IPD mismatch causes blur and fatigue

What it is

The optical centres of the two headset lenses sit 68 millimetres apart; the wearer’s interpupillary distance is 58 millimetres: each pupil is aimed at the outer ring of its lens, and the sweet spot is not on. The picture is milky, there is flare, one eye is softer than the other, and after a while the eyes ache — that is IPD mismatch. It is an optical problem first: the eye is not travelling the optical axis the lens was designed for. Adult IPD runs from the mid-fifties of millimetres into the seventies; a headset with no adjustment, or with only two detents, parks a large fraction of people outside the sweet spot for good.

Changing “IPD” in software without moving the lenses moves camera separation. It does not move this optical misalignment. Blur and fatigue come from a sweet spot that was never fitted, not from the number in the menu.

Why it happens

Headset lenses (Fresnel or pancake) keep the sharp, low-distortion region near the optical axis. A few millimetres off-axis, field curvature, chromatic aberration, and flare rise together. A pupil off the axis is looking through the poor part of the lens. The two eyes usually deviate in opposite directions (too-wide lenses send both pupils outward; too-narrow, both inward), so sharpness and distortion fail to match between the eyes, and fusion takes extra work. Ciliary muscle and vergence fight over which eye’s image counts as clear; fatigue is reported within minutes.

Mechanical adjustment moves the lens spacing so the axes can chase the pupils. With no mechanical adjustment, the only move left is to slide the whole machine on the face — which spends fit and light leak, and barely changes relative spacing, because binocular distance is anatomical. So “it will clear up as you wear it” usually does not hold for an IPD error.

Studying it

Measure the person’s IPD (a PD ruler or a camera method), measure the headset’s current mechanical lens spacing, and take the mismatch in millimetres. Tasks: sharpness judgments, reading, and ocular-fatigue items after timed fixation. On a machine with mechanical adjustment, pair “aligned / deliberately one detent off.”

Independent variables: difference between mechanical IPD and the person, presence of software-separation compensation (as a control only, not to explain size perception), fixation duration. Dependent variables: subjective sharpness, contrast sensitivity or reading speed, eye strain, whether “which side is softer” can be named.

Split “I cannot see clearly” from “I cannot keep it on my face.” The first still holds when seated with the headset on a fixture, and that is IPD; the second is fit. Children have smaller IPD, and adult-spec machines systematically mismatch them; a sample of lab young adults is not enough.

Where it stops holding

Uncorrected refractive error stacks blur onto mismatch and the two cannot be split. Wear the person’s own prescription before talking IPD. A monocular wearer has no binocular fusion load; mismatch shows mainly as monocular sharpness, and the fatigue path is weaker. IPD changes slightly with fixation distance (a small inward shift at near); a headset set to distance IPD adds a little extra error on near work, usually smaller than an unadjusted detent. Some people have a vertical IPD difference; horizontal spacing alone will not align, one eye stays sharp and one stays soft, and it gets reported as “the machine is broken.” Extreme IPD (below the adjustable floor, above the ceiling) is not a “finer adjustment” problem; it needs a different optical module, or a stated non-coverage.

Applying it

• At factory and for every new user, do a mechanical IPD alignment and accept on sharpness, not on the menu number: have them look at a fine central line and adjust left–right until it is sharpest. A machine that only offers a software slider must say in copy that the optical sweet spot does not follow the slider.
• Set the adjustable range from the IPD distribution of the intended population, not from “typical adult male.” If children or small-face users are in the support list, a floor that does not reach them means they are not in the list.
• In the tutorial, use “which side is softer” as a self-check; it is harder to fill in wrong than “please enter your IPD in millimetres.”
• How to check: find someone whose IPD clearly misses the default detent, and have them read the same fine type for a minute on default and on aligned. If aligned is sharper and less tiring, mismatch is being explained by this path. If both are equally soft, look at refraction and dirty lenses; do not keep twisting IPD.

Related

• Same group: N1.07.2 Compatibility with prescription glasses · N1.07.3 Facial size differences affect light leak and comfort
• Nearby: N1.08 Headset Mass and Prolonged Wear · N1.10 Binocular Disparity and Vergence–Accommodation Conflict
• Search terms: IPD mismatch · optical sweet spot · interpupillary distance