The Brighter Choice
THE SHORT ANSWER
A colour rendering index (CRI) is one measure of the quality of white light
In brief, the CRI of a light source (such as an LED lamp) describes how well that light can reveal the colours of objects relative to how they would be revealed under natural daylight. CRI is expressed on a scale from 0 – 100, where 100 indicates that colour will be revealed as well as it would under natural daylight, with lower values indicating that colour revelation will be less complete.
Have you ever bought a shirt and a tie, convinced they matched well, but found when you left the shop that they didn’t go together at all? Or have you ever put on a pair of socks, only to find when you left the house that they weren’t a pair after all? This is what can happen when colours are selected under light with a low CRI. Your eyes are not defective – but the information they are receiving is incomplete because the light source in the shop or over your sock drawer is emitting light with a low CRI.
EN12464 recommends minimum CRI values for a wide range of applications. Light sources with a CRI>90 are recommended in clothing retail, artistic and medical settings where accurate colour perception is required. CRI>80 is recommended for offices and schools, and CRI>70 is acceptable in many outdoor applications. Light sources with lower CRI values than 70 are rarely used.
CRI is only one measure of the quality of white light. Another widely used, complimentary, measure is correlated colour temperature (CCT). For details, please see What is correlated colour temperature (cct)?
THE FULL ANSWER
First, let’s consider how colour vision works
Our eyes respond to different wave-lengths of light. The sun emits the full visible spectrum – every colour of the rainbow, and all the shades in between – but when sunlight falls on an object, some wavelengths are absorbed and others are reflected. It is the reflected wavelengths we see, and these define the colour of the object in question.
In terms of the ability of a light to reveal true colours, daylight is considered to be perfect. This is because it contains at least some of every single wavelength. Put in technical terms, its spectral power distribution (SPD) is continuous – there are no gaps.
An incandescent (tungsten filament) light is similar in that its SPD is continuous, but it is biased towards the red end of the spectrum. However, the SPD of all other light sources contain some gaps, so our ability to perceive colour under, say, a fluorescent lamp or some LEDs will be limited. Their CRI is less than 100.
So, if the light source under which we check the colours of a shirt or our socks is missing a few wavelengths, our clothing will never appear the same under that light source as it would under natural daylight
Now, let’s look at how CRI is measured
There is more than one measure of CRI, in just the same way that there is more than one measure of speed (mph or km/h). The CRI measure that is most widely used today (2022) is Ra defined by the Commission Internationale de l’Eclairage (CIE) in 1995.
Other measures exist, notably R96a, and IES TM-30, but these are not yet in widespread use.
To measure the Ra of a light source you need:
• 14 standard colour swatches, defined in CIE1995. The first eight of these are low-saturated (pastel) shades. Four of them are more highly saturated (strong) colours – red, yellow, green and blue and two are colours occurring commonly in nature
• A standard light source with a continuous SPD at a CCT of 5000K
• The light source to be tested
• The CIE 1960 colour space diagram
• …and some specialist equipment
Calculating the Ra involves some intricate maths, but the basic method is quite simple:
1. Each of the first 8 colour swatches are illuminated under the standard (ideal or perfect) light source and their colour is plotted on the CIE1960 colour space map. The coordinates are noted. Of course, the colour swatches are standard and the light source is standard – so the coordinates are already known. Therefore, this part of the test process does not have to be repeated every time a light source is evaluated.
2. Each of the first 8 colour swatches are illuminated under the light source being tested and their perceived colour is plotted on the CIE colour space map. The coordinates are noted. You now have a pair of colour coordinates for each colour swatch.
3. Calculate the distance between the coordinates in each pair, using the formula Ri = 100-4.6∆Ei In brief, Ri is the rendering index that is specific to an individual swatch. The light source under test could give a high Ri on some swatches (because the light source includes the appropriate wavelengths), but a lower Ri on others.
4. Sum all 8 Ri values together and divide by 8. The result is Ra, the rendering average.The coordinates of the other six samples can also be plotted and the Ri calculated for each one. This gives some additional information about the qualityof light being produced by the lamp under test, but this is not included in the calculation of Ra.
The drawbacks of Ra as a CRI measure
While Ra is the most widely used and understood measure of the colour rendering ability of a light source it has some significant shortcomings.
The chief drawback of Ra is that it includes only the very unsaturated hues and excludes the strong red, yellow, green and blue. It is common that LEDs have a weak performance in the red part of the spectrum, yet they could perform well in an Ra test because the red swatch is excluded.
For this reason NVC Lighting (and some other lighting companies) do not merely specify their LEDs with a CCT and a minimum CRI value. Instead, we specify Ri values for some of the other colours in the CIE 1995 colour swatch deck to ensure that our lighting doesn’t just look good on paper, but also looks good in real life.
