Eyes move fast over small amplitudes for fine pointing; the head moves slowly over large amplitudes for coarse aiming

What it is

Eyes and head can both change gaze direction, at different movement scales. Eye movements are fast, low-inertia, comfortable within about ±15–20°, and fit fine pointing inside the current heading. Head movements are slow, high-inertia, able to turn tens of degrees and, with the trunk, more, and fit coarsely rotating the whole heading to a new region. The split is by scale, not a contest over which is the stabler sole pointer—that is head pointing as a device.

Why it happens

Orbital muscle can finish a saccade in tens of milliseconds; neck muscle has to move the mass of the skull and headset, so the same visual angle is slower and costlier. Gaze shifts beyond about 15–20° usually recruit the head: the eye moves first, the head follows, the eye makes a compensatory return (VOR and eye–head synergy). Below that scale, eyes-only is cheaper. Giving a small target to the head is a low-bandwidth actuator on a high-bandwidth job; giving a large-angle shift to the eyes alone drives them to an uncomfortable limit and soon recruits the head anyway—if the system still treats “only the eye as pointing,” it misreads.

Sidenmark and Gellersen described this natural synergy in VR: the head for range, the eyes within range. A product that forces a single scale (head only or eyes only) is working against that physiology.

Studying it

Have people shift gaze to targets of different amplitudes (5°, 20°, 60°) and record the angular contribution and timing of eye, head, and trunk. Compare free synergy, head locked (chin rest), and eyes locked (head turn only). Outcomes: completion time, eye-in-head eccentricity, neck load. Headset mass and field of view change the recruitment threshold; measure on the target device. Laboratory laser-spot targets underestimate the two-stage “find the region, then read the label” of a UI.

Where it stops holding

Users with almost no neck motion must make even large shifts with the eyes or by turning a chair; the scale split collapses to one channel. When eye tracking is unavailable, fine pointing also falls on the head, and small targets must grow. In some cultures or performance settings people habitually “look with the whole head” and use the eyes’ fine channel less. Recumbent poses change which neck axes are available, and the horizontal/vertical scale split can swap.

Applying it

• Use the head to coarse-turn large panels and distant zones; use the eyes to fine-point small controls inside a zone; do not make the head align a one-degree button.
• Do not lock either channel: locking the head drives the eyes to the limit; locking the eyes turns the neck into a mouse.
• Verify on 5° and 60° targets that the product is not forcing a head turn onto small targets or an eyes-only fling onto large shifts.

Related

• Same group: C8.12.2 Eyes and head often desynchronize; gaze target and head orientation can separate · C8.12.3 When combined, the head typically sets overall orientation and the eyes fine-tune within it · C8.12.4 Head-dependent schemes fail for users with limited neck mobility and need an eye-only path
• Adjacent: C8.05 Head pointing · C8.01 Fixation and saccade
• Search: eye-head coordination · gaze shift · vestibulo-ocular reflex