Frank Durgin¹ (), Alexander Huk²; ¹Swarthmore College, ²University of California Los Angeles

It has been hypothesized that large aftereffects in the perception of visual density (and thus of number) may be primarily due to adaptation of early, density-sensitive channels of visual information. Most studies measuring density aftereffects are conducted on displays at a fixed distance from the observer, leaving it ambiguous as to whether the key representation is real-world (i.e., object-centric) density or retinal density. Here we dissociated these by measuring density aftereffects to adapting textures presented at a different distance than the tests. Two monitors were positioned at 60 cm and 120 cm from the observer, with the far display positioned above the near display. After establishing baseline density matches to 2 dots/deg2 on the near display, adapting textures (6 flashes prior to each test) were presented around a fixation mark on the far display, with fixation changing to the near display prior to the presentation of the test patches. The retinal density of the adapter on the far display was 4 times denser than the near display, but due to the distance difference, appeared to be a larger patch of similar world density. In a control condition, two virtual displays at the same distance (top and bottom of the near display) were employed using a retinally-matched adapter (also 4 times the test density, but now appearing much denser) in the top display with test patches in the bottom display, so that vertical refixation was again required between adapter and test. Observers (n=2) showed equally strong downward shifts in perceived density following adaptation to the distant patch (-46%) as to the nearer patch relative to baseline (-43%), with Weber fractions of 5 to 10% in all cases. Thus, an early representation, corresponding to retinal density, is the likely site of adaptation, with downstream effects on perceived density.

Acknowledgements: Swarthmore College Faculty Research Grant