Compression of space as a default for localizing degraded targets in the context of highly visible stimuli.

The visual input reaching our perceptual system is frequently disrupted: Objects may be temporarily occluded, the lighting might change abruptly, and drastic changes in the retinal image occur every time we move our eyes. Some phenomena, such as change blindness (Rensink, O'Regan, & Clark, 1997), demonstrate the challenges of such visual disruptions for perception. In addition, previous research has reported dramatic localization errors around the time of visual disruptions: Stimuli briefly flashed just before a saccade are perceived closer to the saccade target, a phenomenon known as perisaccadic compression of space (Ross, Morrone, & Burr, 1997). More recently, we have demonstrated that brief probes are attracted towards a visual reference when followed by a mask, even in the absence of saccades (Zimmermann, Born, Fink, & Cavanagh, in press). Here, we ask whether, akin to change blindness phenomena, a transient visual disruption is a critical factor in compression. Alternatively, the decrease in visibility of the probe caused by the saccade or mask may determine compression. We used our mask-induced compression paradigm and varied the regions of the screen covered by the transient mask, including an area centered on the probe, areas where no stimulus was presented, and we added a condition without masking. In all conditions, we adjusted probe contrast to make the probe equally hard to see. Compression effects were found in all conditions, even without the mask. To obtain compression without a mask, the probe had to be presented at much lower contrasts than with the mask, though. Thus, compression of space can also be found with low contrast stimuli, without a transient visual disruption. We suggest that compression reflects how the visual system deals with degraded target onsets in the context of highly visible stimuli that share some aspects with the target. Meeting abstract presented at VSS 2015.