The Problem of Color: Paint Matching

There has been much speculation on various web posts about why paint color for a specific prototype varies from model manufacturer to model manufacturer. Questions typically center around the color of the paint in comparison to the prototype, or original, train car. As I've been working on the next batch of Wheels of Time 70-ft Heavyweight Baggage-Express cars (to be announced shortly), it occurred to me that paint-matching is more of an art than a science -- there are a lot of variables. Here are some of the variables that effect color matching on model trains.

1. Paint on the real railroad rolling stock changes over time. Newly applied paint looks different than paint that has been weathered. The paint formula itself may also change over time.  For instance, Norfolk & Western "Tuscan Red" paint changed around 1958. It went from a tomato soup red to a more maroon red.
2. Variation in reference points. Unless you are modeling a brand new diesel locomotive, you'll be basing the paint samples on equipment that has been retired . . .  not exactly newly painted. Also, Pullman green for Great Northern may not be the same as Pullman green on other railroads. Remember this was before computerized color matching as we have in modern auto body shops. If you're basing your color on a picture, photographs of painted rolling stock will also introduce variation, based on lighting condition and surroundings.  A car photographed under the noon-day sun will look different than one taken in the late afternoon. The advent of computers has not always resulted in more precision: a JPG image on your computer screen will vary according to the monitor's color settings. An image found online may look accurate, but hue, tone and intensity are easily, and often, changed with programs like Adobe Photoshop. Such an image may only meet the standard of "what looks good" in the eyes of the poster.
3. Popular understanding. Popular railroad culture may perpetuates an image which isn't accurate. For example, the Atlantic Coast Line green heavyweight baggage cars appear to have always had black roofs, since most of the available images are from the 1960s. But this doesn't paint an accurate picture. For the record, ACL HW baggage cars prior to the 60's had roofs that matched the color of their carbody. In another example, the Kato SP Daylight painted train set is Kato's interpretation of a prototype's original color. Should another model train manufacturer match Kato's paint, stripe and lettering so the hobbyist can couple together cars from different manufacturers? Or should a modeler strive to match the original, even if the result is clashing models?
4. Paint and inks. Paint on prototype metal differs from paint for plastic hobby models. Today it's common for model train manufacturers to specify color with Pantone color specs. In fact, the Union Pacific Railroad actually specifies their car paint by Pantone color specs. The Pantone system specifies inks by a color index, allowing for standardization. However, Pantone specs are intended for printing and web publishing and not for plastic surfaces. Hue, tone, and intensity all differ on a printed material when compared to a painted surface. Replicating the color from a printed image on plastic is very difficult. 
5. Substrate and painting. Pad printing inks over a darker painted surface can become a problem. For instance, "dulux" lettering applied over a Pullman green paint will look different than over a maroon painted surface. The real railroads didn't have this problem. Also, the hue, tone and intensity of the color depends on a number of factors: how the paint was applied, what percentage it was thinned to be sprayed, the pressure of the sprayer, the surface texture of the object, and the number of coats or applications. Furthermore, when the factory matches our paint chip samples, the matching is dependent on a technician's knowledge and experience of the paint being used and their perception of color.
6. Lights and scale. Since the color that we see is created by reflection of light off of pigment, lighting conditions matter. Model railroad layouts are usually not lit by daylight but illuminated by fluorescents, LEDs or incandescent bulbs. A yellow or bluish cast makes your painted models look different. Further, objects that look "normal" on larger scale, may look too dark on smaller scale.

Now, what would be ideal to help a model train manufacturer in finding paint matches for a specific prototype? I wish the various railroad historical societies could provide paint samples (on plastic!) for the various elements of their specific rail systems  . . .  that would go a long way.

Comments

[Red Beard]

Thank you for taking the time to explain this. Clearly you folks care about accurate color; unfortunately many manufacturers don't. I'm a big Union Pacific fan and trying to find HO models in that fairly bright rusty red oxide color all their 1940s & '50s freight cars were painted is just so frustrating for those of us who remember them. -- Red Beard