Monday, November 6, 2017

Understanding Color Rendering





The automated lighting industry is full of different metrics used to compare diverse properties of various products.  Most people understand measurements such as Lumens, Lux, Foot-Candles, hertz, and dB, but color rendering has recently become the newest concern for many.  Color rendering scores can help designers determine how well colors (sets, costumes, skin tones) will look when illuminated by a particular light.

Many manufacturers will refer to the CRI of their product, while others are using CQS, TM-30, and/or TLCI as methodologies for explaining the color rendering properties of lighting products.  These scales are important as they describe the quality of the light and how colors will appear when illuminated by a specific source.  The following provides a brief overview of each color rendering methodology and how the measurements are achieved.

Measurement Basics
Every color rendering system has the same goal: to describe how well a light source (or fixture) will render specific colors compared to a known source (typically pure daylight or pure tungsten).  Usually there is a defined set of colors and a score is given based on how well each color looks when lit by the source compared to the reference.  If a color does not look good under a specific source (think a red shirt under a florescent) then a lower score is given for that particular color.  The measurements for all the chart's colors are then averaged together to create an overall score (usually on a 0-100 scale). Note that a perfect score is only possible with the original source (pure daylight or pure tungsten).


Because it can be very difficult to actually look at a color reference chart and determine the rendering ability of a source, light meters are regularly used to measure color rendering.  They will examine the light coming from a source and use complex mathematical formulas to calculate the color rendering score based on the source's color temperature and spectrum.



CRI
CRI or Color Rendering Index is currently the most commonly provided metric for color rendering, which is a shame as it is also the poorest.  Basically it consists of eight* colors that are given a 0-100 score and simply averaged together.  The eight colors primarily consist of skin and nature tones, but not any saturated hues.  Because these eight values are simply averaged together it is possible to get skewed results.

For instance, a source that scores high on seven of the values, but very low on one could still result in a high CRI value.  Furthermore, manufacturers can adjust the CRI score simply by filtering certain wavelengths to get to a better average score (although not necessarily better light output).

*The full CRI specification consists of 14 colors, but typically only the first eight are used for calculation.

CQS
The Color Quality Scale (CQS) uses 15 color references and instead of a basic average, it uses a root-mean-square calculation methodology.  This squares each value before averaging resulting in a much more fair result where even one bad value will greatly affect the total result.  The chart contains many saturated colors and is considered a much better scale than CRI.

TM-30
The Illuminating Engineering Society (IES) created the TM-30 metric to replace CRI.  This scale uses 99 color references based on real world color objects and again uses the root-mean-square calculation to determine the final score.  This is known as the TM-30 Fidelity index (Rf).  This standard also provides a Gamut index (Rg) which scores the saturation of the colors as over or under saturated.  An Rg score of 100 means a perfect match, with greater than 100 being over saturated and below 100 being under saturated.  Furthermore, TM-30 reports provide color vector graphics to visibly display which wavelengths are over or under saturated.

TLCI

The Television Lighting Consistency Index is specifically designed to represent  how colors will render on camera as opposed to the human eye like the other metrics.  Using 18 colors and the root-mean-square calculation method, TLCI results also provide data to help make appropriate corrections to the video image.


With so many different color rendering metrics to choose from, it can be confusing to determine how well a specific light will render colors.  It is important when comparing fixtures to understand how scores are calculated and what they are based upon.  CRI is the worst choice, while CQS and TM-30 are much better references.  TLCI should always be used for television.  Further detailed descriptions of each of these color rendering methods can be found on Mike Wood's website


Martin lighting now provides full metrics for all the above color rendering scales in fixture photometric reports starting with the MAC Encore series.