The hypothesis is offered that relative brightness judgments are based upon experience with the way the amount of light reflected by objects changes with their distance from the light source.
The experimental results support this hypothesis. One group of 40 subjects judged half brightness, and another 40 estimated the change in luminance corresponding to moving a hidden point light source to twice the distance from an illuminated standard field. Judgments of the two groups were equivalent.
Under stimulus conditions designed to represent the common visual situation (stimuli subtending wide visual angles, adaptation approximating stimulus levels), one quarter the standard luminance was correctly chosen for the effect of doubling distance from the source, and the same fraction was chosen for half brightness for all standard intensities (0.00086 to 87 millilamberts).
Under less familiar conditions similar to those employed for the bril scale (small stimuli with black backgrounds, indeterminate levels of adaptation) half brightness judgments were again equivalent to estimates of the effect of doubling distance from object to light source. These estimates were less than one-quarter standard luminance.
The hypothesis is discussed in terms of sensory scaling in general, and the neutral value and bril scales in particular.
RICHARD M. WARREN and ROSLYN P. WARREN, "Basis for Judgments of Relative Brightness," J. Opt. Soc. Am. 48, 445-450 (1958)