Abstract
Visual discomfort consists of the adverse effects of viewing certain stimuli such as stripes of around 3cycles/degree (Wilkins et al, 1984, Brain, 107, 989-1017), or more generally images whose amplitude spectra deviate from the 1/f profile that characterises typical natural images (Juricevic et al, 2010, Perception, 39, 884-899). This study was conducted to determine whether the presence of uncomfortable stimuli has a negative effect on performance in a visual search task. We predict that, when presented with an uncomfortable background, observers will have a shorter reaction time to decide that a target is absent, so as to terminate the trial. Stimuli were initially vertical Gabor patches (of either 3 or 0.75 cycles/degree) presented against a random 1/f filtered noise patch, which subtended approximately 8 degrees. This was surrounded by either a mid-grey background, or a vertical sinusoidal grating of either 0.75, 1.5 or 3 cycles/degree that was irrelevant to the task. When the frequency of the grating matched that of the target, accuracy reduced and reaction times increased, in comparison with performance against a mid-grey background. The detrimental effect on performance was greatest when the Gabor and the background were closely matched for spatial frequency, i.e. it showed clear spatial-frequency tuning. This indicates that the effect is due to interference between the target and background. This was further supported by the fact that, when a horizontally-oriented target was used, spatial-frequency tuning was not found, and performance was no worse than that against a mid-grey background. It is concluded that the detrimental effects on performance did not reflect the apparent discomfort of the stimuli, but instead are due to interference between the target and background. This indicates that uncomfortable stimuli do not affect performance on a visual-search task, in a sample drawn from the general population.
Meeting abstract presented at VSS 2012