Abstract
Purpose. Past studies using non-dichoptic viewing show that varying flank duration affects contrast detection threshold (CDT). We further investigated this effect by varying flank duration and the inter-stimulus interval (ISI) under monoptic and dichoptic viewing conditions. Methods. Four observers with normal vision participated. Target and both flanks were collinear, 3-cpd sinusoids, vertically oriented, separated center-to-center by 6-lambda (sigma = 1.5 lambda). We normalized flank contrast (3X flank CDT) and used a forward masking paradigm: flanks and ISI were each presented at 67, 117, 250 and 500ms resulting in 11 (including simultaneous presentation) stimulus onset asynchronies (SOA) ranging from 0–1000 ms. Target was presented to the dominant eye (250ms) and flanks to the dominant (monoptic) and non-dominant (dichoptic) eyes via mirror haploscope with septum. The task was 1-FC detection (yes/no) incorporating a 7-level MOCS. Each CDT was calculated via Weibull function from ≥700 trials. Results and Discussion. Simultaneous presentation produced CDT facilitation under monoptic (mean 19.0% ± 3.86(SD)) and dichoptic (mean 10.1% ± 2.46) viewing. SOA increase generally produced CDT increase. Monoptic mean facilitation ranged from 18.6% to 2.50% for the shortest and longest SOA, respectively. For the same SOA range, dichoptic viewing produced 11.8% facilitation to −8.31% (suppression). Best-fit functions to the relative CDT versus SOA results showed a simple, vertical shift downward for the dichoptic condition. This shift suggests similar mechanisms of contrast integration across space with additional gain reduction for the dichoptic mechanism. Analysis of the two SOA components showed that under monoptic viewing, longer flank duration tended to facilitate CDT more than longer ISI duration. Under dichoptic viewing these contributions to CDT appeared equal. Conclusion. Flank duration and ISI effects on contrast detection under dichoptic viewing appear consistent with current dichoptic, contrast discrimination (pedestal) models of contrast gain control.