A Comparison of Two Hue Scaling Methods

Abstract

Hue scaling functions are designed to characterize color appearance by assessing the relative strength of the red vs. green and blue vs. yellow opponent sensations comprising different hues. However, these judgments can be non-intuitive and may pose difficulties for measurement and analysis. We explored an alternative scaling method based on judging the relative similarity of each hue and identifying where it belongs within the labeled opponent categories. The two methods were compared for 29 observers. Settings on both tasks were comparable though the similarity ratings showed weaker categorical bias and less inter-observer variability. The concordance suggests that the functions measured by the two paradigms reflect properties of color appearance rather than properties of the tasks. Individual differences on both tasks suggest that color appearance depends on multiple, narrowly tuned color processes which are inconsistent with conventional color-opponent theory.