Cite as:
Banks M S, Backus B T, 1998, "Cyclovergence, extraretinal signals, and stereopsis" Perception 27 ECVP Abstract Supplement

Cyclovergence, extraretinal signals, and stereopsis

M S Banks, B T Backus

When a frontal surface is slanted about a horizontal axis, horizontal-shear disparities are altered, but vertical-shear disparities are barely affected. Opposed torsional eye movements (cyclovergence) create changes in horizontal and vertical shear. Howard, Rogers, and others have shown that changes in horizontal shear due to cyclovergence are not perceived as surface-slant changes because the induced vertical shear corrects the induced horizontal shear. This resistance to cyclovergence could also be accomplished by using extraretinal signals to correct induced horizontal shear. Does this means of correction exist?

Observers fixated a large cyclorotated pattern which produced cyclovergence (verified by nonius method). After 10 s, the pattern was replaced briefly by a stereoscopic surface with smooth vertical lines; vertical shear cannot be measured with this stimulus. Observers adjusted the surface slant (horizontal axis) until it appeared gaze normal. The experiment was repeated with random-dot stimuli (allowing vertical-shear estimation).

Slant settings were expressed as horizontal shear at the retina. With the vertical-line stimulus, settings varied systematically with cyclovergence. The change in proximal horizontal shear that was accepted as gaze normal was half the cyclovergence, so the gain of extraretinal correction is -0.5. With the random-dot stimulus, settings (expressed as retinal horizontal shear) varied with cyclovergence with a gain of -1 because vertical shear was available.

Two means of correcting retinal disparities for cyclovergence exist. One compares horizontal-shear and vertical-shear disparities (shown previously); the other uses an extraretinal estimate of the eyes' cyclovergence.

These web-based abstracts are provided for ease of seaching and access, but certain aspects (such as as mathematics) may not appear in their optimum form. For the final published version of this abstract, please see
ECVP 1998 Abstract Supplement (complete) size: 2107 Kb