Failure of binocular fusion produces diplopia or monocular suppression

What it is

The ability to combine two slightly different images from the two eyes into a single percept — fusion — is not unlimited: when the disparity between corresponding points in the two eyes is too large, or the two lines of sight aren't actually pointed accurately at the same target, fusion fails outright, and one of two typical outcomes follows. Either the user genuinely sees two separate, misaligned images — diplopia (commonly called double vision) — or the visual system actively discards one eye's input and keeps only the other eye's image — suppression — in which case the user has no awareness of "missing half the picture" and simply perceives a single, normal-looking image. Both outcomes are different strategies the system adopts once fusion has completely broken down, not fusion continuing to operate in some degraded form.

Why it happens

Binocular fusion can turn a range of disparities into a single, stable stereo-depth percept, but that ability is confined to a limited disparity range known as Panum's fusional area. Disparity within that range gets normally fused into a corresponding depth sensation; once disparity exceeds that range, the fusion mechanism simply cannot process it — it doesn't force the excess disparity into a wrong-but-single depth result. From there, one of two things happens. If the disparity overshoot is mild or intermittent, the two eyes' images get presented separately and independently, and the user subjectively sees two misaligned, overlapping images — diplopia. If the disparity overshoot is persistent and stable (for example, the two eyes' visual axes chronically fail to align on the same target, a condition called strabismus), the ongoing double vision itself causes significant visual interference, and to avoid enduring that interference long-term, the visual system actively and chronically suppresses the input from one eye (typically the one that's persistently misaligned), removing it from awareness and leaving only the other eye's single image. The user therefore doesn't perceive double vision, but the cost is that the suppressed eye contributes essentially nothing to the final percept for as long as suppression lasts, and the stereo depth information binocular vision would otherwise provide is lost along with it.

Studying it

Clinical vision science typically measures the fusional range with a prism bar test: progressively increasing base-in or base-out prism power in front of a participant's eye to artificially induce a gradually growing disparity, and recording the prism power at which the participant transitions from "still maintaining a single fused image" to "starting to see double" — that transition point marks the boundary of that participant's current fusional range. Monocular suppression is commonly assessed with a red-green dichoptic test or a similar dichoptic-detection method, checking whether a participant can simultaneously detect visual information that should be delivered separately by each eye, or whether part of it is systematically undetected. Typical independent variables are the induced disparity magnitude (prism power) and the duration disparity is presented; typical dependent variables are the disparity threshold at which diplopia appears and the detection rate of the suppressed eye's information in a suppression test.

Where it stops holding

• The fusional range itself varies between individuals, and even within the same person it fluctuates with viewing distance and visual fatigue state, so there's no single disparity "safety value" that applies to everyone in every scenario.
• Diplopia and monocular suppression have something of a trade-off relationship: mild, occasional disparity overshoot more often shows up as transient diplopia, while persistent, stable disparity overshoot — especially a binocular-alignment problem present since childhood — more often develops into chronic suppression. The two aren't fully independent, unrelated paths, but which outcome a given individual ends up with also depends on factors like developmental history, so it can't be generalized.
• This entry concerns fusion failure caused specifically by disparity exceeding the fusional range, which is a different starting mechanism from visual discomfort caused by a broken coupling between accommodation and vergence — the two shouldn't be conflated.

Applying it

• When setting the disparity range for stereo display content (VR, 3D screens, binocular augmented-reality overlays), treat the range users can comfortably fuse as a hard constraint, not just a target to maximize visual impact — negative or positive parallax that exceeds this range can make some users see outright double text or double interface elements, not just mild discomfort.
• If a binocular augmented-reality device has a calibration error between the left- and right-eye rendering or optical path, even a small error can be perceived as slight double vision or visual fatigue on static content that requires prolonged fixation (text, icons); calibration-precision requirements should be set against the scale of the fusional range, not estimated by feel as "good enough."
• Verification: have participants fixate on static stereo-rendered text or interface elements on the target device for an extended period and directly ask whether they see double, combined with a before/after comparison of covering one eye (if the double vision disappears when one eye is covered, that confirms it's a binocular fusion issue) to pin down whether the cause is the display's disparity setting or a left/right calibration error.

Related

• Same group: A1.33.1 Conflicting binocular input can produce rivalry instead of fusion · A1.33.2 Most people have a dominant eye, and aiming/pointing defer to it · A1.33.3 Occluding one eye does not switch off its input, and rivalry can still occur
• Nearby: A1.13.3 Binocular Disparity Is Only Effective at Close Range · A1.21 Lens Accommodation and Vergence
• Search terms: diplopia · suppression · Panum's fusional area · strabismus