Confusion prevails! a reply to kertesz and sullivan

Detection of vertical disparties

Psychometric curves determined for the detection of differences between vertical disparities as a function of the reference disparity for three presentation times (20 msec, 160 msec and 2 sec) and for two eccentricites of the stimulus (0.5 degree and 4 degrees) are presented. The results show that the detectability of disparity differences first increases and then decreases when the reference disparity is increased from zero. It further appears that the only effect of the stimulus presentation time and eccentricity on the psychometric curves and their dependency on the reference disparity is that the effective disparities observed are multiplied by a constant factor. The experimental data can be described by a signal detection model with two special features around zero disparity where binocular single vision occurs, viz. Loss of information about the sign of disparity and presence of intrinsic noise at a level up to 25% higher than the minimum level found at intermediate disparities.

The effect of presentation time on detection and diplopia thresholds for vertical disparities

Detection and diplopia thresholds for vertical disparities are found to decrease when the presentation time of the stimulus with disparity is increased between 20 msec and 2 sec. This result supports the following notions previously put forward in literature; (a) the human visual system embodies detection processes which reveal vertical disparities; there is no sensory fusion mechanism in this system that gradually conceals vertical disparities: (b) the detection of vertical disparity in tachistoscopically presented stimuli is not based upon the initial appearance of the stimulus with disparity.