Thanks to LED lighting specified, the signature yellow and purple of the Lakers’ uniforms pop when the team plays at the Staples Center in Los Angeles.
Original photo: Keith Allison via Flickr/Creative Commons
Thanks to LED lighting specified, the signature yellow and purple of the Lakers’ uniforms pop when the team plays at the Staples Center in Los Angeles.

Daylight during the sun’s zenith—in which each wavelength in the visible color spectrum comes together—is the paragon of white light. As recently as five years ago, finding LED products that delivered uniform white light with consistent color rendition was difficult. “Just getting a decent white light was an issue,” says Glenn Heinmiller, a principal at Lam Partners, in Cambridge, Mass.

Today, manufacturers have a number of strategies for producing solid-state lighting (SSL) with color rendering index (CRI) values that surpass 80, and are harnessing the latest science to target specific color points for custom uses. Institutions, including the Lighting Research Center (LRC) in Troy, N.Y.; the Illuminating Engineering Society (IES); and the University of British Columbia have proposed new color rendering metrics more attuned to the unique, digital characteristics of LEDs.

The Standby

CRI, developed by the Commission Internationale de l’Eclairage (CIE) in the 1950s and approved in 1964, remains the standard for measuring color performance despite a consensus among many lighting designers and manufacturers of its inadequacy for solid-state lighting. “Color science has progressed a lot since then,” says Michael Royer, a lighting engineer with the Advanced Lighting Team at the Pacific Northwest National Laboratory (PNNL), in Richland, Wash. The metric uses eight pastel color chips from the Munsell color set (named after painter and art professor Albert Munsell), which Royer says can create an undesirable sensitivity in the metric and allow manufacturers to manipulate the system by targeting those colors. (Six additional color chips are used to provide supplementary information about the light source.)

In 2012, the National Institute of Standards and Technology developed the Color Quality Scale (CQS), which Royer says made only modest improvements to CRI. CQS uses 15 color chips and CIE’s 1976 color space—a chromaticity diagram that maps the color spectrum—that has a slightly modified lightness scale from the 1960 color space.

In 2014, in response to the increasing market share of SSL, the LRC established Class A Color. To be considered Class A, a luminaire must have a CRI higher than 80; a consistent chromaticity, or hue; and a gamut area index (GAI) between 80 and 100. GAI compares a source’s gamut area—which indicates how saturated an object’s colors appear—to that of a reference source. Unlike CRI, GAI can exceed 100.

The Newcomer: TM-30

This summer, the IES is expected to approve TM-30, IES Method for Evaluating Light Source Color Rendition as a new way to evaluate color rendition in LEDs. Developed by a task group chaired by Royer, the method was also submitted to the CIE for consideration as the new international standard.
The eight test color samples used to calculate CRI (left) and the 99 color evaluation samples used to determine TM-30 (right), which is pending review by the IES. The colors shown here are approximations.
The eight test color samples used to calculate CRI (left) and the 99 color evaluation samples used to determine TM-30 (right), which is pending review by the IES. The colors shown here are approximations.

TM-30 uses 99 color evaluation samples, drawing from leaves, flowers, skin tones, paints, and some of the original Munsell chips. It spans the entire color space, including saturated and desaturated colors, and introduces new metrics for fidelity (Rf) and gamut (Rg). Similar to CRI, an Rf of 100 indicates a perfect match with the reference. Rg is calculated by plotting a light source’s chromaticity values in a color space and comparing the area to that of a reference source. “If Rf is 100 … Rg must also be 100,” wrote New York–based Studio T+L principal and IES Color Committee member Jason Livingston on his website, Designing Light. “As the Rf value falls, the potential range of Rg above [or] below 100 (indicating an increase or a decrease in saturation) grows.”

Much of the thinking behind CQS and Class A informed TM-30, Royer says. His task group, however, took issue with GAI because it relies on the same set of color samples as CRI as well as on the equal energy spectrum, a theoretical source that emits the same amount of energy at every wavelength. “Thus,” he says, “GAI and CRI do not work as well together as a system since a given source will be compared against different references for the two measures.”

Though Lam Partners’ Heinmiller is anxious for a replacement for CRI, he acknowledges that most designers don’t have time to research what these new scores mean. “Once [the new metric is] established, the manufacturers will pick up on it and test to it, and then we can all start using that.” Until then, he says, most designers will continue to look to CRI for guidance.

The Preference Factor

One thing Class A and CQS attempt to measure, which TM-30 does not, is human preference. Research has shown that people sometimes prefer light that renders colors more vividly—that is, with a higher gamut area than the reference source. And the LRC has found that when it comes to white light, human preference does not always follow the blackbody curve.

In its 2011 paper “Perceptions of White Light Sources of Different Color Temperatures,” the LRC found that humans perceived untinted white illumination with chromaticities that map to an “S”-curve rather than the smooth arc of the blackbody curve. Perceptions of white light at and above 4000K fall just above the blackbody curve and have a greener tint, while perceptions of white light below 4000K fall just below it and are slightly more pink. Historically, these subtle tints were viewed as imperfections but, as the research suggests, they may be preferable at certain color temperatures or for certain situations. The term “whitebody curve” has informally emerged to describe these differences, says Ken Bruns, a controls product manager at Lumenpulse.

