Fig. 9: Selective and unselective saccadic suppression measured using full psychometric curves.

a We repeated the real saccade experiments of Fig. 8, and obtained full psychometric curves by using different Gabor grating contrasts (Methods). Different colors indicate different spatial frequencies of the flashed gratings. When the gratings were flashed ~42 ms after saccade onset (Methods) and there was a coarse surround texture, perceptual suppression clearly depended on spatial frequency: detection thresholds were highest for the lowest spatial frequency, and they progressively decreased with increasing spatial frequency. Each curve shows mean ± s.e.m. of four subjects’ psychometric curves. Dashed psychometric curves show perceptual detectability without saccadic suppression (obtained similarly to Fig. 8). b When the surround context was fine, rather than coarse, perceptual suppression was not selective for low spatial frequencies (consistent with Fig. 8). c Detection thresholds from a, b as a function of grating spatial frequency for flashes ~42 ms after saccade onset. With a coarse surround, detection thresholds were highest for low spatial frequencies and progressively decreased with increasing spatial frequency (1-way ANOVA, p = 0.0168, F = 6.6608; p = 0.0133 for post-hoc comparison between lowest and highest spatial frequency, indicated by *). With a fine surround, detection thresholds did not depend on spatial frequency. d Same as in c but now for grating flashes occurring ~65 ms after saccade onset. For both surround textures, detection thresholds decreased, indicating perceptual recovery. There was still a trend for dependence of perception on spatial frequency in the coarse condition, consistent with c.