Accuracy Of Rendered Depth In Head-Mounted Displays: Role Of Eyepoints Location


Eyetracking is typically not available in head-mounted displays, and eye motions are thus simply ignored when 2D virtual images are displayed, giving rise to rendered depth errors in generating stereoscopic image pairs in head-mounted displays. We present an investigation and quantification of rendered depth errors linked to natural eye movements in binocular head-mounted displays, or Albertian errors, for three possible eyepoint locations: the center of the entrance pupil, the nodal point, and the center of rotation. Theoretical computations based on the intersection of chief rays concluded that, while the center of rotation yields minimal depth errors if no eyetracking is used, rendered angular errors may in some cases be significant (i.e. up to six degrees). Based on the analysis presented in this paper, we suggest that the center of entrance pupil be chosen for far field applications. The center of rotation of the eye should be chosen for near field applications under the assumption that they emphasize position accuracy versus angular accuracy. Preventing or minimizing rendered depth errors may be required for some high accuracy tasks related, for example, to medical or military visualization.

Publication Date


Publication Title

Proceedings of SPIE - The International Society for Optical Engineering



Number of Pages


Document Type

Article; Proceedings Paper

Personal Identifier


Socpus ID

0033714316 (Scopus)

Source API URL


This document is currently not available here.