Abstract
Static images produce torsional eye movements known as optostatic torsion, which may be greater when adding spatial clues. In this study we investigated static tilted scenes, and images with an illusory tilt. Illusory images were of the Café Wall type. Eye movements were recorded in real-time, participants gazing straight ahead. Two participants sat with torso and head upright, supported by chinrest and forehead rest. After 20 seconds of darkness, stimuli were projected (4096 x 2160, 60Hz) at 90 cm. Tilted scenes were of the Lincoln Memorial, diameter 51° visual angle (86 cm). Illusory stimuli consisted of a black square 19.7 cm diagonal (12.4° visual angle) surrounded by 40 rows of black and white alternating tiles, 20 tiles/row, 51° visual angle. Thin gray lines separated the tile rows, as in the Café Wall illusion, producing a perception of tilt. Clockwise (CW) and counterclockwise (CCW) stimuli appeared for 60s, 4 times each, ±30° in alternating sequence. Illusory stimuli were always upright, configured to appear tilted CW or CCW. Eye movements were recorded at 100Hz by infrared camera (FLIR Grasshopper3 USB3), located 24 cm in front of the participant. Torsional eye position was calculated by tracking the iris pattern. Optostatic effects were replicated using tilted scenes. Average ocular torsion was -1.2 ± 0.02° for CCW stimuli, and 0.49 ± 0.48° for CW stimuli. Illusory stimuli did not produce appreciable differences in ocular torsion, comparing CW and CCW configurations. Average ocular torsion was -0.28 ± 0.16° for CCW stimuli, and 0.03 ± 0.13° for CW stimuli. Optostatic torsion is replicable in different experimental setups with various methods of eye tracking. Under comparable conditions, static illusory images did not produce comparable amounts of torsion. Future investigations should probe additional parameters of stimuli that produce illusory tilt phenomena.