Abstract
Coordinated binocular vision permits a clear fused image of objects at different depths. With a curved display, exacerbated change of vergence is needed for stimulus displayed at different parts of screen. Here we tested whether visual processing is affected by the change in image depth as determined by screen curvature and viewing position. Participants (n = 30, age 18 to 40) were asked to determine whether the opening of central supra-threshold Landolt C (threshold visual acuity + .2 logMAR size) was the same as a second near-threshold one (VA+.1 logMAR), with the stimulus size adjusted to viewing distances relative to various stimulus locations. They were seated at a perpendicular central position (3 or 4m distance) or one of three lateral off-perpendicular positions (15, 30, or 45° of viewing angle). The screen had one of five curvature radiuses while the target was presented at one of 24 screen locations. Results show that visual identification was poorer when the depth difference between the central and peripheral targets was larger, and when the viewing angle was larger. Different amplitudes of screen curvature and viewing distances had no direct effect on response accuracy after depth difference and viewing angle were controlled for. Therefore, an optimal curvature of screen surface for different screen sizes can be determined to allow the largest area of viewing angle/distance while maintaining good image perception, by calculating the resultant viewing angles and depth differences.