“Going beyond the blackbody is exciting,” says Maria Topete, director of applications engineering at Bridgelux, because it raises questions about which characteristics make a particular type of white light desirable. “The vernacular to talk about [people’s preferences] and the ability to quantify those characteristics are evolving.”

Perceptions of white illumination deviate slightly from the blackbody curve depending on the light source’s correlated color temperature.
Mark S. Rea / Lighting Research Center Perceptions of white illumination deviate slightly from the blackbody curve depending on the light source’s correlated color temperature.

Optical Optimizers

As lighting metrics evolve in conjunction with the increasing prevalence of SSL, the quality of white LED light is approaching the standard embodied by daylight with the advent of several new and updated technologies.

Typically, manufacturers designing a high-CRI white LED luminaire must find a balance between diffusion and reflectance. Diffusers typically lessen a luminaire’s efficacy, whereas high-reflectance surfaces increase output but can create hotspots and glare. “If you have a bunch of LED dots bouncing off a specular surface, it’s going to look like an arcade or a ’70s disco,” says Eric Teather, president and founder of WhiteOptics.

A spin-off from DuPont and developed in partnership with the U.S. Department of Energy and the University of Delaware, WhiteOptics, in New Castle, Del., makes patented composite materials that allow for up to 98 percent reflectance while offering a ratio of particles and microvoids that scatter light so that the optic artifact—the beam itself—is diffused. This allows an original equipment manufacturer to maintain a high lumen output and color quality while diffusing the reviled LED dots.
WhiteOptics
WhiteOptics WhiteOptics’ aluminum product (upper portion of photo above) reflects up to 98 percent of light and diffuses hotspots, as compared to a standard finish.

White rendition, or how well an LED renders a white object, is often overlooked as a measure of color quality. Optical brighteners, chemical compounds that make objects appear brighter and whiter, are embedded in everyday products, such as apparel and printer paper. The brighteners work by absorbing and then remitting ultraviolet and violet light as longer-wavelength visible light, which adds a blue tint.

Also known as fluorescent whitening agents (FWAs), optical brighteners have been in use for decades. Last year, they made headlines following the study “Whiteness Perception Under LED Illumination” (Leukos, April 2014) by Kevin Houser, a professor of architectural engineering at Penn State University (and a member of architectural lighting’s editorial advisory board). Houser tested the fluorescence of optical brighteners under a halogen lamp, a violet-pumped LED (a white source that uses a violet-emitting LED), and a blue-pumped LED. While the first two performed in a similar manner to incandescent light, the blue LED failed to activate the FWAs, and the products appeared dingy and yellow.

“The results indicate that engineering of an LED source’s spectrum is necessary for an accurate rendering of whiteness,” Houser wrote. This research has spurred manufacturers, including Soraa and Lumileds, to develop violet-pumped LED arrays that will render white objects such as paints, appliances, and fabrics as their makers intended.

Color and the Real World

Color performance can make a significant difference in healthcare environments, where the tone of a patient’s skin can inform a diagnosis, and in museums, where works of art can be damaged by low-quality light. But where it really garners interest is in retail applications. The quality of light can affect a shopper’s experience and, moreover, their decision to make a purchase.

When Aaron Merrill, the senior director of channel marketing at Bridgelux, was recently leaving a grocery store near his company’s headquarters in San Francisco, he observed: “In the checkout aisle, all the candy bars were illuminated with LEDs. All the colors were popping, and I couldn’t keep my kids [away].”

Similarly, the LED lighting used at the Staples Center in Los Angeles was developed in direct response to the colors of the Los Angeles Lakers’ uniforms. “The [owners] wanted specific color points that [made] the yellows and the purples pop, and something that looked good on high-def television,” Merrill says.

But even custom LED lighting still remains subject to the whims of human preference. Merrill points to the example of a luminaire in Europe developed specifically to display bread. Responses to the light varied by region, he says, a reminder that the measure of light quality goes beyond what can be measured in numbers and indexes. “There’s a whole new world of behavioral science still to be discovered.”


Resources
A list of references that discuss color metrics and white lighting.

“Class A Color Designation for Light Sources Used in General Illumination,” by Jean Paul Freyssinier and Mark S. Rea, Journal of Light and Visual Environment, 2013. Available at: bit.ly/1gK223S.

“Whiteness Perception Under LED Illumination,” by Kevin W. Houser, Leukos, 2014. Available at: bit.ly/1fYqSNr.

“Perceptions of White Light Sources of Different Color Temperatures,” by the Alliance for Solid-State Illumination Systems and Technologies and the Lighting Research Center, 2011. Available at: bit.ly/1LngV9b.

“TM-30 and Color Gamut,” by Jason Livingston, Designing Light, June 2015. Available at: http://bit.ly/1VJziWm.

“How New Methods for Evaluating Color Rendering Will Affect You,” Michael Royer, Lightfair, May 2015. Available at: 1.usa.gov/1EbDc2f.

More about Lumenpulse
Find products, contact information and articles about Lumenpulse