13 Replies Latest reply on May 12, 2010 4:58 PM by Larry Tseng

    Why are Gamut Warnings Inconsistent?

    Dave505

      The color picker gamut warning and the proof setup gamut warning are inconsistant and when I change the rendering intent in the proof setup the gamut warning area changes.

       

      I thought maybe the gamut warning shows where the color is modified instead of which colors are out of gamut, but that does not explain what I see.

      When I use an image with just one color and the absolute colormetric rendering intent in the proof setup, I get a gamut warning on the image.  If I look at the color in the color picker there is NO gamut warning.  If I use the relative colormetric rendering intent in the proof setup I get NO gamut warning.

       

      I'm using PS CS3, on XP SP3, with the Adobe Color Engine.

       

      Why are these Gamut Warnings Inconsistent?  Are these inconsitencies just due working with colors near a poorly defined gamut boundary?  I have seen major differences in gamut warning with different proof set-up rendering intents.

       

      Thanks.

        • 1. Re: Why are Gamut Warnings Inconsistent?
          Lou Dina Level 3

          David,

           

          I have often wondered the same. I rarely use gamut warning, and rely on custom proof preview and sample points for soft proofing with a properly calibrated monitor. I'm not sure how accurate gamut warning is, and whether it shows just those colors that are out of gamut, or all colors that change.I suspect it is only the colors that are out of gamut, since changing from perceptual to relative seems to make zero difference to the gamut warning plot on screen. (I'm on a Mac and don't see any difference in gamut warning when changing rendering intents for a given CMYK space).

           

          Where I do occasionally use gamut warning is when trying to adjust colors before conversion to a CMYK space from RGB. For example, a bright red, saturated flower that will be partially, or completely, out of gamut in CMYK. With gamut warning turned on, and my intended CMYK destination space set up in soft proof mode, I can quickly get an indication as to which colors will present a problem, since they are grayed out. This may help to select sample points in a few critical areas of the flower, where I want to insure I have different tones and detail. Once my sample points are selected, I turn gamut warning back off (or sometimes turn it off before sampling). Then I make adjustments with curves, or whatever, (in RGB) to maintain some detail and differentiation in the reds, so I don't end up with a detail-less blob of red color when I do the conversion. When faced with this situation, I sometimes reduce saturation to keep the entire image within CMYK gamut, and to insure detail does remain, then do the conversion to CMYK. Once in CMYK, I'll check my sample points and make curves and saturation adjustments so some of the reds are totally maxed out, while others are not. This gives me the reddest, most saturated red I can get in the destination CMYK space, but gives me some detail too. That's the extent of my use of gamut warning.

           

          For me, the moral of the story is to remain a bit conservative when editing in RGB if I know my final destination is CMYK or a fine art print on a matte fine art paper with limited gamut.

           

          Maybe I'm missing some other good uses for gamut warning, but I don't know what they might be.

           

          Lou

          • 2. Re: Why are Gamut Warnings Inconsistent?
            Printer_Rick Level 4

            You are seeing the fundamental difference between relative and absolute

             

            Relative rendering compensates for media white. BPC compensates for the black point and helps to preserve shadow detail.

             

            Absolute does not compensate for either white or black points.

             

            Assume you are in RGB. Fill half a blank canvas with 0R 0G 0B. Now you have white and black.

             

            If you go to Proof Setup with Relative, select the destination CMYK, gamut warning does not show the white and black as out of gamut.

             

            In a normal file conversion context, the white and black are indeed in gamut. No information is lost. With relative and BPC, you could happily bounce mode CMYK, mode RGB, Mode CMYK endlessly, all day long, and the white stays white and the black stays black. But in reality, the paper is not as bright as 255R 255G 255B. And the darkest color (75C 68M 67Y 90K in US Web Coated v2) is not 0R 0G 0B.

             

            If you change the Proof Setup to Absolute, it gives you a better idea of just how many colors are out of the destination gamut. The problem of course is that Absolute should never be used for a file conversion.

             

            As for the color picker I believe that is based on the rendering in your Color Settings.

             

            The best proof setup for most (not all) RGB images going to CMYK is:

             

            Relative with BPC

            Simulate paper white on

             

            The simulate paper uses Absolute rendering to create the soft proof, but the file conversion to CMYK is still relative. It's the best way to see on your display a more accurate representation of paper and ink.

             

            With this proof setup, you will see most all the RGB colors shift when you turn on Proof View. However gamut warning will only flag those areas that are clipped in the file conversion, not in the final printed result.

