(Issue: December 2020)

Color Research

By Craig DiLouie, CLCP, LC


Despite its limitations, the color rendering index (CRI) endured as a lasting measure of color fidelity in light sources. A rule of thumb is an 80+ CRI is considered “good” for many commercial applications such as office buildings and schools. A 90+ CRI, meanwhile, is considered “excellent” for discerning applications such as groceries and retail. 

The biggest limit of CRI is it measures only fidelity—the average of how eight test colors appear for a light source compared to an ideal reference source—but not chroma or hue. This results in two light sources having the same CRI but possibly a different gamut (increase or decrease in chroma)—e.g., one being oversaturated in blue and the other in yellow, resulting in objects and the space looking different under each source. 

In 2015, the Illuminating Engineering Society (IES) announced TM-30-15, a technical memorandum that unveiled a new method for evaluating color in light sources. In 2018, a new and improved version was released. TM-30 measures both color fidelity and gamut. Fidelity is measured using a metric called Rf, analogous to CRI but based on 99 color samples. Gamut is measured using Rg. TM-30 also provides for color vector graphics that visualize average hue and chroma shifts. 

Why is TM-30 important? While CRI may have traditionally been seen as convenient and sufficient, research suggests fidelity alone is not enough to do the job, and that fidelity is actually less important than gamut in predicting human color preference. CRI also does not indicate saturation in red, which studies also suggest is a key element in human color preference. 

In one such study published by the Pacific Northwest National Laboratory in 2016, participants entered a room filled with common items—fruit, clothes, a mirror to view skin tones, and more—all illuminated at 3500K in a series of lighting scenes. Each scene received light with a different spectral power distribution, with variations in fidelity and with color gamut ranging from under to oversaturated. The participants then rated each scene. The researchers found that color gamut, not fidelity, was most strongly linked to preference. More specifically, saturation in red was found to be important. 

This resulted in a simple equation: Preference = Fidelity + Red Saturation. The researchers concluded a high level of preference correlated with fidelity being 75+ with a 2-16 percent oversaturated in the red spectrum and overall saturation. The WELL Building Standard adopted similar TM-30 measures (Rf 78+, Rg 100+, 1-15 percent red oversaturation), with an alternative being either >90 CRI or >80 CRI coupled with an R9 of >50. 

After the PNNL study, a similar study was conducted by Penn State University. In this study, the researchers similarly kept the light source at 3500K while varying fidelity and gamut, but showed the participants only one lighting scene per day. Despite the different methodology, preferred light sources fell within the same parameters as the PNNL study. PNNL subsequently conducted a follow-up study in which they varied chromaticity from 2700K to 4300K, and again achieved similar results for preference. 

Lighting upgrades provide the ability not only to provide comfortable vision at the highest level of energy efficiency, but also enhance overall occupant satisfaction with the space being illuminated. TM-30 provides a tool with which lighting management companies can evaluate the color performance of light sources with much greater precision. Further, research provides the basis for ensuring light sources are selected that produce sufficient red saturation. 

Craig DiLouie, CLCP, LC, principal of ZING Communications, Inc., is a consultant, analyst and reporter specializing in the lighting and electrical industries, and a regular contributor to LM&M. You may contact Craig at cdilouie@zinginc.com.