Home>Artist Resources> Color > Delta E: A Key to Understanding Lightfastness Readings

Delta E: A Key to Understanding Lightfastness Readings

21 thoughts on “Delta E: A Key to Understanding Lightfastness Readings”

  1. I am absolute;y fascinated with information like this. I have been working with JND as a concept for a little while but you have added a completely new dimension. Thank you!

    • You are SO welcome! We love that you find this information fascinating and valuable as clearly it is part of our passions as well.

    • Great question. The measure would actually be the same since Delta E is neutral about how the change comes about and simply measures the difference in color between two readings. So, for example, when Cadmiums or Ultramarine Blue are exposed outdoors in an environment with moisture or humidity, chemical changes take place that are not really classic “lightfastness” issues, but which will cause an Ultramarine Blue to bleach out, or a Cadmium Red to get darker. Or we will see cases where a particular pigment will appear lighter after outdoor exposure, but due to chalking of the surface rather than fading, which is a form of erosion different from purely a fugitive color. Ultimately, however, while all of these changes can be described in terms of Delta E, that does not mean all of them are reflected in ASTM’s lightfastness rating even though they can impact how a pigment is used. So, for example, returning to our earlier example, while Ultramarine Blue and Cadmiums are all given a Lightfastness rating of I by ASTM, their use is restricted to indoor applications. As most paintings are exhibited and live their lives indoors few artists notice this, but for those who paint murals they will find them absent from our list of recommended colors for exterior applications. For more information on that aspect, the following are good links to look at:

      Exterior Mural Tech Sheet – Recommended COlors

      Selecting the Best Exterior Mural Pigments

  2. Technical question. I am shooting a polyurethane powder that is required to be less then 1.0 delta E. For some reason I keep coming out at 2 to 2.5 Delta E. What can I do to get this lowered the manufacturer of the powder came and he couldn’t get it below 1.0 either. It has to be something with the cure schedule or timing, took the same Box of powder to another vendor and he got it below a 1.0. The manufacturer did test my oven, and it has absolutely no issues, the temperatures are right on, no cold spots, everything came back 98.2 % efficient. I have provided a copy of the Cielab results below for the approved piece and the piece I shot.
    Cure schedule is part @ 400F at 10 mins.
    Approved delta E piece: Delta was 0.9 CIELAB – L:78.3 a:2.5 b:-7.3

    Piece I shot.
    2.2 Delta E
    Lab: L:83.2 A:3.5 B:4.1
    You can visually see that the piece I shot is brighter then the approved piece what can I do to get this darker and extract the blues while curing.

    • Hi Breet –

      I am not going to be able to help in terms of the preparation or formulation of your materials – that is just not something we work with here, being solely an artist paint company specializing in acrylics, watercolors, and oils. However, I do have some questions that might help me understand things better.

      What settings are you using for your spectrophotometer, and what type do you have? Different settings as well as the different equations used to calculate Delta E will give different results. As mentioned in the article, we use CIE 1976, or dE76 for calculating the difference – but that is not the only option and is not the one used by most commercial paint companies, who often rely on CIE 98 or 2000. Our use of de76 has to do with the particular history of lightfastness calculations for artist materials. So you might see what your results are when using these other formulas. And on the reason why these can give different results, I thought this article was helpful:


      and the difference in the equations is decently covered here


      In terms of the light source, we use D50, and set the standard observer angle to 10 degrees. Lastly, make sure to choose Specular Component Included (SCI) for your readings to compare actual color divorced from the impact of specular gloss on shiny surfaces. For an explanation of that see the following:


      We hope that helps some and if you can let us know what you are using and the settings – for both you and the other people who are getting different readings – that can help if further questions arise.

  3. Can a color, say pms 144, be in the blue green quadrant? Wouldn’t that become a different color? Become dirty?
    Should quadrants be considered for color consistency? Meaning I should be in that quadrant each time I want to repeat the color?
    Or lastly, Isn’t a color either in its correct quadrant or not? It can’t be moved from the correct area into another area without “changing” the color? Please help. Drowning under the -b value!!!

    • Hi Tony –

      Happy to help, but first am a little uncertain about the question as it relates to -b values. The Lab equivalent to pms 144 (assuming you mean Pantone Matching System’s 144) is reported as L = 66.88, a = 30.02, b = 72.67, so there is no -b value involved. Perhaps I am misunderstanding the issues with that specific color? In any case let me answer your questions in a more general way:

      Can a color pms 144 be in the blue green quadrant?

      Not if we are talking about that specific swatch, as described above. It will be defined as laying in the Red Yellow quadrant, positive a, positive b.

      Wouldn’t that become a different color? Become dirty?

      Yes, absolutely, as you would be introducing blue (or more technically a ‘blue bias’) into a color that started off as orangish.

      Should quadrants be considered for color consistency? Meaning I should be in that quadrant each time I want to repeat the color?

      Yes. A color’s Lab values are absolute, in the sense that a specific swatch will always read the same. That said, there are almost always some variations that might have slight differences that we perceive as the same. But the variation would be small.

      Or lastly, isn’t a color either in its correct quadrant or not? It can’t be moved from the correct area into another area without “changing” the color?

      Correct! Moving a color from where it is, meaning changing its Lab values, will always define a different color. And definitely changing quadrants would describe a very different indeed.

      Please help. Drowning under the -b value!!!

      Happy to throw you a colorful floating ring buoy to help prevent drowning! What is it about ‘-b’ that is causing distress? Is it that Yellow is described by -b and you are thinking of ‘b’ as meaning ‘blue’? Which is not really the case – it’s merely a coincidence of a how this geometric space was set up, labeling the axis’ as ‘ab’ rather than, say, ‘xy’.

      Does that help? Keep treading water if not!



Leave a Comment