The Color Project: Part One

The National Coil Coating Association Technology Committee has been investigating color measurement, color difference, and how best to establish meaningful color tolerances. “Color” is a small word, but one with lots of tentacles. You see a blue car, you call it a blue car. The person you’re walking down the street with also describes this same car as blue. So you both call it “blue.” What’s the big deal? Seeing a “blue” car as it travels down the street is one thing. Putting two metal panels next to each other and comparing their colors closely and carefully is quite another thing. It’s all a matter of perspective.

The funny thing about color—in our industry, anyway—is that even though our response to it is visual, we more often than not simply look at the numbers represented by the familiar L, a, and b values. When you walk through a QC department, whether in a paint plant or on a coil line, you will hear something like this: “The color is plus 05, minus two tenths, and plus 30.” And we—the smart coil people that we are—know that the person is indicating the color of the sample is 0.05 ΔL lighter, 0.2 Δa greener, and 0.30 Δb units yellower than the standard. Of course, there is nothing wrong with this shorthand; we all have our own abbreviated way of discussing certain topics. And there’s nothing wrong with expressing color numerically, either in absolute terms or as a numerical difference between a standard and a sample. It seems we do not have much of a color problem in our industry, so numbers—rather than visual observations—seem to work just fine.

Currently, the coil coating and extrusion industries are the only industries that use Hunter color space (the rest of the world uses CIE color space), so NCCA took on the task of considering whether the North American coil coating industry should consider converting to a new method for color measurement. This new method—now 17 years old (hardly new)—is ΔE₂₀₀₀. You may have heard of it; you may have heard people talk about elliptical tolerancing; and you may have heard that the coil industry is about as old school as it gets regarding color. Rather than debate the issue with these folks, why not find out? And that’s what NCCA’s “The Color Project” is all about.

As we approached this project, we decided to take the easy path first. Why not? Maybe we’d get lucky. What if we just converted to CIE color space and ΔE₂₀₀₀ as our color difference approach. Sadly, there is nothing easy about color. For a number of reasons, which won’t be discussed here, we quickly realized that a simple approach would not work. This forced us to challenge ourselves: what were we really attempting to accomplish? After much thought (remember, nothing about color is easy), we set out to find the best numerical approach to establish meaningful color tolerances. There seemed to be only three options: 1) Do nothing, 2) arbitrarily convert to the new ΔE₂₀₀₀ system without due diligence, or 3) perform a visual observation test, in which a large group of individuals tell us how they see color differences and then we determine how best to express their visual “opinions” numerically. We did not want to do nothing, and an arbitrary switch to ΔE₂₀₀₀ would be imprudent, so we proceeded with the last option. We knew that this would be an arduous journey for all kinds of reasons, but, heck, we’re coil people; we’re tough!

But what on earth can be so tough about asking people to comment on color? Well, everything! The deeper you get into this topic, the crazier it gets. So, rather than inundating you with all the details at this time, stay tuned for “The Color Project: Part Two.” I can assure you that you are not going to believe what the data says.

David Cocuzzi

NCCA Technical Director