Abstract
When a row of 5 squares ranging from white to black are suspended in midair and illuminated by a spotlight thirty times brighter than overall room illumination the perceived range of gray shades is compressed relative to the actual range. That is, plotting lightness range against actual range produces a slope less than 1. According to anchoring theory, the lightness of a surface is a weighted average of its lightness within its local framework (spotlight) and within the global framework (whole field). The compression stems from the provision that, when the spotlight is 30X, all 5 squares are assigned the same value of white, because each has a luminance equal to (or exceeding) that of a white in the global framework. However, no further compression is predicted even if the spotlight becomes brighter, because their global values are maxed out. Purpose: (1) test the prediction that maximum compression occurs with a 30X spotlight; (2) test whether compression depends on absolute or relative illumination levels. Fifteen subjects matched 5 squares for lightness in each of 3 conditions. In Condition 1, the spotlight was 30X room illumination. The slope of the plot was .42 (white to middle gray). In Condition 2 spotlight intensity was increased by 4X, producing a significantly lower slope of .30 (white to light gray). In Condition 3, spotlight intensity was equal to that of Condition 1 but overall room illumination was reduced by 4X, producing a slope of .29. These results show that the compression depends on relative illumination, not absolute. But they also contradict the theory's prediction of maximum compression and appear to require a major modification of the theory.
Meeting abstract presented at VSS 2014