Communicating about Color

Rick Roth is president of Pawtucket, R.I.-based Mirror Image. In recent years, his business has taken home numerous Golden Image awards in various categories, as well as top honors in the industry media's various printing competitions.

Print burgundy please.

Sure, you mean a dark purple, like the wine?

No, not that purple, more reddish.

You mean like Harvard Crimson?

I don’t know what that looks like, more like Boston College.

You mean maroon?

I don’t know. Well, a real dark maroon, but brighter.

You mean very dark red?

No, that’s too brown.

More like wine.

Now would that be Pinot Noir, Merlot or Carmenere?


We happen to have some bottles here in the shop but do you mean from the Montrachet or Corton vineyard?

Communication about color can be as much fun as an Abbot and Costello routine. It can be frustrating, exasperating and, if you get it wrong, very expensive. Let’s examine some ways to communicate about color and some fine points about making our customers happy about the color of ink printed on their shirts.

What are we matching?

This should be part of any screen printer’s mantra—repeat after me: “What are we matching?” Is it the previous screen printing company’s poorly printed batch of shirts? Or is the ink on this shirt supposed to match the color of another garment? Is it supposed to match other promotional materials? 

Most printing professionals use the Pantone Matching System (PMS) for communicating about color. The Pantone folks produce books of carefully-produced printed color chips that have many, many colors and a corresponding number to each color. So, a customer can ask for Pantone 195 Burgundy and the print shop down in Tierra del Fuego can print that same color the customer’s looking at in Boston. 

However, even Pantone colors pose issues. For example, Pantone 505 doesn’t indicate if it is U (uncoated) or C (coated). Or, for another, Pantone 19-1629 is a Pantone fabric color, but most customers and most printers don’t have fabric books. This would be the reason for the offset U and C colors. Still, inks for screen printing on textiles don’t really match either the U or C. But they are the best shot we have for communicating about color. 

In recent years matching Pantone colors has become easier as ink companies have made formulas for print practitioners to follow. And, quality ink companies have consistent ingredients so the formulas achieve the colors customers want. There was a time it felt like one needed an advanced degree in fine arts with color theory to mix inks properly, but now you can buy a scale and follow the recipes that the major ink companies provide and voila, accurate ink colors.

Good on paper

Let’s go back to the aforementioned burgundy. While some burgundy wines improve with age, the Pantone books that communicate what burgundy looks like do not age well. Watch out for old Pantone books, particularly if they have been in the sun. The colors can fade, and even fade unevenly. As painful as it may be to plunk down another $75 for a new book, printing 36, or worse, 3,600 sweatshirts with the wrong color could cost more than the $5,800 price tag on a bottle of Romanee-Conti 1990 vintage burgundy wine.

Sometimes it’s necessary to grill customers when they specify colors. If they are getting their Pantone colors from a computer monitor, ink printed on fabric is never going to look like a color from a backlit computer screen. There is also the little issue that only about 1 percent of color monitors in the world are color corrected. Customers that expect bright colors printed on shirts to match what they see on their monitor are, well, both uneducated (formerly known as dumb) and commonplace, and the print shop is the one who has to deal with these expectations.

Worse than customers expecting colors on the monitor to show up on a shirt are those who look at a paper printout and expect that color. Even the best color printer isn’t going to match ink on a shirt. If producing a mock up sample on paper, it’s important to communicate the colors may not be accurate.

A matter of perception

Another pitfall to color matching is metamerism—that is, the biology of how eyes function that describes variation in color. There are several types of metamerism or, more exactly, metameric failure, that wreak havoc with color matching.

Geometric metameric failure is where a color looks different depending on the angle from which it is viewed. The cure is to look straight at a printed sample, not to stand too close, and sometimes just squinting helps in comparing samples. This is an issue particularly when trying to match very glossy samples to metallic inks. Look at them from a few angles. Also, tell customers that the ink may look different depending on how you look at it.

