The increasing availability of 3D input and output devices demands a better understan ding and comparison of their quality. This paper describes an empirical experiment that provides quantitative results of the viewing quality of stereoscopic and perspective display modes. In this study such results accuracy and time of eightyone users were measured performing realistic 3D tasks to research the structures of organic molecules. A subsequent comparison of cost vs. performance can represent a meaningful help for decision-making of users and designers. The experiment was designed and conducted in cooperation of computer and chemical scientists. The experience of both sides was necessary to get a controlled testing environment with appropriate tasks. The considered display modes were perspective mode, anaglyph mode and shutter glass stereo mode. Additionally the effect of the level of expertise of a subject on accuracy and response time were explored. Mean response errors and mean response times were computed separately by a two-way analyses (ANOVA) for a series of six tasks and eighty-one subjects.

output devices, 3D, display mode, empirical evaluation, stereo perception