Influence of Scanning Eye Movements on the Scintillating-Grid Illusion

Abstract

The scintillating grid is an example of a brightness contrast illusion in which bright disks are superimposed on a Hermann grid. Dark spots are perceived within the bright disks as flashing with each flick of the eye (Schrauf et al Vision Research in press). With steady fixation, stimulation must be brief for the illusion to occur. Varying exposure duration with steady fixation, we found the maximum illusion at durations between 210–350 ms (Schrauf et al, 1996 Perception25 Supplement, 78). In the present study voluntary saccadic eye movements were used to produce brief exposures during fixations. The task of 5 trained subjects was to scan scintillation grids of five different spacings (0.5 – 2.0 deg separation of grid elements, observation time 60 s) in a manner yielding a maximum scintillating effect. Eye movements were recorded binocularly at 100 Hz with the Demel Debic 90 infrared corneal reflection tracker (resolution: 6 min arc). For each fixation we calculated latency and position. Durations and amplitudes of the saccades increased with decreasing angular separation of grid elements. When the illusion was maximal, roughly 60% of fixation durations were between 250 and 550 ms. These durations overlap with those obtained with steady fixation combined with variable exposure durations. Unlike the Hermann grid illusion, where illusory dark spots can be perceived at very brief exposure durations, the scintillating grid illusion takes much more time to develop. The long optimal durations found here are discussed in relation to mechanisms of scanning eye movements.