Abstract
Visual discomfort describes the adverse effects, such as headaches, eye-strain and illusions, that can be experienced when viewing some visual patterns. The ability to measure discomfort reliably has many important applications, for example in the development of three-dimensional displays, and the understanding of migraine. Previous research has used a variety of methods, including rating scales, to measure discomfort. Typically, an observer is presented with a stimulus and asked to judge (e.g. on a scale of 1 to 10) the degree of discomfort experienced. One limitation of such techniques is that there is an unknown mapping between the experience of discomfort, and the scale values reported. In the current study, we assessed whether maximum likelihood difference scaling (MLDS) could be used to quantify discomfort. In the first experiment, observers were presented with simple Op Art pictures, and asked firstly to judge whether they were uncomfortable, and secondly to quantify the degree of discomfort. In the second experiment, observers were presented with pairs of pairs of these stimuli. They were asked to judge which pair was more similar in (i) spatial frequency and (ii) discomfort. Their responses were used to create perceptual scales for spatial frequency and discomfort using MLDS. In the first experiment, discomfort depended on spatial frequency. In the second experiment, when asked to base judgements on perceived spatial frequency, MLDS uncovered a perceptual scale that varied with the log of spatial frequency. However, when asked to base judgements on perceived discomfort, 5 of the 8 observers failed to consistently carry out the task. These results show that observers are readily able to separate judgements of discomfort from judgements of other stimulus dimensions (in this case spatial frequency), but that the measurement of visual discomfort presents a significant challenge to MLDS.
Meeting abstract presented at VSS 2014