DOE OSTI2019
Pacific Northwest National Laboratory (PNNL) has released the results of a study – published in Lighting Research & Technology – that validates color-preference specification criteria based on ANSI/IES TM-30-18. The result provides additional data to support TM-30 Annex E (in press), which will include tiers of recommended specification criteria for different goals (color preference, color vividness, and color fidelity), among other information related to color rendition specification. Standardized methods for evaluating the color rendition of light sources have progressed rapidly in the past five years, beginning with the publication of IES TM-30-15, followed by the adoption of CIE 224:2017, and culminating with the issuance of ANSI/IES TM-30-18 (which revised TM-30-15, made it a national standard, and unified the calculation framework of CIE 224 and IES TM-30). Importantly, none of these documents provided guidance on how to specify color rendition using the included measures; that is, no value or combination of values was identified that leads to any desired quality of the light. Dozens of experiments over the past 60 years have examined subjective evaluations of color quality—such as preference, naturalness, or vividness—although the focus has traditionally been on developing a single-number color quality metric rather than specification criteria. Further, many of these pre-LED and pre-TM-30 studies were limited in their ability to vary the lighting characteristics, preventing them from generating a complete and accurate picture of what influences color rendering quality in architectural spaces. A renewed emphasis in this area of research over the past four years has yielded a growing consensus about how measures of color fidelity (Rf), gamut area (Rg), and red chroma (Rcs,h1) can be used in combination to effectively describe several important perceptual attributes. In the new study, which was the third in a series conducted at PNNL, 90 lighting conditions with systematically varied color rendition and chromaticity were evaluated by 25 participants, who provided ratings for normalness, saturation, preference, and acceptability. The lighting conditions were specifically designed to test the bounds of previously proposed color preference criteria, while simultaneously verifying the applicability of the criteria across different values of CCT and Duv. In line with the preceding work, there were substantial differences in ratings based on color rendition, but minimal differences based on chromaticity. The tested criteria performed well, although after the experiment, some small adjustments in lower limits provided more effective divisions between tiers of performance. The adjusted criteria are: • Level 1: Rf = 78, Rg = 95, -1% = Rcs,h1 = 15% • Level 2: Rf = 74, Rg = 92, -7% = Rcs,h1 = 19% • Level 3: Rf = 70, Rg = 89, -12% = Rcs,h1 = 23%. These criteria were concurrently checked against the results from other, independent experiments, and will be included in ANSI/IES TM-30 Annex E as the specifications for color preference. The new criteria will be a valuable tool for improving lighting quality and balancing the tradeoff between color rendition and energy efficiency.