If stereopsis is significant in in-depth perception, the disparity-defined depth should not change with distance. Studies demonstrating stereoscopic depth constancy, on the other hand, frequently showed systematic distortions in in-depth assessments across distance, particularly for virtual stimuli. By simulating actual objects in virtual equivalents, researchers hoped to learn more about how viewing distance and display-based cue conflicts affect depth estimation. They used virtual textured half-cylinders and identical three-dimensional (3D) printed equivalents at two viewing distances under monocular and binocular circumstances to quantify perceived depth. A mirror stereoscope and an Oculus Rift head-mounted display (HMD) were used to see virtual stimuli, while real stimuli were observed in a controlled test setting. 

Depth assessments were identical in both virtual apparatuses, suggesting that modifications in the HMD’s viewing geometry and optics did not affect perceived depth. When observing physical stimuli binocularly, judgments were correct and stereoscopic depth constancy was seen. However, with virtual stimuli, depth was consistently underestimated and depth consistency was not achieved. It was obvious that depth constancy was only perfect for cue-rich physical stimuli, and that the failure of constancy in virtual stimuli is caused by the presence of the vergence-accommodation conflict. In addition, the post-hoc study found that past experience with virtual and real surroundings had a significant impact on depth perceptions. That is, short exposure to a similar activity using actual items improved performance in virtual settings.

Reference:jov.arvojournals.org/article.aspx?articleid=2778683