Abstract
The luminance contrast produced by a reflectance edge was judged as apparently equal to almost three times higher luminance contrast produced by an illumination edge (Logvinenko, AIC, 2005). Such illumination contrast discounting was subject to large inter-individual variability. As lightness constancy also exhibits considerable individual differences, we decided to ascertain whether lightnessconstancy and illumination contrast discounting are related to each other.
The stimulus consisted of a large sheet of black paper with a rectangular spotlight projected onto the lower half and 40 squares of different gray shades printed on the upper half. The luminance ratio at the edge of the spotlight was 25:1 while that of the squares varied from 2:1 to 16:1.
Observers were asked to find a square in the upper half which (1) had the same luminance contrast as the illumination edge (contrast match) and (2) which looked as if it was made of the same paper as that on which the spotlight fell (lightness match). Each observer made ten matches under each instruction.
The median luminance ratio for the contrast match was 9.2. Hence, the luminance contrast produced by illumination was underestimated by a factor 2.75. While the tasks were different, the median match for the group of 40 observers was the same for both, Friedman's ANOVA showing no significant effect of the task (Chi-sq=1.04, df=1, p=0.31). Only for 14 observers the lightness match significantly differed from the contrast match according to a Kruskal-Wallis test (p [[gt]] 0.05). The Brunswick ratio for the lightness match varied from 0.44 to 0.88, the median being 0.68. A perfect correlation was found between the two indices: the higher the Brunswick ratio, the higher the illumination contrast discounting.
EPSRC EP/C010353/1 (ADL).