Interface distance must fall in a comfort zone

What it is

The depth range that can be fused is wider than the range that can be looked at for a long time without tiring. The latter is what interfaces get. The zone of comfort (Percival’s comfort band) is the region where, with accommodation locked at the headset screen distance, vergence demand still sits in the comfortable part of fusional reserve — not on the edge of barely single vision. Panels, labels, type that must be read, have to land in this band, not in the band of “within arm’s reach” or “looks near and stereo.”

The zone is a dioptric window relative to screen distance, not a universal world-centimetre figure. When screen distance moves, the comfortable world distance moves with it.

Why it happens

Percival placed the comfort band in the middle portion of the zone of clear single binocular vision: fusional reserve is not spent, and the accommodation–vergence couple does not have to be wrenched by will. Once a headset nails accommodation, that band becomes “do not let vergence distance leave screen distance too far” — especially not toward near, because each centimetre nearer is a larger dioptric step. One metre to two metres is only 0.5 D; forty centimetres to one metre is already more than 1.5 D. So moving an interface nearer drops it out of the band faster than moving it farther.

Reading, staring at status, selecting, are minute-scale fixations, and they use this band. A virtual hammer that flashes through at thirty centimetres has a large conflict and a short fixation, and need not follow interface rules. A menu built as a thirty-centimetre “phone” inverts the rule: the interface borrows a tool’s distance and demands the continuous fixation the tool never needed.

Studying it

On a fixed screen distance, place the same panel at a set of world distances, run a reading or selection task, and collect discomfort, reading speed, and error as a function of distance. The operational definition of the band is the stretch of distance where speed and error have not yet clearly broken and ocular discomfort stays below a threshold written down in advance.

Independent variables: dioptric gap of the panel from screen distance, task duration, angular type size (angular size must be compensated with distance, or the distance effect is confounded with readability). Dependent variables: ocular discomfort, completion speed, errors, the distance to which people voluntarily push or pull the panel.

Letting people set the distance and seeing where they stop is a strong contrast: the spontaneous stop is often near the screen, not near the hand. If the experiment gets angular size wrong, people pull the panel in only to make the type larger, and that stop measures type size, not the comfort band.

Where it stops holding

The band narrows with the session: a near distance acceptable at the start is not acceptable forty minutes later, so acceptance cannot run only the first five minutes. Headsets whose screen is already near (some optics sit a little past one metre) shift the whole band forward; a “put the UI at two metres” line written for a two-metre screen cannot be copied across. Directly manipulated objects must enter reachable space — that is an input constraint; they should not also be the reading surface. The readable layer steps back into the comfort band; the grabbable layer stays in the hand. A one-shot toast can leave the band briefly; a resident settings page cannot. People with weak accommodation have a narrower band; a product that sets the default from young participants’ stop points parks another group outside the band for good.

Applying it

• Look up the current device’s screen distance and put the default of the main panel near that screen, then trim for the task — do not start from “an arm” and bargain back.
• Split “to be read” from “to be grabbed”: grabables can be near; the reading surface returns to the comfort band. Do not let a thirty-centimetre settings wall carry both reading and selection.
• Offer a distance control, and put the factory default in the middle of the band. Log where people push the panel, and use that to change the default, rather than offering a control while leaving the default near.
• How to check: wear the main interface for the designed single-session duration, no doffing in the middle. Ocular items still below threshold at the end, and no spontaneous push of the panel farther away, is what it takes to say the distance landed in the band. Passing a five-minute demo does not count.

Related

• Same group: N1.04.1 Vergence distance and accommodation distance do not match · N1.04.2 The conflict causes eye strain and discomfort
• Nearby: N1.10 Binocular Disparity and Vergence–Accommodation Conflict · N3.02 Interface Distance
• Search terms: Percival zone of comfort · comfort viewing distance · HMD focal plane