How well are color components of samples of the Natural Color System estimated?

The aim of this study was to determine how accurately color-normal subjects that have received basic information about, but do not have practical experience with, the Natural Color System (NCS) can estimate the Heringian components of a representative selection of samples. Twenty-five color-normal subjects, taking part in two trials with at least a 24 h gap between assessments, selected four samples representing individual unique hues (uHs) from a set of 40 highly chromatic NCS samples on a rotatable tray. The samples selected for assessment of components were displayed to the subjects who estimated the hue components of 16 high-chroma samples, hue and white/black components of 16 tonal color samples, and three achromatic samples with different blackness values. Variability in selection of samples representing uHs as well as the relationship between the subjects' estimates of unique hue components and the defined values of the system was obtained. It was found that hues alone are easier to correctly estimate than hues together with white and black and that the components of colors of higher chroma are easier to estimate accurately than those of lower chroma. It was also found that, for R and G, the mean uH choices of subjects differed very little from the NCS's R and G, whereas selections for yellow and blue deviated, the former by 1.22 hue steps (slightly greener than G90Y), and the latter by 1.36 hue steps (represented approximately by R85B). This may impact the accuracy of color models that employ NCS unique hues. (C) 2011 Optical Society of America