Observer metameric failure describes the fact that not every person sees color the same, with the extreme example of color blindness. There are subtle differences in just about everyone’s color perception and not much can be done about it. The differences are not usually that great, so no fair using this as an excuse for every color issue with customers. An everyday solution is to have a few people in the shop judge on whether colors match.

When a little area of color doesn’t look the same as a big area, we are dealing with field-size metameric failure. This most commonly occurs in our T-shirt world when customers are shown a swatch of T-shirt fabric and it doesn’t look the same to them when they get a whole T-shirt. One cure is to send a whole sample shirt. If a swatch is the only option, make it a fairly large swatch. Unfortunately, this is what makes using a Pantone book so frustrating as well. Often a large field of printed ink on a shirt doesn’t look the same as our little Pantone chip, but when you hold the chip directly against the ink it matches exactly. There is little cure for this except to simply be aware.

Metameric failure from context is when colors look different to the viewer when next to different colors. I.e. that little green spot of ink on a white shirt does not look the same when surrounded by a large amount of orange ink. The color swatch of green also might not look the same on an orange shirt. There is not much to be done about this except to be aware and inform customers of it.

The most common issue with color communication is illuminant metameric failure. Without light, there is no color. The issue then becomes, “what kind of light.” Different wavelengths of light cause colors to look different to human beings. Incandescent and halogen lights (common light bulbs) are very “warm” and make colors look more yellow or reddish. Fluorescent tubes are very “cool” and enhance or add blue and green to color. Natural sunlight is particularly touchy; in the morning can be warm and add yellow, in midday add blue or bleach colors out, and late in the day, tends to add red again. Even further complicating it is that winter light is warmer (color, not temperature!) and summer light is cooler.

So what is a poor printer to do? There are light boxes that are “color correct.” Still, it doesn’t really provide how color will look to real, live human beings in real, live situations. More practical is to find a place in that has fluorescent and natural light and a neutral background. For good measure, take a look in just fluorescent and just natural light and see if anything wacky goes on with the colors. Advise customers to do the same. 

One might think that, when comparing colors, two samples would be influenced by the light in the same manner. But that is wrong. Minute, almost imperceptible differences can be exaggerated tremendously by lighting changes. 

Communication about color is communication, and as with all communication, the more information that can be given to customers about the variables, the better. Provide a little technical background so they know what to expect, as nobody likes surprises. There is not a perfect way to keep every customer happy, but with some systems in place and some realistic expectations from customers, you can make some nicely-printed shirts that are the “correct” color. Or at least close enough to get paid.

Color Conversation Odds and Ends

• The words aqua, burgundy, crimson, green, gold and fire-engine red do not mean the same thing to everyone. They are the most commonly misconstrued colors. Blue and green are synonymous in some languages, so particularly be wary of these colors and the mixes thereof.

• Communicating fluorescents or metallic ink colors by Pantone doesn’t work very well. Swatch them onto fabric and send them to customers to look at.

• Avoid good color matches going bad by placing a shirt that has been approved at the end of the dryer. Color shifts can happen slowly, so shirt one and shirt 15 may be close in color enough to look alike, but shirt one and shirt 400 may look unacceptably different.

• Test colors when they are cured. A wet sample of ink does not necessarily look like a cured sample. Also, a sample that is still hot may not look the same color as when it gets back down to room temperature.

• Some manufacturers, American Apparel and Hanes for example, can provide the nearest PMS color numbers to their shirt colors. This is good for preliminary communication but always get a full shirt in the customer’s hands when possible.

• Don’t promise exact color matching with discharge inks because it is too difficult. 

• Have a set of inks your shop offers. Instead of matching every Pantone color that comes along, keep certain colors mixed and ask customers to choose from those. Charge more for other colors.

• Ask customers to sign off on a printed shirt, literally signing a sample shirt that is used to match. Don’t risk matching a customer’s unsustainable color direction.

• When taking a digital photo to send to a customer prior to production, put your Pantone book, open to the color in the print, in the photo as well. Even if the customer doesn’t have a color-correct monitor, they usually can see how closely the print and chip match.