Photo: Shutterstock

LED light bulbs need describing in new ways

Thursday 27 Jul 17


Carsten Dam-Hansen
Senior Researcher
DTU Electro
+45 46 77 45 13


Anders Thorseth
Project Manager
DTU Electro
+45 46 77 45 47

Ra expresses colour rendering

The Ra index expresses how well the light is capable of rendering colours. Daylight has a Ra value of 100, as daylight renders colours perfectly. The lower the Ra value, the poorer the light is at rendering colours. An incandescent bulb and a halogen bulb both come close to daylight with Ra values of 99. However, even with a Ra value of 80, the human eye will find it difficult to distinguish between different colours.

Source: Light! by Tor Nørretranders and Olafur Eliasson
The Ra index is a set of values that describe light quality. However, the index stems from the era of the incandescent bulb and the fluorescent tube, and it makes it difficult to fully exploit the potential of LED technology. Entirely new ways of describing light are needed, according to two researchers from DTU Fotonik.

The quality of light is, among other things, perceived as its ability to render colours accurately. For decades, colour rendering been expressed by calculating the light bulb’s so-called Ra value. This is based on incandescent light, which renders colours perfectly, and until now there has been general agreement that the incandescent light bulb (and daylight) provide the ideal lighting. The incandescent bulb’s Ra value is therefore close to the highest possible—100. A fluorescent tube on the other hand typically has poorer colour rendering properties, and its Ra values are therefore lower.

The Ra index is based on light’s ability to render eight selected pastel colours. However, the emergence of LED technology has led to a global discussion about the continued justification of using the Ra index. With LED, it is technologically possible to design white light so that it highlights red colours more, for example. Or blue and green. However, if the colour rendering of the eight selected pastel colours is not good, then such an LED lamp will have a low score on the Ra index.

At DTU Fotonik, Senior Researcher Carsten Dam-Hansen and Project Manager Anders Thorseth are busy measuring light, and enjoy extensive cooperation with LED manufacturers who need help performing professional light measurements in the laboratories.

“With LED technology, it has become possible to ‘cheat’ the Ra index. In other words, you can have two different LEDs which render the colours very differently, but which have the same Ra score. Conversely, you can have LED light sources with different Ra values, but whose colour rendering is almost the same,” explain the researchers.

DTU Fotonik

Experimental setup with fruit (and cake) and a colour checker chart shows how differently two light sources render the colours, even though the light from each has the same value on the Ra index. Both light sources have a Ra value of 95. In the box on the left, the light source is a cold white light with a colour temperature of 5,747 kelvin, while the light source in the right-hand box is a warm white light with a colour temperature of approx. 2,631 kelvin.

A single number is not enough

The industry and researchers are measuring, calculating and discussing these issues back and forth—and some have even developed proposals for new measuring methods to replace the Ra index. But according to the two light researchers, the basic premise that a number can express something as complex as the quality of light is mistaken:

“The experience of the quality of light is extremely subjective. At the same time, how you experience the light also depends on what you need to use the light for. Do you need to sleep, wake up, perform a surgical procedure—or eat dinner? We need to find a new way of describing light, as LED technology has opened up the door to using light in completely new ways. The Ra index excludes these possibilities if, for example, the light requires a minimum Ra value of 95. We’re wrong in thinking that we’re helping consumers by imposing strict requirements on industry to only supply LEDs with a very high minimum score on the Ra index,” says Carsten Dam-Hansen.

Anders Thorseth agrees:

“Numbers aren’t the answer. It’s simply too inadequate. My vision is that we start talking about light in a new way. We could start by using words with a reference that consumers understand. Here in the Nordic countries, for example, we use expressions such as the blue hour, dusk, or autumn.”

The researchers refuse to set themselves up as judges as to what is good or bad light. And nor should the decision be left to the lighting industry, they say:

“Three American research studies of people’s preferences for light have independently shown that people actually prefer white light which highlights the red hues better than incandescent bulbs. So for years the industry has been trying to imitate the incandescent bulb, while there is strong evidence that people actually prefer something completely different,” explains Carsten Dam-Hansen, who is hoping to conduct a similar study in Denmark to determine whether Danes have the same preferences as their American counterparts. The two researchers believe that neither industry nor researchers should decide what good light is—it should be up to the consumer. 

“We must create light for people. We can calculate and measure everything, but unless the human factor is included in the equation, we might as well forget it,” says Anders Thorseth.


The two diagrams show the spectral distribution of the light sources in the two demo boxes. The top diagram shows the cold white light, where the blue hues are dominant, while the bottom diagram shows the spectral distribution of the warm white light. Here, it is clear that the red hues are dominant. Source: DTU Fotonik/Carsten Dam-Hansen.

Experts test light

The DOLL Quality Lab is a high-tech lab at DTU which is fitted out with advanced equipment that enables researchers to test the quality of different light sources. The laboratory carries out precise and impartial technical measurements from one day to the next, and cooperates extensively with Danish lighting manufacturers. At the DOLL Quality Lab, it is possible to measure:

  • How much light is being emitted;
  • In which direction the light is being emitted;
  • The energy consumption and efficiency of a light source;
  • How well a light source renders colours (CIE Ra index or IES RF and RG-index);
  • The colour temperature of the light;
  • Flicker (variations in the light). 


Read more about the laboratory

News and filters

Get updated on news that match your filter.