Perception of Spatial Color Variation Caused by Mass Variations about Single Separations

Abstract

When developing image quality specifications for a printer or copier, it is important to understand people's threshold and sensitivity for spatial color variation. Much information exists in the literature for thresholds for lightness variations about a neutral base color. There is much less information about thresholds for color variation about a neutral base color. The general question of thresholds for spatial variation in any direction in color space about an arbitrary base color is very complex and has yet to be answered. This work is part of the general study.In tracking the behavior of a printer against process changes, it is useful to examine the quality of printed single separations. Visual thresholds to color variations representative of those caused by mass variations about 30% tints of cyan, magenta, yellow and black have been determined from psychophysical tests. Thresholds were determined in each of 3 common color difference metrics (ΔECIE-Lab, ΔECMC 2:1 and ΔECIE-94) as a function of spatial frequency from 0.02 to 2.0 cycles/mm at normal reading distance of ∼40 cm.The actual ΔE values depend heavily on which metric is used, varying by a factor of two or more. In any of the 3 ΔE's used, there is a broad peak in sensitivity to spatial variation between 0.1 and 0.7 cycles/mm, with perceptibility thresholds in the neighborhood of ΔE∼0.2 (twice as high for ΔECIE-Lab). The exception is yellow, for which the peak is between 0.05 and 0.2 cycles/mm and the perceptibility threshold is ΔE∼1 (again twice as high for ΔECIE-Lab). This difference is because a yellow mass variation creates only a chroma variation, whereas for the other colors there is a significant lightness variation.