Lighting quality evaluations using images on a high dynamic range display

Limited research comparing participant ratings of luminous environments to ratings of images of those environments indicates that images can be a reasonable surrogate for the real space, particularly on ratings related to aesthetics. However, the realism of such images when presented on computer screens is potentially limited by conventional display technologies that cannot reproduce the full range of luminances in real spaces. In this pilot experiment we used a new, high dynamic range (HDR) computer monitor capable of producing screen luminances and contrasts comparable to those in a real space. Fifty-four participants viewed three images of a conventional office in two display modes: HDR monitor and conventional monitor. Participants rated each image for room appearance, environmental satisfaction and realism. These ratings were also compared to similar ratings made by participants in an earlier experiment (reported in 1998) who occupied the real spaces depicted in the images. Results indicate that computer screen images are perceived in a similar way as real luminous environments. HDR images are perceived differently than images on a conventional monitor: they are rated as brighter and less attractive, as expected. Given their more authentic luminances, HDR images should be perceived as more similar to the real space, but our results neither support nor refute this. Introduction The traditional method of exploring preferred luminous conditions involves participants evaluating full-scale physical mock-ups of spaces lit in different ways. While high in external validity (the fit between the experimental condition and a real world setting), these studies are expensive, especially if manipulation of the lighting design is desired. This is not only a drawback for the researcher: Lighting manufacturers and designers who wish to present design solutions to their clients are often required to create expensive physical mock-ups. Partly as a response to this, there has been some interest in other, cheaper presentation methods, such as scale models, photographs, or renderings from computer simulation packages. Researchers in areas such as forestry and architecture [e.g. Daniel & Meitner, 2000; Danford & Willems, 1975] have established that images can be a reasonable surrogate for the real space, particularly on ratings related to aesthetics. The limited research in lighting on this topic concurs with this, when representing the real space with photographs [Hendrick et al., 1977], or with detailed simulations [Eissa & Mahdavi, 2001].