Experimental validation of colour rendition specification criteria based on ANSI/IES TM-30-18

Abstract

An experiment was conducted to examine colour rendition specification criteria. Twenty-five participants each evaluated 90 lighting scenes in a room filled with objects. The lighting scenes included nine chromaticity groups, each with 10 systematically-varied colour rendition conditions designed to meet or not meet previously proposed colour preference specification criteria using the ANSI/IES TM-30-18 measures: Fidelity Index (R_f), Gamut Index (R_g) and red Local Chroma Shift (R_cs,h1). The colour rendition conditions did not meet the criterion for none, one, two or all three of these measures. Participants, who chromatically adapted to each chromaticity group, rated the objects' colour appearance on eight-point scales for saturated–dull, normal–shifted and like-dislike (preference), as well as a binary for acceptable or unacceptable. The findings corroborate past work, but also indicate that colour preference criteria could be adjusted slightly to improve performance, with Tier A having R_f ≥ 78, R_g ≥ 95 and −1% ≤ R_cs,h1 ≤ 15%, Tier B having R_f ≥ 74, R_g ≥ 92 and −7% ≤ R_cs,h1 ≤ 19%, and Tier C having R_f ≥ 70, R_g ≥ 89 and −12% ≤ R_cs,h1 ≤ 23%. A companion regression analysis shows models based on R_f, R_g and R_cs,h1 were superior in predicting colour preference compared to those using other measures of colour rendition.