An Experimental Exploration of the relationship between Subjective Impressions of Illumination and Physical Fidelity

Two experiments were carried out to explore the effect of rendering and interface fidelity on subjective impressions of illumination and perceived presence after exposure to a virtual environment (VE). In particular, a study that compares a real-world task situation to its computer graphics simulation counterpart is presented. The computer graphics simulation was based on photometry data acquired in the real-world space and was displayed on either a Head Mounted Display or desktop display utilising either monocular or stereo imagery and interaction interfaces such as the common mouse and head tracking. 105 participants across five conditions were exposed to the real and computer graphics environment and after completing a spatial task, subjective impressions of the illumination and sense of presence assessments were acquired. Relevant results showed a positive correlation between presence and subjective impressions of lighting (e.g. warm,comfortable,spacious, etc.) for the HMD monocular conditions. For the second study, the computer graphics scenes were rendered in varied levels of shadow accuracy utilising flat-shaded and radiosity rendering and were displayed on a stereo, head tracked Head Mounted Display (HMD). A total of 36 participants across three visual conditions were exposed to the scene and after completing a spatial task, subjective impressions of the illumination and sense of presence assessments were acquired. Relevant results showed a positive correlation between presence and subjective impressions of lighting (e.g. warm,comfortable,spacious, etc.) associated to the high-quality, full-shadow accuracy rendering condition. There was no effect of viewing condition upon subjective impressions of illumination for both studies, because of constant luminance levels. How real-world impressions of illumination could be simulated in a synthetic scene is still an open research question.