Matching Paint Code to RGB Color

Is there a way to translate RGB color to a paint code so I can get paint mixed by a local paint shop?

The guy I talked to at my local automotive paint shop only spoke in terms of brand codes and did not know of any type of translation. I made a prototype avionics panel for a customer and per his request picked out a color for the first prototype. Now he is requesting a color in terms of RGB. Should I look for a way to come up with a paint code or tell him to go pick out a paint code?

Also any advice on paint is welcomed. I used an automotive adhesion promoter then automotive 2 stage urethane acrylic. The part I am painting is polycarbonate. The first prototype came out good but I have not done any durability testing on it. I might get a chance to do a small production run on this part so if I should be using something different I would love to know.

Thanks!

Try this.

Color Converter

Short answer: "no". There are a lot of reasons you cannot reliably go from RGB to paint code. I am not going to get into additive vs subtractive colors, nor CIE Lab color space, nor gamut limitations, nor color models, nor any of the other highly technical crap behind the standard "your monitor may not show these colors correctly" warning.

Your customer needs to pick a color based on a paint chip, preferably a real physical paint chip. Seriously: send him/her to a hardware or big box store with a paint department. A decent reproduction of the RAL or Federal standard paint color charts would work, although the colors are fairly limited. The Pantone process color chart would also work although it's intended for ink on paper, not paint on metal. (BTW, Pantone has several color specification systems. Process color is what you probably want.)

Your paint vendor can scan a paint chip and give you the closest possible match in their color code system. The paint guy should be able to find cross-reference charts for RAL, Federal standard, and Pantone process colors. If he can't, one of his competitors probably can, although I am sure they would prefer a paint chip to scan. It probably will not be a perfect match, but it will be a very close match unless your customer picked some wild out-of-gamut color that requires a special pigment. Examples: highly intense purples or other secondary colors, flourescent or "neon" colors.

A simple way to match RGB in automotive paint MIGHT to to print it out and have the paint store scan the sample. IF you have a printer that will reproduce the color accurately and nicely on a smooth glossy paper with no striations (laser not inkjet) AND they can scan it, they MIGHT be able to match it to a standard color or else custom mix it. There are lots of ways to go wrong in that process, not the least of which is the scanning. For example I tried to match a RAL 6011 green sample by scanning a piece of sheetmetal, and the pebbled texture on the part caused the scanner to make a different result on each attempt. Sometimes you get lucky, a friend scanned a '60s Yamaha side panel and matched it to an '80s Mercury color. And sometimes just close enough is OK.

Matching with a known code can be a crapshoot too. I have a handful of RAL6011 samples I've ordered from several sources and none of them match the machine or each other.

I tell my clients I'll do the ''known'' farm machinery colours and RAL - and nothing else, if they want fancy they can supply the paint.

If you do get lumbered with something fancy or, and I'd advise this - any colour, .......even RAL, do a sample and get it signed off.

Possible stripping and repainting id a costly giant PITA - BTDT at my own costand carry the scars.

No, process color is the four colors used for the dots in a four color half tone, cyan, yellow, magenta, and black. Pantone colors are meant to be "SPOT" colors, premixed to the exact shade rather than approximated with a blending of dots. If using a Pantone book for paint, you want to be using the "C" swatches printed on coated (glossy) stock. Biggest problem with a Pantone book is the swatches are likely too small to match to. You can get bigger swatches from Pantone, but they ain't cheap.

If you are angling for a production run, don't send the customer to a big box store, send him to an automotive supply that carries the brand paint YOU like to use, and have him pick something. Then you have something reproducible.

Dennis

Modelman said:

No, process color is the four colors used for the dots in a four color half tone, cyan, yellow, magenta, and black. Pantone colors are meant to be "SPOT" colors, premixed to the exact shade rather than approximated with a blending of dots.

Click to expand...

Yeah, that's valid. Thanks for the correction! I was trying too hard to avoid the half-dozen other Pantone color systems like cloth, textured plastic, etc, etc.

Yeah, have him hold in his hand the item that's the color he wants, and go to an auto paint store and have it computer matched. Bonus -- that leaves you out of it!

