Fig. 2.

Spatial-frequency-selective microsaccadic suppression in behavior. A: mean RT as a function of spatial frequency. On no-microsaccade trials (black), RT increased with spatial frequency, consistent with dependence of visual response dynamics on spatial frequency (Breitmeyer 1975). If the same gratings appeared ~20–100 ms after microsaccades (gray), RT increased relative to no-microsaccade trials (a behavioral correlate of suppressed visual sensitivity), but more dramatically for low rather than high spatial frequencies (compare gray and black curves at different spatial frequencies). B: difference in RT between microsaccade and no-microsaccade trials (i.e., difference between gray and black curves in A), demonstrating the diminishing effects of microsaccades on RT behavioral costs with increasing spatial frequency. C: difference in the likelihood of express RT trials between microsaccade and no-microsaccade trials, demonstrating diminishing effects of microsaccades on reducing the likelihood of express RTs. D–F: same analyses as in A–C but for a second monkey. G and H: time courses of mean RT (G; as in A) or difference in RT (H; as in B) as a function of the time of grating onset after microsaccade end. The time courses are from 2 sample spatial frequencies (complete time courses from all spatial frequencies, and for each animal individually, are also shown in Fig. 6). For the difference in RT time course, RTs on trials with no microsaccades within <250 ms from grating onset were taken as the baseline. The initial RT cost caused by microsaccades was weaker for higher spatial frequency gratings (compare vertical arrows, consistent with A). I: likelihood of express RT trials as a function of time after microsaccade end, for the same spatial frequencies as in G and H. Immediately after microsaccades, there was an express RT cost (i.e., fewer express RTs), with gradual recovery in time. Moreover, the recovery dynamics were different for different spatial frequencies, as with overall RT (G and H). Also, note that the baseline fraction of express RTs (i.e., long after microsaccades) was different for different spatial frequencies so that the recovery for different spatial frequencies is toward different absolute values (as in G). J–L: same analyses as in G–I but for a second monkey. Error bars, when visible, denote SE; n = 8,153 trials for monkey N, and n = 7,117 for monkey P.