Additivity failures in mesopic brightness matching.

Abstract

Additivity tests were carried out for a centrally-viewed 10° field using heterochromatic brightness matching at 6 retinal illuminance levels from 0.01 to 1000 photopic troland (Td). The results obtained by the mixture of 487 nm and 660 nm monochromatic stimuli showed two types of additivity failure. At the high levels of the retinal illuminance of 10,100 and 1000 Td, the brightness of the mixture was perceived to be lower than strict validity of the additivity law would predict. This is referred to as additivity failure of the reduction type. At the low levels of the retinal illuminances of 0.01 and 0.1 Td, the mixture appeared to be brighter than the brightness that the additivity law would predict. This phenomenon means additivity failure of the enhancement type. At the intermediate level of 1 Td, both types of additivity failure were seen. The type was determined by the luminance ratio of the two stimulus in the mixture. The results from another pair of mixtures, 463 nm and 567 nm, showed that additivity held at low levels, additivity failure of reduction type at intermediate level, and both types of additivity failure at high levels. The additivity failure observed at the mesopic levels are attributed to interaction between rod responses and responses from the red, green, and blue cones. These interactions are complex functions of luminance level.