Spatial scaling of the binocular capture effect

Binocular "capture" occurs when the perceived visual direction of a monocular stimulus is displaced in the direction of the cyclopean visual direction of nearby binocular targets. This effect increases with the vertical separation of broadband monocular stimuli. The present study investigated whether the "capture" effect exhibits a systematic relationship with the spatial frequency composition of monocular lines and vertical separation.

Subjects judged the horizontal misalignment of 66 arc min vertical spatial frequency ribbons that were temporally interleaved with a random dot depth edge (3.2 degrees) for 108 ms. Spatial frequency ribbons were constructed from horizontal cosine gratings windowed by a 4 arc min vertical Gaussian envelope. The bottom half of the depth edge was presented with zero relative disparity, whereas the top half was presented with 10 arc min of crossed or uncrossed relative disparity. Four vertical separations (8, 16, 30, and 60 arc min) and three ribbon spatial frequencies (1, 4, and 8 cpd) were tested. The horizontal ribbon offset corresponding to 50% performance was calculated for each combination of depth condition, ribbon spatial frequency, and vertical separation.

The magnitude of the "capture" effect was consistently larger for higher spatial frequency ribbons and decreased with decreasing vertical separation. When vertical separation was expressed as multiples of spatial periods of the respective ribbon spatial frequency, the magnitude of effect was significantly larger for separations greater than about one spatial period.

The systematic scaling of the "capture" effect with spatial frequency and vertical separation is strongly suggestive of the operation of multiple spatial scale mechanisms; similar to those advocated for the processing of relative positional acuity with increasing vertical separation of monocular targets.