Chip

Color matching sucks, and trying to do it long distance is fraught with the danger of misunderstanding. Everybody sees color differently. I'm reminded of a story an old time modelmaker, now recently passed on, told me years ago.

The customer had supplied a color sample on a metal panel (often called a "drift card") for a new product they were building a model of for the early ad and catalog photos, which was common in the day. It was a teal/turquoise color and he was having a hell of a time matching it... he'd mix a batch, spray up a sample, let it dry (just a couple minutes, this was back in nitrocellulose lacquer days) and take it and the drift card into his boss. "Too blue," the boss would say, even though to his eye he thought he nailed it. So, he'd add a bit more green, and do the whole sequence again. "Too blue," the boss would say. Finally after four or five tries, in desperation, he painted the drift card with his mix when he sprayed his sample, so they were both painted with the same paint. "Too blue," said the boss.

Don't let yourself get trapped in this endless exercise. Make the customer pick the color and specify it by manufacturer number.

Dennis

I have a book "Paint Technology Handbook" from CRC Press. Chapter 5 "Color Matching and Color Control" (just a small paragraph). Get the book pdf if you care to read more.

Colorimeters measure color in terms of three numbers called tristimulus values—
X, Y, and Z. These values are the amounts of three primary colors—red (X), green (Y),
and blue (Z)—needed to match or specify a color under illuminating and viewing
conditions as standardized by the ICI. Each color has its own tristimulus values, and
identical values are identical colors. Differences in tristimulus values between colors
can be analyzed to determine the direction and magnitude of the color difference.
To facilitate the translation of the color difference from numbers to the more
meaningful visual descriptive terms and quantify color difference, the XYZ values are
converted into the Hunter L, a, b Scale. This transformation is simple, and all modern
instruments offer it as an option. L is lightness–darkness, a is redness–greenness, and
b is yellowish–bluish. If the difference between the trial values and given color in
L, a, b values was −0.5, −0.5, and +0.5 respectively, the trial values indicate that
the color is darker, greener, and yellower than the given color. If the color difference
is unacceptable, then the trial must be corrected by adding white to make it lighter,
adding the bluest colorant to make it bluer or less yellow, and adding the reddest
colorant to make it redder or less green.

Without getting into any of the science, I recently worked with the paint guy at the local ACE hardware to get a shade that I wanted. The first thing that I learned was that they did not have a sample chip that came close enough to what I wanted.

But the second thing was the real surprise. I wanted him to record the changes that he made to a standard color that we came up with and he did. The paint he mixed for me was the ACE brand of latex. When I asked him if that set of numbers would work with an oil based paint he said "NO!" Each brand of paint that they carried, and they had several, needed a separate set of numbers for mixing the pigments into the base. There were even differences between the latex and oil based paints for the same brands. In short, each paint had to have it's own set of numbers to get the same color.

I doubt that any of them can be translated from any R-G-B values. That kind of numbers are generally used with respect to video signals and even there the actual colors displayed on a TV or computer monitor will be somewhat different depending on the actual colors of the phosphors or other coloring means used for that particular display as well as how it is set-up in the factory. There are just plain too many variables in the process.

Mud said:

Matching with a known code can be a crapshoot too. I have a handful of RAL6011 samples I've ordered from several sources and none of them match the machine or each other.

Click to expand...

I can TOTALLY agree with Mud on this one. I have found over the years that no two suppliers can equally match a specific RAL colour despite "supposingly all mixing to the SAME standard" - I think "BUYER BEWARE"

When you find a supplier with a colour that suits, stay with that one - don't go elsewhere if you need more !!

John

Good pantone colour charts use to have a RGB table, not got a new one here only a really old one but its there. Colour matching in the print trade use to be a full time all out task. Its use to be std practice to calibrate monitors back when CRT was king and it made a fair difference. The RAL colour range is a lot less than the pantone colour range, but theres more than enough cross over to make most people happy.

Matching any RGB colour combo is pretty easy, going that way, thing to remember is its not as easy to go the other way and several colours can not be made accurately using RGB alone.

