Ask the Expert Question-and-Answer Archive

by Ron Joseph

October, 2004

Converting Fed. Std. 595B Color to RAL Color

Q. I have been asked to produce a plastic part in Federal Specification 595 colour 36320. As I am in Europe my material suppliers are requesting that I supply them with a RAL or Pantone colour reference for this colour. Do you know the equivelent reference in RAL or Pantone?

Thank you for your interesting question. Since this topic is as little above my head, I referred it to a colleague, Danny Pascale who has designed a color software program, BabelColor. Danny replied to your question as follows:

FED-STD-595B color #36320 is a flat/lusterless color. If your plastic part has a shiny finish, it may be difficult to exactly match the appearance that your customer wants. You should ask your customer how he determined that this was the correct color and what is his tolerance for color difference (DeltaE).

However, lets look at how we can get an equivalent in another color system. I selected "RAL Design" because you do not need to have an actual sample to convert to it.

Most colorimeters will provide colorimetric data in L*a*b* D50 (2 degrees CIE 1931 Observer). Many spectrophotometers will enable you to get spectral data which can be processed to give you colorimetric coordinates for any Illuminant or Observer; its mainly a matter of how much you want to spend on add-on software. Most spectrophotometers will give you raw spectral data plus L*a*b* D50.

RAL numbers correspond to L*C*h* values, presented in the h*L*C* order (called HLC by RAL). L*C*h* vales are simple derivatives of L*a*b*. The problem here is that RAL is defined for the 10 degrees CIE 1964 Observer AND for the D65 Illuminant.

Thus you cannot take L*a*b* coordinates obtained with a 2 degrees Observer and D50 Illuminant and just compute L*C*h* from them. You need the spectral data, and this is what I did. (In this very particular case, the error obtained by using a 2 degrees Observer would not be that bad, about 0.6 DeltaE*ab, but please do not do that, the error could be way more!).

I used spectral data from a measured FED-STD-595B #36320 chip which I converted to 10 degrees D65 CIE 1964 XYZ data. In turn, I determined the L*a*b* and L*C*h* values from there.

The exact converted HLC values are: 243 61.6 5.7
The closest RAL chip card is: 240 60 5

The color difference between the two is 1.7 CIE94. Most of the difference comes from the lightness (L* of L*a*b*), where the converted values are brighter than the closest RAL chip (do not compare the HLC numbers individually, they will give you a wrong impression).

If you take into account my measurement you should expect a CIE color difference of about 2.0, to which you should add the precision of the RAL patch on the other end, if the final product is compared visually.

In conclusion, I would first verify how your customer got the color spec (visual assessment, measurement, selection in the FED-STD-595B book of color, guesstimate). The more reproducible the method, the more confident you will be in matching the color he wants. You should also agree on a verification method and a color difference limit. Will your customer accept a barely noticeable visual color difference between the Fed. Std, color and the RAL color? If so, I will need more specific information from you.

Finally, you may want to add a gloss specification. Ask your customer what gloss level he expects as well as the gloss tolerance.

When you have all this information please get back to me.

Danny Pascale