The measurement of color always includes measuring some of the gloss component as well. Not everybody is aware of that implication and many articles exist on the subject (see below). In fact a complete book is needed to cover all aspects of the color and gloss interaction. In this article I focus on the correct way to measure high gloss dark colors. Measuring colors including or excluding the gloss component can have a substantial influence on the calculated color differences and jeopardizes the agreement between visual and instrumental assessment.
After reading this article you will better understand the disagreement between color measurements and visual assessments of glossy dark colors.
A practical situation
More than once we run in to the following situation: a dark high gloss color is measured and compared to its high gloss standard. The calculated color difference seems to be much less then visually observed in e.g. a light booth. Conclusion: “color measurements give a general indication of the color differences but not the right answers”. Wrong conclusion. What is often the case is that the standard and sample colors are measured with a D/8 sphere spectrophotometer including the gloss component. And this is not representing practice. So, should you measure high gloss (dark) panels excluding the gloss component? It certainly is a possibility but there is more to say about it as we explain below.
Let’s look at an example of a dark blue color.
The reflectance curve including measured gloss looks like this
Now let’s compare the curve of the Standard color measured with and without the gloss component.
There is a very substantial difference (about 4%) between the two curves. See the pictures for an indication of the differences. So, realize that the spectrophotometer “sees” the color completely different in these cases! Hint: look at the panel in a D/8 spectrophotometer through the observation port yourself with and without gloss (gloss port open and closed). Perhaps you need a small mirror to do that.
Because, the human eye is much more sensitive to dark colors, differences between a Standard and a Sample color will be more noticeable excluding gloss than including gloss.
The influence on color differences
Let’s consider the following Standard and Sample colors including and excluding gloss:
In fact, they have the same shapes because gloss is almost constant over all wavelengths. The graphs on the left (including gloss) are just shifted about 4% as compared to the graphs on the right (excluding gloss). Realize that the average reflection level is only about 5% which means that a 4% shift is very substantial. Because of the limited capabilities of the display the color differences are hard to see on screen but very real in practice.
The Lab values including gloss indicate a color differences of 1.4.
The Lab values excluding gloss indicate a color differences of 4.3!
NB: please realize that there is no gloss difference between the Standard and Sample i.e. both are high glossy. So, we are not looking at panels with different gloss levels where most publications refer to.
Implications of calculating color differences with or without gloss can be severe. Color differences for dark high gloss colors can be much higher if color differences are calculated without the gloss component. This means a substantial discrepancy between visual assessment and instrument digital results because visually we look outside the gloss so the gloss component is not taken into account. Or perhaps not completely, read on…
Practical visual assessment in a light booth.
In everyday life colors are compared in all sorts of light situations. Light can come from everywhere and this can greatly influence the differences we see between colors. Especially with dark colors. Usually professionals use light booths to compare colors because these mimic practical situations in a very reproducible way. It is important (as we just saw) to consider how the gloss component of the light influences the color.
Let’s look at visual assessment in a light booth from different perspectives.
The visual observer looks at the standard and sample in the color light booth. Light coming from above and illuminates the complete light booth. The blue part in the white light is primarily reflected by the blue panel. So, that is the reason why the observer perceives a blue color.
At the same time white light from the light source also illuminates the grey back panel which in turn will reflect a small portion of the white light. Some of light/reflection from the grey back panel is exactly opposite of the observation angle. The spot where this light/reflection comes from is indicated as the “Gloss spot” in the drawing.
This light/reflection causes some gloss to be added to the color reflection thus making standard and sample a bit paler. Inserting a black back panel would get rid of most of the direct gloss making the color less pale while inserting a white back panel would increase the direct gloss observed a lot and the color would appear significantly paler. In addition, the visual color differences will be less.
In the case of a standard light booth the walls vary between light and dark grey and hence influence the visual perception in some way. Looking at the color differences above the right correction must be found to correlate with visual assessments.
One way to do such a correction is to measure including the gloss component and to subtract a certain amount of gloss from the reflection values of both standard and sample. For high gloss this would typically vary between 2.5 and 3.5% reflection. This is less severe than measuring the color excluding the gloss component with a diffuse/8 spectrophotometer because the difference is then around 4%. In addition, measurements including the gloss component are more stable than measurements without the gloss component.
Another possibility is to measure the standard and sample using a 45/0 instrument. This is not a bad idea however the gloss component is then excluded almost completely and the reflectances will be even lower than using D/8 excluding gloss. To still achieve a good correlation, it would be a good idea to even add some gloss of around 2% to the measured reflectance values.
Some companies provide automatic ways to deal with gloss, however they are not perfect and probably not tuned to your situation.
If you need to tune visual observation of standards and samples with digital measurements make sure to choose the right methods to handle gloss in your calculations and visual assessments.
Need to know more about this subject or other color tech related subjects please contact us: Color Technology Consultancy coltechcon.com; firstname.lastname@example.org
Roel Gottenbos is an experienced color research manager and developer of game changing color tools. He helps companies to improve their color processes from raw material to end product. Within Color Technology Consultancy Roel has the privilege to work together with Erik van Biemen who is as passionate and knowledgeable about color.
Some references relevant for the article: