Cleantech Market Intelligence
Light Color is Complicated
I recently viewed a webinar titled, “Light and Color Measurement of Today’s LED Technology.” The presenters were from Konica Minolta Sensing Americas. While I wasn’t part of their intended target audience of measurement equipment purchasers, I wanted to take the opportunity to learn more about the color-related complexities of the light emitted by LEDs. To start with, “complexities” is probably not a strong enough word. I’m not going to get into anywhere near the technical detail that these gentlemen presented. My head still hurts from just the parts I tried to understand, and don’t even ask me why “y10 is not the same as y or V(λ)” as it said on one of their charts. With apologies to people who really do understand this stuff in detail, here goes.
In the good old days, the color properties of a light source were described by color temperature and Color Rendering Index (CRI). Color temperature, as it relates to the appearance of the light, is based on the principle that when an object is heated to a high enough temperature, it will emit light. The measurement, then, is based on the colored light a “black body” will emit as it is heated. As the temperature increases, the emitted colors shift from red to orange to yellow to “white hot.” The color temperature of a light source is the temperature at which the heated black body is emitting the same color as the light source.
The purpose of the CRI is to describe how colors will appear under a light source. The Index is based on how a set of eight specific colors appear when illuminated by the light as compared to a reference source at the same color temperature. If the specific colors appear the same, the light source is assigned a CRI of 100. The classic incandescent bulb has a CRI very close to 100 because it renders those eight reference colors very well. This is one of the reasons it is sometimes difficult to replace incandescent bulbs with more energy efficient light sources that have lower CRIs. People are used to the way things, and they themselves, look under incandescent lights. That doesn’t mean that incandescent lighting is perfect, though. Incandescent bulbs tend to be weak in the amount of blue light they emit. For example, do you have a problem telling the difference between dark blue and black socks in a space lit with just a bulb, like in a closet?
And that’s one of the difficulties with light color. Even light sources with high CRIs don’t necessarily give off “perfect” light. In fact, the traditional CRI is pretty much an irrelevant measure for LEDs because the typical spectrum that makes up white LED light doesn’t work well with the eight reference colors. The National Institute of Standard and Technology is developing a new Color Quality Scale (CQS). The method for calculating the CQS is based on modifications to the method used for the CRI, starting with a more representative set of 15 reference colors. The CQS is being proposed to the International Commission on Illumination (CIE) Technical Committee that is working on “Colour Rendering of White Light Sources” (TC 1-69).
The illustration below shows that light from different sources, even light that looks “white,” can be comprised of very different combinations of wavelengths. (The “Tungsten Lamp” is the representation for an incandescent bulb.)
Here is a more specific illustration of the difference between daylight and white LED light.
Color temperature also doesn’t exactly apply to LED sources since their light cannot be matched by a heated black body. There is a standard approach for computing color temperature. Suffice it to say that it involves equations that represent a two-dimensional plot of the component colors emitted by the source. Unfortunately, the emittance of LEDs does not follow the standard curves this approach assumes. So, while people are trying to use the familiar terms related to color temperature, LEDs with the same computed color temperatures can still have perceptibly different colors.
Most light sources used to date radiate their light in many directions. Think of the standard bulb, a CFL, or a fluorescent tube. The light is often directed with reflectors and lenses, but that’s outside of the source itself. LEDs are point source emitters that give off light in a much more directed manner. However, measuring the brightness of the emitted light gets complicated and the fact that the component colors of the white light might not all exactly “line up” adds another dimension of complexity.
First of all, there are differences between the mechanical center line of the LED (the “shaft construction” in the illustration below), the Optical Axis (sort of the “weighted average” center line), and the Axis of Maximum Brightness (the direction of the brightest emittance). And remember, the diagram below is just a two dimensional representation of a three-dimensional situation.
So, to measure the brightness of an LED requires carefully defined, standardized methods. (Here’s where the Konica Minolta guys described the different types of equipment they provide.) The diagram below with representations of different LED emission patterns shows why measuring the color characteristics is even more difficult.
My head is hurting again, so I’d better stop here for now. I didn’t even get to all of the factors that influence the color and color rendering characteristics of LEDs. There is also an LED’s color stability, the influence of junction temperature, and the difference between color temperature and correlated color temperature (CCT – which doesn’t strictly apply to LEDs, either).
The bottom line is that light color is a complicated topic and it gets even more complicated when LEDs are involved. If you’re purchasing LED lighting and color is important, make sure you get the results of tests that really will describe product performance. What test methods were used? Who did the testing? Above all, probably, try before you buy.