            1 person found this helpful
            • 3. Re: Why are Gamut Warnings Inconsistent?
              Lou Dina Level 3

              Thanks, Rick. That was interesting. I so rarely mess with Absolute Colorimetric, I didn't even check it out with gamut warning turned on.

              • 4. Re: Why are Gamut Warnings Inconsistent?
                Dave505 Level 1

                Thanks Rick, you provided 2 very helpful peices of information:

                1. The color settings rendering intent is used for the color picker gamut warning.
                2. Absolute Colormetric does not use black point compensation.

                 

                My gamut warnings are more consistant if I use this info.  I discovered that the proof gamut warning does not use the black point compensation setting in the proof setup...The proof gamut warning uses the black point compensation setting in the color settings.

                 

                Even with this knowlege I am still seeing an inconsistancy in the gamut warning using the relative colormetric w/o BPC vs the absolute colormetric rendering intetents.  The absolute rendering intent in the proof setup generates significantly more gamut warnings than the relative colormetric w/o BPC.  Here's a link to a these two gamut warnings: http://www.naturesgrandeur.com/images/gamut_warnings.gif

                 

                I'm interested in the gamut warning because I want to: 1) know what colors a printer is capable of printing, and 2) try to use colors which are within the printer's gamut.  I am using a pantone color fan in addition to soft proofing.  I'm using the adobe RGB working space.

                 

                Thanks.

                • 5. Re: Why are Gamut Warnings Inconsistent?
                  Printer_Rick Level 4

                  It's neat but a little bit half baked. What would be better is to have the two conversions of the Proof Setup honored properly in the gamut warning. First, the file conversion (RGB – CMYK) using relative for example. Then the soft proof (CMYK – Monitor RGB) using Absolute (simulate paper)

                   

                  When you enable the paper option it is usually easy to see there's more getting clipped than gamut warning indicates.

                  • 6. Re: Why are Gamut Warnings Inconsistent?
                    Printer_Rick Level 4

                    The absolute colorimetric has no white point compensation or black point. That's the difference you're seeing, I think. With the relative no BPC, you still have the white point adjustment.

                     

                    It is the same type of thing happening in the display options in Proof Setup. The Simulate Black Ink checked is equivalent to Relative with BPC unchecked. The Simulate paper white is equivalent to Absolute. But the display options relate only to the soft proof conversion (CMYK – Monitor)

                     

                    Also an edit to my last post. There would be a big problem with Gamut Warning trying to honor this soft proof conversion (CMYK > Monitor). There are many CMYK colors that fall outside of monitor gamuts. These colors would end up being included in the warning, and it could very confusing and misleading. Just because the monitor can't proof the color doesn't mean it's not printable.

                     

                    The only way one could know about CMYK colors that can be printed, but not displayed: the image must already be CMYK. Then Proof Setup. Device to Simulate, Monitor Profile. Rendering Absolute. Gamut warning shows what CMYK colors fall outside the monitor space.

                     

                    As far as I know, it is impossible to know about these out-of-monitor gamut CMYK colors if the image is still source RGB.

                     

                    To summarize, what would be great is to have a source RGB image, and see ALL the source colors that are beyond the CMYK print gamut. The proper file conversion (default relative) needs to be a part of this. However, this does not yield a fully accurate print appearance.

                     

                    It would seem that Absolute could be used to take care of the appearance, but you can't go down that road because that's not the true file conversion. To understand, open any RGB. Proof Setup, CMYK. Absolute.

                     

                    The white point gets lighter when compared to relative. But what needs to happen is the paper white should appear duller. Absolute is dropping out highlights because of the dullness of the paper. It's reversed, because it's being used for a file conversion, instead of just a proofing conversion.

                     

                    Message was edited by: Printer_Rick added a few things and clarified confusing statements

                    • 7. Re: Why are Gamut Warnings Inconsistent?
                      Dave505 Level 1

                      Ok, Thanks Rick.  I think I understand now, what I need to do to get the proper gamut warning (what colors will be clipped, because the printer cannot print them).  I need to use the Relative Colormetric with BPC.  This is the rendering intent normally used for printing.  The Absolute Colormetric rendering intent does not scale to the paper white color and (for reasons I probably don't entirely understand), this causes more light colors to show as out of gamut.

                      • 8. Re: Why are Gamut Warnings Inconsistent?
                        Printer_Rick Level 4

                        Yes, that's correct. Relative is normally used for file conversions, especially for images.

                         

                        I apologize for my convoluted explanation of Absolute rendering. This may help:

                         

                        In any decent CMYK profile, the paper white Lab values are included. In a CMYK file, paper white should be 0C 0M 0Y 0K, and the relative Lab equivalent is 100L a0 b0.

