Demystifying daylight (Part 1)
By David Crowther
In this article David Crowther looks at the myths and misconceptions about simulating daylight that everyone involved in evaluating colour should know.
Ok, you may think that we are starting to go slightly overboard with this whole daylight thing and evaluating colour. But hang on: One of the rules of colour management is that we use a consistent illuminant for colour measurement and print assessment. Illuminant meaning a daylight light source (D50 or D65) with the Spectrophotometer for measuring a printed ICC profile chart.
The monitor used for viewing our images should be set up to a similar daylight standard 5000~6500K and our viewing system (light box) should accurately simulate D50 or D65. Every part of being able to control colour is pivotal on accurate simulation of daylight. The component that lets us down most frequently is the viewing system or daylight simulator.
It is proven that natural daylight is the only source that will not distort our perception and ability to evaluate colour. As stated at the outset, most well managed colour software applications specify a form of daylight as the light source of choice for making accurate colour evaluations.
The problem with natural daylight is that its appearance and spectral characteristics can change dramatically from day to day, season to season and even during a single day. Morning sunrise tends to have a reddish hue.
An overcast day appears grey and drab, and a crystal clear bright sky appears blue. Changes in daylight quality are affected by atmospheric conditions, the change of seasons, time of day, pollution, altitude and even your location – city versus country.
Many times I have heard the comment from visitors of the Northern Hemisphere, to our shores, comment about the clear, bright days we have. All of these factors affect our ability to accurately evaluate the colour quality of a product, hence the need for daylight simulation.
In the following we will discuss and dispel the common myths and misconceptions about daylight simulators (viewing systems). Our review of the technologies, industry standards, practices and issues relating to simulating daylight in colour evaluation may seem arduous and drawn out. However, we firmly believe that it is only proper to lay down the facts in an open forum.
Top Four Myths about Daylight Simulation
Myth #1: All daylight simulators are the same. False!
Natural daylight has been thoroughly defined by the CIE, International Commission on Illumination, as having three specific attributes: (1) spectral power distribution (SPD); (2) chromaticities and (3) correlated colour temperature. The ability to simulate all three characteristics is critical for accurate daylight simulation.
There are some simulators that meet the criteria based on one or two of these attributes. Without meeting the criteria for all three attributes, the simulator will be deficient in rendering colour.
The facts about spectral power distribution (SPD)
The accompanying graphic (Fig. 1) displays the relative energies found in different phases of daylight.
The energy content of a light source determines its ability to render colours accurately. This energy is known as spectral power distribution and is determined by measuring the lamp output with a spectroradiometer. It should be noted that the human eye is sensitive to light energy that falls within the 400 to 700 nanometer wavelength range, which is why the X axis on each graph displays that wavelength range.
Figure 1 shows ideal curves for three standard daylight illuminants: D55 or noon sky daylight, D65 or average daylight and D75 or north sky daylight. Although the curves are uneven, all colours of the spectrum are present in relatively equal proportions.
Therefore, all three phases of daylight have similar curves, indicating similar spectral content. Thus, colours that match under one phase of daylight will also match under the other phases of daylight.
Why all daylight simulators are not the same
Simulated daylight can be achieved through several different lighting methods, including wide band fluorescent (commercial daylight) seven-phosphor wide band fluorescent (patented GretagMacbeth AG* technology - * an X-Rite company) and filtered tungsten halogen (patented GretagMacbeth AG* technology - *an X-Rite company). All produce light energy with a correlated colour temperature and chromaticities equal to the phase of daylight that they simulate. However, closer evaluation of their energy content reveals more obvious differences in the quality of the simulators as shown in Figure 2.
and correlated colour temperature.
The next myths reveal the truth about these two attributes.
Myth #2: As long as the light source is D65, it’s a good daylight simulator. False!
Let’s take another look at the spectral power distribution curves in the previous chart. All the sources have the same apparent or correlated colour temperature of 6500 degrees Kelvin, yet different spectral power distribution curves. This discrepancy between colour temperature and spectral power distribution uncovers a fundamental flaw in using colour temperature alone to specify a light source.
To further explain, let’s take a look at the relationship between chromaticity and correlated colour temperature. This relationship reveals just how widely the spectral power distribution can vary based on colour temperature alone.
Why two light sources can have the same colour temperature, but different colour rendering quality The colour temperature designation for a light source identifies the appearance of the light source as compared to the 1931 CIE Chromaticity Diagram (see Figure 3).
The chromaticities are the x and y coordinates for a light source. These coordinates are plotted and correlated to a tolerance acceptable for a specific colour temperature. Herein lies the flaw. Using the correlated colour temperatures provides an infinite number of chromaticity coordinates – all with the same colour temperature! So lamps that match the correlated colour temperature can vary widely in their ability to simulate CIE D65 daylight as well as their ability to render colour accurately or
In our next articles we will continue to look into the details of demystifying daylight, covering the myths and misconceptions about simulating daylight that everyone involved in evaluating colour should know, by discussing Myths #3 & 4.
National Manager, X-Rite products