Mixing colours to high precision takes care, few machines are really accurate and if you want bang on, you need to test paint - print and tint to perfection, its not hard, just takes a bit of time. Application method + different manufacturers bases all changes colour mix formulations. Human eye is more than capable of telling parts per million difference in colour pigment mixes too which makes it hard. Then you get into the whole who sees what colour and how and it gets ever worse.

I have a friend with a pretty big body shop. At some point a few years ago he told me they didn't bother with all this color-matching by eye and paint codes stuff anymore. Cars fade, the original paint color never matches anyhow. He has some kind of scanner which interfaces to a paint mixer to match the color. Scan the car near the fix, machine mixes up the paint, matches much better than anything they did before.

I dunnno but it's worth a few phone calls.

EPAIII said:

But the second thing was the real surprise. I wanted him to record the changes that he made to a standard color that we came up with and he did. The paint he mixed for me was the ACE brand of latex. When I asked him if that set of numbers would work with an oil based paint he said "NO!" Each brand of paint that they carried, and they had several, needed a separate set of numbers for mixing the pigments into the base. There were even differences between the latex and oil based paints for the same brands. In short, each paint had to have it's own set of numbers to get the same color.

Click to expand...

A couple more comments...

The Hunter l,a,b system mentioned above gives a set of numbers that are akin to map coordinates; they tell you where the color is in "color space", but not how to get there.

The mixing formulas are like a recipe to make the color, using that brand's tinting colors. Since each manufacturer uses different pigments in those tinting colors (each trying to get the "best" color palate at the lowest cost) it stands to reason that the formulas won't be the same from one manufacturer to another, or between alkyd and acrylic resin systems. Given a standard color to match, each manufacturer will come up with a formula that is close; a so called commercial match... That is, if you paint two boxcars with different brands of paint, the result will appear to be the same color, but if you touch up either one with the wrong brand paint, the patches will likely show.

Trying to match through the RGB system is even worse. The RGB colors on your monitor are colors of the light source (the glowing phosphors), not light reflected from a painted surface.

Just like film photography, or old time photocopiers (remember them?) the more steps between the original and the end product, the more uncertainty is introduced in the outcome. If you want to finish these items with automotive paint, send the customer to a supplier of that paint to select his color, and specify it by manufacturer number. If you are really concerned about customer approval, spray up a sample of the paint you source and send it to him for approval... but be sure to build into the price sufficient money to pay for several samples and the dink-around time involved.

Dennis

You also need to be careful about the light you are viewing color samples in, should be the same for you and customer

I'm going to stick with my original short answer: "No, you can't." Have the customer pick a paint chip.

sfriedberg said:

Short answer: "no". ... I am not going to get into additive vs subtractive colors, nor CIE Lab color space, nor gamut limitations, nor color models, nor any of the other highly technical crap behind the standard "your monitor may not show these colors correctly" warning. [remainder of excellent comment deleted]

Click to expand...

And let's not go into metamerism either. This is an extremely technical subject, and I echo the short answer.

richard newman said:

You also need to be careful about the light you are viewing color samples in, should be the same for you and customer

Click to expand...

Finally! And thank you!

Advert dept. next door to me at one Day Job had a rig that could select among four different 'Kelvin temperatures' of light.

Some Big Box have a taste of that on paint-brand displays now as well.

Otherwise, fluorescent of half a dozen Kelvin ranges, various incandescents and substitutes, natural daylight and with what latitude, sky conditions, time of year, time of day, and 'direction' (a good North light...) worst of all any energy-saving lamps with narrow spectrum, and that change with age or even 'on time' since started that day, and how hot or cool a day as well?

Can't get there from here until the LIGHTING variables are nailed down.

Sami's way "if they want fancy, they can supply the paint" is the only sure-thing.

And customer supplies the paint, not just 'specifies' it.

Well, in theory, the color matching "machine" should be providing a suitable light source. The problem comes, even with a perfect machine match, because of metamerism, if the customer compares the samples under light of a different color temperature, the perfectly matched samples may appear to not match. As I said, color matching sucks. Make the customer pick the paint to be used.

Dennis