A narrow field of view produces tunnel vision
Aliases: scuba-mask effect · restricted FOV · visor tunnel · field restriction
What it is
A scuba mask, a pair of binoculars, the peephole in a door: the shared experience is the world collapsed into a tube — clear straight ahead, nothing beyond the wall of the tube. When a headset field is cut narrow enough, the same tunnel vision appears. It is not blur. A ring of solid angle has been sliced off, and the remaining tube becomes the default working window. A dashboard has to be swept with the whole head, a teammate standing outside the tube drops out of the scene, walking starts to honour the line the tube is aimed along.
The tunnel is a shape problem. Spreading the same pixel count over a narrower angle can sharpen the centre and people will still report being in a tube. Sharpness does not restore the missing ring.
Why it happens
In natural vision the periphery cannot read type, but it continuously supplies a base of “there is still a world”: optic flow, self-motion, space on both sides of the body. Cut the field into a tube and that base is removed; the fovea becomes the only sampler. To know what lies outside the tube, the whole head has to turn so the tube can be fitted over that patch. Head inertia is an order of magnitude above a saccade, so scanning becomes a posture change rather than a millisecond flick.
The tube also changes motor strategy. People rotate the body to drop a point of interest into the centre of the tube, rather than sweeping with the eyes. While walking, the tube points down the corridor and doors and people to the side are treated as absent until contact. In driving and flight simulation this shows up as staring at the runway and losing the aircraft to the side. Tunnel vision is a behavioural adaptation after the hardware has forced the working window to shrink. It is not a separate eye disease.
Studying it
Make several tube diameters with a field restrictor or a software crop. Choose tasks that themselves depend on the periphery: a lane change, a pass to a teammate at the side, watching several instruments on a desk at once. Compare with natural vision or a wide-field headset.
Independent variables: field diameter (horizontal and vertical separately), whether a hard black bezel is visible, how much the task depends on the periphery. Dependent variables: number and amplitude of head sweeps, missed lateral events, coupling between body heading and gaze heading, a subjective “in a tube” rating.
An eye tracker inside the headset can only say what was looked at inside the tube. Misses outside it need an external event log. Reporting that “gaze stayed central” under a narrow field is close to tautology and cannot, on its own, prove a tunnel.
Where it stops holding
A deliberately narrow field (a scope, a camera viewfinder, a telescope tool) is the task; the tunnel is a function, not a defect. On exit the field has to be given back, or the tunnel leaks into the main experience. Dark adaptation, stress, and alcohol also produce a functional tunnel; that is the person’s state, not to be merged with headset crop as one cause. A headset that is very tight vertically and acceptable horizontally tunnels mainly on looking at the hands, the keyboard, the feet, and may look fine on horizontal tasks. Users already trained to scan with the head (motorcycle helmets, fire visors) adapt to the tube faster; a novice’s head-sweep cost in the lab will overestimate their long-run cost. Video see-through that still shows degraded reality outside the tube feels less like a tunnel than a hard black frame — the black frame is a stronger tube wall.
Applying it
- Mark tasks that require simultaneous knowledge of lateral state (collaboration, driving, carrying) as high-risk under a narrow field. Prefer a wider-field device, or fold the lateral state into an indicator inside the tube, rather than assuming continuous head turning.
- A visible black frame strengthens the tube. Prefer a mild optical falloff to a solid black wall. In demos, do not crop the field with software to something narrower than the lens “for looks.”
- Teach a head-scan pattern for narrow-field experiences, and accept a slower pace. Do not sign off against a wide-field time budget.
- How to check: put a changing status source on each side and a main task in the centre. If both sides have changed by the time the main task finishes and the person has not reacted, the tunnel is in force. Count head sweeps in that segment — almost none while claiming “I saw everything around,” or a very high count needed to finish, are two faces of the same tube.