Abstract
When a pattern of broadband spatial content is viewed, the multiple spatial components in the pattern stimulate detecting-mechanisms that suppress each other. We've shown that this suppression is anisotropic, being relatively greater at horizontal and least at obliques (Essock et al., VisRes 2003; Essock et al., JOV 2009). This suppression leads to better thresholds and greater saliency of obliquely-oriented structure when viewed in the presence of broadband or natural-like structure. Thus this anisotropic suppression (the “horizontal effect”) appears to “whiten” the neural response to typical scenes (Hansen and Essock, JOV 2004; Essock et al., JOV 2009). We have measured the timing of this anisotropic suppression by varying the stimulus onset asynchrony (SOA) between a brief (50 ms) test pulse of a (spatial) Gabor and the spatial mask (50 ms pulse of 1/f broadband noise pattern) for different test orientations (Essock et al., JOV 2009). With an 8 cpd test, masking is greatest at 0 ms SOA and with backward more effective than forward masking. Modeling of the temporal properties of the masking suggested that at least two suppression parameters underlie the performance anisotropies: a fast horizontal vs vertical bias, and a slow cardinal vs oblique bias. Here, we evaluated the time course of anisotropic broadband masking with test spatial frequencies of 2, 8, and 12 cpd to assess the suppression anisotropies at different speeds. Results show strongest suppression at horizontal at all three test spatial frequencies, however, the time-course of the anisotropic suppression differs. Suppression is maximal at the 0 ms SOA, but at the low spatial frequency, suppression is brief with little temporal asymmetry. That is, for transient, “high-speed” conditions (Meese and Holmes, ProcRoySoc 2007; Essock et al., JOV 2009) the anisotropic suppression follows a different time-course than at the relatively sustained “low-speed” conditions.