Subjective Impression of Discomfort Glare from Sources of Non-Uniform Luminance

A paired comparison experiment was performed to study the effects of spatial frequency and position on discomfort glare. The experiment required the subject to choose which of two stimuli, when presented together, caused more discomfort ( termed the "choice" data), and then also rate how much more discomforting the one is than the other ( termed the "magnitude" data). A rating scale experiment was also performed. Significant Findings: 1. The results show that both position and spatial frequency are significant predictors of discomfort. As position increases above the line of sight, glare decreases. As spatial frequency increases, discomfort increases, with a uniform stimulus causing the most discomfort ( where "uniform" is defined here as infinite cycles per degree). 2. The results show that the effect of spatial frequency is more pronounced at positions closer to the line of sight. 3. The findings here contradict the CIE's complex extension to the Unified Glare Rating (UGR). The CIE complex extension clearly states that a complex source will always predict more glare than the uniform one, which is in direct opposition to the results found in the paired comparison experiment discussed here. 4. A relationship between contrast sensitivity and discomfort sensitivity was found. Contrast sensitivity peaks at a particular frequency, and it drops on either side of that peak frequency ( at any given eccentricity). That peak frequency decreases as eccentricity increases ( Fig. 1). This research shows that discomfort is minimized at a particular frequency, and it increases on either side of that minimum frequency ( at any given position), which is contrary to the effect of contrast sensitivity. 5. For each position, the lowest glare perception occurred for the frequency which has approximately 3 receptive fields per cycle ( see Fig. 23 through Fig. 26). 6. The paired comparison technique gave data with much less variability than the rating scale technique, as can be seen from the size of the error bars in Fig. 9 and Fig. 16. It is the recommendation of the authors that researchers consider using a paired comparison experimental technique for glare research in the future.