Binocular capture: The role of non-linear position mechanisms

- Binocular capture varies with the vertical separation between monocular targets.
- Capture magnitudes are higher for feature-defined stimuli.
- Capture depends on the mechanism processing the relative offset.

When monocular Vernier targets are presented with binocular disparate elements, an increase in vertical separation elevates alignment thresholds and also shifts its perceived visual direction towards the visual direction of the binocular disparate surround. This observation has been termed binocular capture. There is increasing evidence that this shift in the visual direction of the monocular target may be related to the type of position encoding mechanism involved in processing the relative position signal. This study investigated the interaction between capture magnitude and vertical separation for stimulus conditions that favored the recruitment of linear or non-linear position encoding mechanisms. Relative alignment thresholds and bias were measured for a pair of vertically separated (8′, 30′, 60′, 120′) monocular Gabor gratings (1, 2, 4 and 8 cpd). Grating stimuli were constructed to constrain relative alignment judgments to the carrier grating (CO) or to the envelope (EO). Relative alignment thresholds and bias were also measured for a pair of vertically separated monocular Gabor gratings comprising a 1 cpd vertical square wave grating (SQ) or a 1 cpd missing fundamental grating (MF). Capture magnitudes were significantly larger across vertical separation and varied proportionally with relative alignment threshold for the EO and MF conditions. This was not evident with the CO and SQ conditions. The stark difference in capture magnitudes between the stimuli conditions suggest that the increase in capture magnitude observed with increasing vertical separation is intimately related to the transition from a “capture-immune” first-order spatial filter mechanism to a “capture-vulnerable” non-linear/feature-based position encoding mechanism.