Interaction between stimulus contrast and pre-saccadic crowding

Abstract

Objects that are briefly flashed around the time of saccades are mislocalized. Previously, robust interactions between saccadic perceptual distortions and stimulus contrast have been reported. It is also known that crowding depends on the contrast of the target and flankers. Here, we investigated how stimulus contrast and crowding interact with pre-saccadic perception. We asked observers to report the orientation of a tilted Gabor presented in the periphery, with or without four flanking vertically oriented Gabors. Observers performed the task either following a saccade or while maintaining fixation. Contrasts of the target and flankers were independently set to either high or low, with equal probability. In both the fixation and saccade conditions, the flanked conditions resulted in worse discrimination performance—the crowding effect. In the unflanked saccade trials, performance significantly decreased with target-to-saccade onset for low-contrast targets but not for high-contrast targets. In the presence of flankers, impending saccades reduced performance only for low-contrast, but not for high-contrast flankers. Interestingly, average performance in the fixation and saccade conditions was mostly similar in all contrast conditions. Moreover, the magnitude of crowding was influenced by saccades only when the target had high contrast and the flankers had low contrasts. Overall, our results are consistent with modulation of perisaccadic spatial localization by contrast and saccadic suppression, but at odds with a recent report of pre-saccadic release of crowding.

Pre-saccadic perception: Separate time courses for enhancement and spatial pooling at the saccade target

“…For example, those authors reported that pre-saccadic enhancement in performance was found also for un-flanked letter targets followed by a mask, consistent with a temporal rather than a spatial effect. Moreover, in a similar experiment to Harrison et al [ 27 ], Ağaoğlu and Chung [ 38 ] reported that during the pre-saccadic period the level of crowding was not in fact reduced compared to a fixation condition when different combination of high/low contrast target/flanker stimuli were used. On the contrary, saccades actually made crowding worse in some of the conditions tested.…”

“…The first main finding was to confirm that masking did indeed have a strong effect on orientation discrimination performance in this task. This was important in order to confirm the idea that the main influence of saccades on performance are primarily due to un-masking rather than un-crowding [ 36 ], [ 37 ], [ 38 ]. Second, we found a general pre-saccadic improvement in performance (significant in 4 out of 6 conditions) that was present already quite early and remained constant throughout the time course of the pre-saccadic time period.…”

“…Ağaoğlu and colleagues [ 37 ] also found a small enhancement on saccade trials, similar to that found here. Moreover, in a recent study, Ağaoğlu and Chung [ 38 ] tested the influence of eye movements on crowding with different combinations of high/low contrast target/flanker stimuli and did not observe any evidence of a release of crowding in any of the conditions tested. In their paper, the authors supported a framework whereby the pre-saccadic modulations are mostly driven by spatial factors, such as localization of the stimuli and saccadic suppression.…”

Coupling perception to action through incidental sensory consequences of motor behaviour

“…The kinematic properties of visual actions might also contribute to spatial perception. Evidence for this potential link comes from trans-saccadic localization and the study of crowding [152][153][154][155] , a bustling domain of vision research that, with few exceptions [156][157][158][159][160] , studies peripheral vision exclusively during visual fixation.…”