Edge content needs a head turn to be seen clearly

What it is

A menu parked at the far edge of the view is not “glanced in peripheral vision and done”. The head turns so the fovea can land on it. That extrafoveal head turn is the real cost of edge placement: ocular range is limited, and beyond it the head must join. Putting something at the edge adds a neck movement to every read, not a mild loss of prominence.

Why it happens

Useful reading lives in about a degree of fovea; the parafovea can identify letters, but speed and accuracy fall fast. Gaze shifts beyond roughly fifteen to twenty degrees recruit the head — eye first, head follows, eye then counter-rotates. A headset adds a second friction: the head now moves the mass of the device, and the neck is less willing to pay that trip for a caption than it is in front of a monitor. Edge content is therefore not “a bit harder to read”; it is unread by default unless the task forces a turn. While the head turns the world slides on the display, so reading waits until the head settles, and edge information is timestamped a beat later than centre information.

Studying it

Use a task that cannot finish unless an edge label is read, and record eye–head coordination. A post-hoc “did you notice it?” questionnaire is not enough — noticing and reading are different events.

Independent variables: eccentricity of the target from view centre, label complexity (swatch / icon / word). Dependent variables: whether a head turn occurs, amplitude and latency of that turn, time to first correct read, miss rate.

Guitton-style eye–head coupling and gaze-shift work in HMDs both sit here. A seat back that limits head motion inflates the cost of eccentricity; the chair is not a free choice.

Where it stops holding

When the head is blocked by a seat, a collar, or a tight cockpit, edge content is unreachable, not “one turn away”. On wider-field devices with eye tracking, people may cover moderate eccentricity with a large saccade and the head-turn threshold moves out. Content that stays locked to the edge while the head rotates always costs a turn to read; world-locked content that can be faced once is a different bill. Low vision narrows the readable cone, so the same eccentricity forces a head turn sooner.

Applying it

• Put elements that must be read as type, numbers, or state into the zone the head need not move to foveate. Leave the edge for things that do not need letter identification.
• Count edge reads along a task path; charge each one as a head turn against the energy budget.
• If a turn is required, give a centre-visible pointer. Do not assume the periphery will finish the read.
• How to check: record head pose during the main task. Log every head movement made in order to read UI. If that count exceeds the heading changes the task actually needs, the edge is taxing reading.

Related

• Same group: N3.07.2 Critical information must not sit at the edge · N3.07.3 The edge is fit for presence cues · N3.07.4 Pull edge elements inward to leave margin for optical falloff · N3.07.5 A ring layout fits the shape of the field of view better than a rectangle · N3.07.6 Content outside the binocular overlap is seen by only one eye · N3.07.7 Elements that enter and leave the view with head motion need hysteresis or they flicker
• Nearby: N1.03 Field of View · N3.04 Text Readability in 3D Space
• Search terms: extrafoveal head turn · gaze shift · eye-head coordination