Home | Our News | Light Quality Metrics Explained

Light Quality Metrics Explained

Even though vision is so inherently part of our everyday lives, it is extremely difficult to describe, quantify and scientifically measure. The human eye responds differently depending on brightness and colour, not all eyes are the same and even if they were, individual colour memories of familiar objects and each person’s individual tastes and preferences mean that there is very little common data to work with. This doesn’t mean however that attempts haven’t been made to objectively quantify light. If the performance of a light source can be accurately described, then it can be accurately reproduced, which as consumers is what we require. The complete output of a light source or SPD (spectral power distribution) can easily be measured with a photometer and the typical white LED’s output looks like this:

graph1 new

The SPD above has peaks in power at the blue and yellow wavelengths. White LEDs actually produce blue light which is changed to white light when it passes through the specialised yellow phosphor coating of the LED – this is why white LEDs are yellow when off. Although all of the light’s characteristics are on this graph, it is obviously not possible by looking at it to visualise how it will perform in the real world when illuminating objects of different colours. The most common metrics used in the industry to describe a light source today are:

CCT                    Correlated Colour Temperature (K) – The colour of light that a perfect radiator would emit if it was heated to this                                temperature. The sun, which is regarded as a black body (perfect) radiator as seen in the sky is at about 5600K.

CRI                      Colour Rendering Index – Commonly described as the ‘naturalness’ of a light source.


It is widely assumed that these two metrics alone are able to wholly describe the characteristics of a light source e.g. an 840 58W fluorescent would emit light at 4000K and a CRI of 80. CCT and CRI both have shortfalls however. Specific CCTs cover colours which fall on either side of the black body locus and so two light sources both with a CCT of 3000K can have noticeably different hues. The constant CCT lines can be seen running perpendicular to the black body locus in the CIE chart below. The 3000K line can be seen ranging from yellowish white to purplish white.

graph2 new

CRI is calculated by comparing the illumination of 8 colours by the test sample to that of the standard illuminant, with a perfect match equalling a CRI of 100. The 8 standard colours are shown below.

graph3 new

These colours are all regarded as pastel shades, with the light source’s rendering of more saturated colours not being regarded. Further problems include:

  • CRI was developed over 40 years ago for fluorescent technology and has remained largely unchanged since, even though lighting technology has developed hugely
  • Because the rendering of the eight samples is averaged, the contribution of any single large deviation is diminished
  • Two luminaires with completely different SPDs and completely different colour rendering capabilities can have the same CRI.
  • Manufacturers can ‘tune’ their LEDs to have an artificially high CRI, with poor individual colour performances (often red rendering)

Very important differences in light source performance which are not included in common specifications are:

graph4 new

Attempts are constantly being made to add metrics to improve the way in which light sources are specified. Some of these present in the industry today include:

MacAdam Ellipses  – The results of a statistical test based method of determining the average person’s ability to notice the differences between similar colours. The specification of finer MacAdam steps means that fewer people will notice a difference in colour between different test samples. An example would be a 3 Step MacAdam specification.

graph5 new

Colour Quality Scale (CQS)  – Similar to CRI, but with 15 colour samples instead of 8. This results in a much wider measure of rendering performance.

graph6 new

Gamut Area Index (GAI)  – Area encompassed by the chromaticity coordinates of the 8 colour samples used for CRI. This evaluates the saturation or vividness of the rendered colours. GAIs of 90 – 100 are considered ideal.

graph7 new

White Body Locus  – An experimentally determined line in the colour space showing which hues are considered the whitest at each CCT by most people.

graph8 new

Even more detailed metrics are being proposed which will specify other areas of performance:

Feeling of Colour Index (FCI)                                    – High colour saturation results in high image clarity

Harmony Rendering Index (HRI)                              – Quantifies the distortion of colour harmony caused by different light sources

Categorical Colour Rendering Index (CCRI)           – Errors in colour rendering will shift the categorical colour names of some object colours

Memory CRI (MCRI)                                                   – Memory colours of familiar objects are used as reference to evaluate colour rendering


As can be seen from this article, the complexities involved in quantifying and ultimately specifying a light sources performance are steadily increasing as the industry’s demands become more detailed and specific. Keeping up to date with all of the metrics involved is vital to ensuring the best solutions to lighting challenges are provided.

In our next article, we will describe which metrics are suitable for each application so be sure to follow LEDwise on LinkedIn to stay updated.


Written by Daniel Prozesky, LEDwise Lighting

Product Search

Which product are you looking for?