Abstract
What happens during the first milliseconds of object recognition? If a vernier is followed immediately by a second vernier with identical spatiotemporal parameters, but opposite offset direction, performance is at chance level, i.e. offsets cancel each other. However, if a grating with 25 aligned elements follows after the second vernier the first displayed vernier determines performance. Subjectively, only one vernier is perceived clearly shining through the grating (shine-through effect; for gratings with less than seven elements the verniers remain invisible). With increasing offset sizes of both preceding elements dominance of the first vernier increases. We determined the presentation time of the second vernier, in relation to the first one, for which offsets cancel each other leading to performance at chance level. This point of temporal equivalence is constant for all offset sizes of the verniers. However, presentation time and offsets size covary strongly if either one of the verniers dominates performance. Hence, temporal equivalence represents a singularity. The presentation time of each of the two verniers is quite short (smaller than 20ms) and gratings last for 300ms. Dominance of the first vernier arises with grating durations around 40-60ms. At this point of time the first vernier has disappeared quite a long time. Where is the offset of the first vernier stored? In summary, the shine-through effect allows detailed analyses of visual processing during the first milliseconds.