Is it time to stop using CRI?
The Colour Rendering Index or CRI is one of the key metrics lighting specifiers review during the specification process. It is the relative scale on which we compare how well light sources reflect colours back to the human eye. CRI has been a standard part of reputable specs for over 40 years. Yet, I think that if you’ve ever really evaluated light sources you would know that CRI is often an inaccurate way of expressing the quality of a light created.
In my past life, as a lighting designer for Abercrombie and Fitch we evaluated halogen and LED light sources as options for display lighting. What we found was that CRI had very little to do with how our eyes actually saw the products under light. Halogen sources with a rated CRI of 100 muddied navy blue colours, while making reds pop with vibrancy. LED sources had the exact opposite effect, they muted bright reds lessening their vibrancy, yet navy blues were made much richer.
There is, of course, another problem: the fact that CRI corrects for Colour temperature (thereby implying that colour temperature of a source has no bearing on CRI). Yet colour temperature is essential to how a light source affects our perception of colour. This all begs the following question – in a world where types of light sources are expanding exponentially and need to be understood more easily, don’t we need to rethink this means of evaluation?
Here in the US, our National Institute of Standards and Technology has developed an alternate scale called the Colour Quality Scale or (CQS). I think CQS is superior to CRI in at least four key ways:
- CQS updates the tested colour space to a more modern and recommended colour scale.
- CQS does not correct for CCT but rather enters it into the model.
- CQS indexes the results of 15 test colors without averaging them (CRI averages the results of 8 test colours).
- CRI punishes light sources for enhancing contrast and colour saturation, where in most applications this effect is actually desirable.
I recommend you read the IALD’s paper on CQS for a thorough analysis.
So if CQS is so much better than CRI why hasn’t it been fully adopted yet? Well that’s hard to say. First of all, I should say that most of the information I found published on CQS dates to 2010. These white papers are introductions to the scale with explanations of why it is superior to CRI. However, it appears not all parts of the industry agree. In this well-reported piece by Dan McGovern of Lumenistics in March 2012 the following explanation was offered:
“We’ve been working for over five years now and have been unable to reach an agreement, so far,” Davis said. “The CIE requires unanimous agreement to issue a recommendation, and this committee has [approximately] 40 members. As you can imagine, it’s been tough. I’m not sure if the committee will be successful.”
It certainly hasn’t been in the past. The CIE has attempted to revise the CRI several times throughout the 1980s and 1990s “and every committee ended in failure,” Davis said. “If that happens now, which is very much a possibility, I wouldn’t be surprised if some other standards organisation were to take up the issue.”
So, for now, CRI remains the only internationally recognized colour rendering system. Whether the lighting industry, especially the solid-state category, will continue to use it for evaluating lighting applications remains to be seen.
The next logical question is if the arguments for CQS are so good, why is there still disagreement? Research scientist Jean Paul Freyssinier of the Lighting Research Center of Rensselaer Polytechnic Institute gave a talk at Strategies in Light last year. He said, “None of the metrics that we use are perfectly predictive of peoples’ perception or assessment of light sources.” On the topic of colour rendering he would go on to say, “no single metric can characterise colour rendering.”
All this leaves the lighting designer with few options. Use a barely trustworthy but ubiquitous metric or use nothing. As such we are forced to assess light sources using the old eye test. There’s no substitute for looking at a light source and asking it to perform the function you’ll need it to perform in the field.
The simple truth is that we are living and working through a time of incredible explosion of lighting technologies. There will be winners and losers over the coming decade and perhaps we’ll normalize how LED sources are produced to provide consistency across manufacturers. Until then every manufacturer is convinced they know how to make the best LEDs. It’s up to the lighting design community to assess these sources and let our specifications do the talking. In the absence of definitive evaluation metrics, it looks like there’s still just as much art to lighting design as there is science.
Do you think James is right about CRI? Join the debate in our Lighting Discussions LinkedIn group here