Select your location

How MacAdam ellipses affect LED color consistency

The perceived quality of a lighted space is greatly affected by the color consistency within the space; therefore, the consistency of individual luminaires’ output is crucial.    

That’s where MacAdam ellipses play a role. Today, experiments conducted by physicist and color scientist David MacAdam in the first half of the 20th century help us understand the perception of slight differences between colors. 

But how do MacAdam ellipses factor into LED lighting? We spoke with Reed Bradford, director of optical design and services at Eaton’s Lighting Division, about the correlation. 

What is a MacAdam step? 

RB: The human visual system is unique for each of us, just as our sense of taste, hearing and pain tolerance. Around 1940, David MacAdam’s vision research determined that observers could match colors within a consistent statistical range. Using MacAdam’s method to show a plot of this range on the CIE XY color space results in an elliptical area of color that observers would identify as being the same. 

As the graph shows, human tolerance depends on the color the person is attempting to match. For instance, we have a more difficult time distinguishing shifts in green, so the ellipse is larger than, say, shifts in blue. For the average observer, any color within the elliptical boundary is indistinguishable from the center point; the boundary of the ellipse is the point at which we would begin to detect a noticeable difference in chromaticity.

Why is color consistency important in LED lighting fixtures?

RB: The human visual system is amazing in many respects. Color perception is one of our highly tuned senses, and it allows us to detect even slight changes in color. For example, if an observer views warm white LEDs at 3,000K, it will only take a small change of +/- 30K for the observer to visually detect a difference between two sources.

The importance of color consistency depends on the nature of the environment and surface to be lighted. For instance, illuminating a light-colored wall with wall wash luminaires should make it easy to detect slight differences in color projected from each luminaire. Likewise, if the luminaire itself is an intentional visual element in the space, such as an array of suspended luminaires or row of wall sconces, the lens and diffusers should appear as the same color. 

Conversely, color detection for external lighting applications is diminished due to the visual receptor shift of our eyes at lower light levels (i.e., the Purkinje effect). Color identification is not extremely critical in most lighted outdoor environments. Parking lot and roadway luminaires are not typically a visual element of the lighting design; plus, the wide spacing and high brightness of these luminaires makes tight color consistency less critical. 

Can do CCT and ANSI standards play a role in lighting product selection?

RB: The color of light can be defined by its spectral power distribution, coordinates on a color space graph or correlated color temperature, CCT. CCT is a single number that defines the proximity of the light source’s chromaticity along the blackbody curve (locus). As the physical temperature of a filament increases, the color shifts from a deep red through orange and yellow to white, and finally bluish-white. The color of light emitted by any light source can therefore be translated or correlated to a point along this curve with a resulting CCT. 

In most lighting applications, LEDs produce light by converting the blue light generated by the diode and passing it through a phosphor layer that converts the blue to white light. White light can vary from a very cool to very warm appearance, much like is typical with fluorescent lamps. 

To ensure consistency in appearance, the American National Standards Institute (ANSI) documented* the various incremental steps in color along that spectrum. As shown in the graph, the boxes along the black body curve define the allowable tolerance for the given CCTs defined as white light. At the blue end of the curve, the highest CCT defined is 6,500K, and 2,200K is at the warm end of the spectrum. Each quadrilateral or ANSI bin is defined with a center point (on or very near the blackbody curve), a tolerance in color temperature (along the curve), and a deviation above and below the curve.

The largest quadrilateral (bin) defined by ANSI would allow a MacAdam ellipse of seven steps to be inscribed within (see below). In order to meet the criteria for any given CCT, an LED’s chromaticity coordinates must fall within these defined bins. 

What is the difference in nanometers from one MacAdam step to the next?

RB: There is not really a direct correlation between nanometers and MacAdam steps. White light generated by LEDs is composed of all wavelengths (measured in nanometers) within the visible spectrum. CCT is affected by the wavelength composition; more blue content leads to cooler color (high CCT), and more red content leads to warmer color (lower CCT). As seen in the preceding graph, the ellipses at any given CCT are centered around a point near the center of quadrilateral. A one-step ellipse will be small and therefore only allow a small color deviation from the center point. However, as the size of the ellipse grows, the amount of color variation also grows. The size of the ellipse and defined distance from the center point are defined by ANSI C78-377. Based on the nominal color temperature, allowable upper and lower boundaries are given in degrees Kelvin (e.g., 3,000K is centered at 3,045K with a tolerance of +/- 175K; 4,000K is centered at 3,985K with a tolerance of +/- 275K.) 

How do large variations affect the average observer’s experience?

RB: Large color shifts are visually distracting and create an impression that a space is poorly designed or maintained. For example, the appearance of items such as fabrics in a store or cars at a dealership can vary depending on the consistency of the light. 

What is the selection process for an LED fixture in terms of number of MacAdam steps? What are the best practices?  

RB: Tighter control of the color of LEDs is desirable, but this must be balanced with the associated costs of doing so. The production of LEDs has improved over time, and today manufacturers are able to control the resulting color output. But there are still color variations based on certain manufacturing tolerances. Production of LEDs is sorted (binned) by proximity to the center point, resulting in varying quantities in each bin. If an LED manufacturer has good control over the various inputs, the majority of production will fall within the tighter ellipses, but there will still be LEDs that fall into the larger ellipses. Pricing of these LEDs follows the law of supply and demand, so the smaller, more desirable ellipses are more costly.  

With the evolutions of LEDs and rapidly advancing technology, are MacAdam steps accurate from year to year in LED lighting?

RB: Yes. ANSI defines the standard, and it’s been consistently followed as technology has changed from fluorescent, to HID, to LED. The most critical advancements are in in the LED manufacturer’s ability to reduce the variations in production and yield more LEDs with tighter color control. 

As we move into the realm of solid state lighting and introduce further variables such as LED binning, knowledge of how closely the eye can discriminate between colors, and how we can measure and specify that discrimination, will become increasingly important to everyone.

*ANSI C78-377-2015, the American National Standards Institute literature on this topic, is available for purchase.