foveated rendering
• Vection
• Cybersickness
• Dynamic FOV Reduction
• Tunneling
• Eye tracking