Abstract

Brown and MacAdam’s experimental procedure, based on color-matching variability, is a good laboratory model for predicting small color differences of foveated objects. Their analysis played an important role in the development of the CIELUV and CIELAB metrics which specify color tolerances. With the increasing use of color monitors to display information and enhance human performance, the same color metrics have been called on to guide the selection of color differences in visual tasks that require somewhat larger color differences (medium color differences). Since Brown and MacAdam’s seminal work, a number of investigators have investigated medium color difference tasks using a variety of performance criteria (reaction times, probability correct discrimination) and spatial configurations (position uncertainty, retinal location). I review the mathematical representation of metric color differences used by Brown and MacAdam. I describe medium color difference tasks in which performance can be described using Brown and MacAdam’s metric representation, permitting us to compare small and medium color differences. Finally, I describe some tasks requiring medium color differences in which performance is qualitatively different from performance in small color difference judgments.

© 1988 Optical Society of America