Abstract
Sensitivity functions for horizontally- and vertically-oriented depth corrugations both peak in modulation spatial frequency at 0.2-0.4 cycles/degree, with higher sensitivity for horizontal than vertical corrugations (both defined by horizontal disparities). In order to elucidate the spatial frequency and orientation tuning of underlying processes, we analyzed two data sets using factor analytic techniques developed to estimate the number and tuning of spatial channels (Peterzell et al.,1993; 1995; 1996; 2000). The first set (Widdall et al., unpublished; 16 deg displays, n=30 individuals), was for 0.1, 0.2, 0.4, 0.8, and 1.4 (1.2 or 1.6) cycles/degree. A principal component analysis of disparity sensitivities (log arc sec) determined that two significant factors accounted for 70% of the variability. Following Varimax rotation to approximate "simple structure," one factor clearly loaded on to low spatial frequencies (0.4 c/deg and below), while the second was tuned to higher spatial frequencies (0.8 and 1.4 c/deg). Both factors had nearly identical tuning for horizontal and vertical patterns. In a second very small data set (Bradshaw & Rodgers, 1999; 20 degree displays, n=6 individuals), two or three factors accounted for 89% or 96% of the variability, with just one factor underlying the horizontal data, and at least two factors underlying the vertical data. The finding of separate factors for low and high spatial frequencies are consistent with previous studies (Witz & Hess, 2013; Serrano-Pedraza et al., 2013). The failure to find separate factors for horizontal and vertical corrugations in the larger data-set is surprising. Individuals' sensitivity to horizontal and vertical gratings were highly correlated at most spatial frequencies, even though the neuronal mechanisms are believed to be different, suggesting that sensitivity is limited mainly by the initial encoding of disparity.
Meeting abstract presented at VSS 2016