Vision Experiments on Chroma-Related Preference for Color Quality of Lighting

ABSTRACT

Two vision experiments were conducted to quantify the chroma-related preference in color quality of lighting in a real room-size cubicle simulating interior lighting environment with spectrally tunable light sources. The first experiment used 9 different lights with increased or decreased chroma in nine levels in red-green directions. The second experiment used 11 different lights with increased or decreased chroma in three hue directions, for red, green, and yellow, controlled as independently as possible. About 20 subjects participated in these two experiments evaluating the appearance of various fruits and vegetables as well as skin tone under these test lights by numerous pair comparisons. The results of the first experiment showed that chroma increase in ΔC*_ab ≈ 4 to 8 from the reference light was most preferred for all target objects, and desaturated lights were disliked in all cases. The second experiment showed that chroma change in red is significantly important for the preference, green was second, and yellow had the least effect on preference. The results of the second experiment suggested that preference increases with increasing red chroma in light, and after a peak in preference (between ΔC*_ab ≈ 5 to 10), preference starts decreasing, which is consistent with findings from the first experiment. Neither the viewing targets nor the correlated color temperature (CCT) conditions were found to have significant impact on the subjects’ preference.

KEYWORDS:

• Color rendition
• color preference
• chroma
• perception
• vision experiment

Acknowledgments

This research was conducted when Yuki Kawashima and Semin Oh were guest researchers at NIST. We thank Mira Fein, who stayed at NIST as a guest researcher under the NIST-NSF Summer Undergraduate Research Fellowship Program and assisted with the experiments, and Prof. Yongshin Kwak of Ulsan National Institute of Science and Technology, South Korea, for her contributions to Experiment 2 and other support for this research. We also thank our NIST colleague Dr. Julia L. Sharp in the Statistical Engineering Division for her advice and help in statistical analysis. The experiments using human subjects reported in this article were approved by the Institutional Review Board (study protocol PML-16-0001).

Disclosure statement

No potential conflict of interest was reported by the authors.