6 Replies Latest reply on Jun 13, 2009 1:23 PM by Chris Cox

    Adobe Swatch Exchange (.ase) files: which illuminant/profile?

    Jason Thompson

      Since Adobe Swatch Exchange files appear to contain no data about illuminants or color profiles, I'm curious to know which ones they use by default when representing each of the various color models.

       

      Color models that appear to be supported for individual color swatches appear to be:

      • L*a*b*
      • CMYK
      • Gray
      • RGB

       

      Specifically, I'm interested to know the answers to the following four questions:

      1. Which illuminant is presumed for color swatches defined in L*a*b* coordinates?
      2. Which color profile is presumed for color swatches defined by CMYK ink proprotions?
      3. Which color profile is presumed for color swatches defined with RGB components?
      4. Which profile is presumed for color swatches defined as a gray percentage?

       

      My thanks go out to anyone who can help answer these questions!

        • 1. Re: Adobe Swatch Exchange (.ase) files: which illuminant/profile?
          Lou Dina Level 3

          JT,

           

          To the best of my knowledge, D50/2 observer is used for all of the color modes in Photoshop.  I just went through an exercise with Wendy Frost on this forum and measured a specific Pantone color (defined as Lab in the Photoshop Solid library).  I had four swatch books with this color in it, and they all were very close if I used the D50/2 observer with my spectrophotometer. She wanted me to report readings to her usind D65/10 observer, and those numbers are different from the D50/2 numbers.  The D50/2 matched the Pantone Lab definitions in PS, within tolerances for variance of my printed charts.

           

          Yes, some color spaces, such as sRGB and Adobe RGB have an internal white point of 6500K, but I am reasonably sure everything is brought back to D50 as the baseline.

           

          If you want a lot of information on color, formulas, conversion tables, etc, visit www.BruceLindbloom.com.

           

          Lou

          • 2. Re: Adobe Swatch Exchange (.ase) files: which illuminant/profile?
            Jason Thompson Level 1

            Thanks for your quick response, Lou! I'm aware that most Adobe applications presume the D50 illuminant/2 degree observer when interpreting L*a*b* coordinates. (This also happens to be the connecting color space for many ICC profiles.) I feel this is a reasonably safe presumption to use when working with .ase swatches saved with L*a*b* coordinates.

             

            However, given an RGB model swatch color in an .ase file, how do Adobe applications interpret the RGB values of that swatch?

             

            For example, the swatch may specifiy 0.5, 0.5, 0.5 as its floating-point RGB values. The color RGB floating point color specified as 0.5, 0.5, 0.5 may appear very different depending on which color profile is presumed.

            • 3. Re: Adobe Swatch Exchange (.ase) files: which illuminant/profile?
              Lou Dina Level 3

              Jason,

               

              I did a little experiment.  I created a swatch in PhotoshopCS2, with my color space set to sRGB.  The color was defined as 200R, 50G, 50B.  I made sure to start with a small color space and selected moderate numbers that would not be out of gamut in a decent CMYK space (my CMYK space was set to GRACoL#1 Sheet, downloaded from the GRACoL Website).  Here are the equivalents:

               

              sRGB   200R, 50G, 50B

              Lab      46L, 59a, 38b

              CMYK  9C, 92M, 80Y, 9K

               

              Still in PS, I changed the default color space to Adobe RGB.  Here are the new numbers:

               

              sRGB   200R, 50G, 50B

              Lab      52L, 69a, 48b

              CMYK  1C, 93M, 84Y, 0K

               

              I then exported the swatch as ASE and opened it in InDesignCS2, also set to Adobe RGB.  After importing the swatch, I double clicked on it, and it comes up as 200R, 50G, 50B.  If I then change the default color space in IDCS2 to sRGB, and double click the color, it still comes up as 200R, 50G, 50B.

               

              So, from what I can tell, there isn't any observer operating, nor any concern about consistent color appearance.  They're just numbers, and in my experiment, it looks like ASE remembers those RAW numbers, along with the color mode in which the swatch was originally created (RGB, LAB, or CMYK).  It ignores the color space (sRGB, Adobe RGB, US Web Coated SWOP, etc).  This seems confirmed, since changing the color space in PS or ID changes the appearance of the swatches in the palette, but not the original numbers.  This is analagous to assigning a profile.  I'm not sure if this is the same in CS3 or CS4.

               

              D50/2 is the correct white point and observer for conversions in and out of Lab.

               

              Not sure if this answers your question, or whether there is more going on under the hood.  From this very limited experiment, it doesn't seem so.  I have no other information on this subject, but would love to know if you find something conclusive.

               

              Lou

              • 4. Re: Adobe Swatch Exchange (.ase) files: which illuminant/profile?
                Jason Thompson Level 1

                I've been doing my own experiments with CS3 applications and have come to a different conclusion for swatches defined using the RGB and Grayscale models.

                 

                So far, my experiments with color swatches defined in the Grayscale color model are that they are consistently saved and interpreted using the sGray profile. (This is a grayscale profile that shares the same gamma function as the sRGB IEC61966-2.1 RGB profile.)

                 

                In color swatches defined in the RGB color model, they are being consistently saved and interpreted using the sRGB IEC61966-2.1 profile.

                 

                Regardless of my document and working profiles (and regardless of if they are the same) when I create the swatches, I can confirm by decoding the binary color data in the .ase file that these two profiles are obviously being used for these color models, even though the profiles are never named in the .ase files.

                 

                When I use these swatches in a document having a different profile, using the swatch colors creates document colors that are correctly converted from sRGB/sGray obeying my color management settings.

                 

                This is actually a very good thing. Just because the swatch colors are encoded using sRGB doesn't limit the swatches' gamut to the sRGB gamut. Since floating point numbers are used in the .ase files, it is possible to encode out-of-gamut sRGB colors of less than zero and greater than one. When using the swatches in a document with a color space/profile having a different gamut, they can potentially be converted to in-gamut colors, depending on one's color management and conversion settings.

                 

                To sum up what I have learned so far:

                • Swatches defined using the Grayscale model are encoded as big-endian 32-bit floating point sGray values with a nominal gamut of {0.0..1.0}.
                • Swatches defined using the RGB model are encoded as big-endian 32-bit floating point sRGB values with a nominal gamut of {0.0..1.0, 0.0..1.0, 0.0..1.0}.
                • Values that would be nominally out-of-gamut are allowed.

                 

                I'll do some further research to see how CS3 handles CMYK and L*a*b* values in .ase files.

                 

                Edit: The behavior I am observing is probably not really any different from yours. My color management settings in CS3 are set to always preserve embedded profiles. This has the side effect of preserving the appearance of .ase swatches when they are used, rather than preserving their absolute values when they are used, as would be analgous to the behavior you describe in CS2.

                 

                The gray/RGB swatches are definitely being saved as sGray/sRGB though. How your applications handle these swatches is just dependent on your color management settings.

                • 5. Re: Adobe Swatch Exchange (.ase) files: which illuminant/profile?
                  Lou Dina Level 3

                  Jason,

                   

                  Very interesting.  That is more like I would have expected.  My test in CS2 suggests RAW numbers in the originating color mode are used.

                   

                  I wonder if Lab gets into the act during conversions?

                   

                  Lou

                  • 6. Re: Adobe Swatch Exchange (.ase) files: which illuminant/profile?
                    Level 7

                    LAB - D50 2degree standard observer, same specs as the ICC version2 PCS

                     

                    The rest are uncalibrated, no profiles associated.  The numbers are loaded by the applications and then interpreted on a per-document basis.

                    (sorry, best I could do with all the apps involved - even getting them to agree to a decent exchangeable format was fun)