                         

                        However, no paper is completely white. It has to absorb some light, if only a little. Some papers also have a slight color cast (affecting a and b values)

                         

                        The same principle applies to the black point. Now matter how much ink goes on the sheet, it can only get so dark. Some light will bounce off. And it may not be perfectly neutral either (but should be as close to neutral as possible).

                         

                        Try this. Take any CMYK image. Duplicate it twice so you have 3 copies.

                         

                        The first, convert to Lab, with Relative, BPC enabled.

                         

                        The second, convert to Lab, with Relative, BPC disabled.

                         

                        The third convert using Absolute.

                         

                        These 3 conversions represent what's going on in the Proof Setup display options. The first – relative with BPC –  is equivalent to no display options checked. The second is equivalent to Black Ink checked. The third – Absolute – is equivalent to Simulate Paper White.

                         

                        Now make a blank white CMYK, convert to Lab using Absolute. Look at the Lab numbers. If the CMYK was US Web SWOP v2, the Lab is 89L 0a 4b. A dingy paper, the temperature a little yellow. If you had a Grade 5 sheet (cheap coated like you see in the newspaper ad inserts) the Lab measurement would be close to this, and the appearance you see on a calibrated profiled monitor should be close to the blank paper.

                         

                        Think back to a source RGB image, before conversion to CMYK. The goal is this – you need to see what exists in the file but cannot exist on the press sheet. Unfortunately, once the image is actually printed and on the press sheet, the CMYK file is no longer relevant. All that matters is the color appearance, which can only be measured using Lab values.

                         

                        So from Source RGB capture to printed result, there are actually 3 conversions – first a file conversion to CMYK (4 separate inks), and then a non-file conversion to Lab (composite color you see on the paper). This final conversion is always going to be Absolute, and you can't change that. And as far as I know, the colors that get clipped as you move from source RGB to press sheet Lab is something Photoshop can't tell you. I wish there were a way.

                        • 9. Re: Why are Gamut Warnings Inconsistent?
                          Larry Tseng Level 1

                          Hi Rick,

                           

                          So from Source RGB capture to printed result, there are actually 3 conversions – first a file conversion to CMYK (4 separate inks), and then a non-file conversion to Lab (composite color you see on the paper). This final conversion is always going to be Absolute, and you can't change that. And as far as I know, the colors that get clipped as you move from source RGB to press sheet Lab is something Photoshop can't tell you. I wish there were a way.

                           


                          You're including monitor clipping when going from CMYK to press sheet Lab, right? Just trying to follow your line of thinking here.


                          --Larry

                          • 10. Re: Why are Gamut Warnings Inconsistent?
                            Printer_Rick Level 4

                            Actually I was thinking about the press sheet Lab, independent of the monitor. Lab is the only color space that can define the printed result.

                             

                            I think of it in a press check context, where the client has no knowledge of CMYK. This person just wants a certain appearance. On the printed sheet CMYK separation values aren't relevant anymore, and the only way to measure appearance is to use Lab.

                             

                            This is a difficult concept for old schoolers like me. I still think of CMYK numbers making colors, but in a color managed world it doesn't work that way. Yes 100C 50M is blue, but what blue depends on a lot of things. Press sheet Lab gives the blue true meaning.

                             

                            The problem with soft proofing is of course the monitor space, as you mentioned.  There are a few CMYK colors that stand to be clipped. It's an issue.

                             

                            I am uncertain how many designers use the simulate paper option when soft proofing for CMYK. I am also uncertain how much the color temperature of the monitor white point affects the soft proof accuracy.

                            • 11. Re: Why are Gamut Warnings Inconsistent?
                              Lou Dina Level 3

                              Hi Rick...good discussion. I appreciate your clear explanations and examples.

                               

                              The problem  with soft proofing is of course the monitor space, as you mentioned.   There are a few CMYK colors that stand to be clipped. It's an issue.

                               

                              Even for those designers using Adobe RGB for their source RGB files, CMYK colors that lie outside the monitor or ARGB gamut are indeed an issue. I'd wager that most monitors being used for soft proofing still come closer to an sRGB gamut than Adobe RGB, so this compounds the potential issue.

                               

                               

                               

                              In my experience, these CMYK colors that cannot be displayed on a monitor pale in comparison with the RGB colors that get clipped on a press sheet, especially, pure whites & blacks, blues and light pastels (especially light magenta flowers). My solution to this is to remain conservative when editing in RGB, leaving plenty of headroom. I also edit in soft proof mode, with my final CMYK space selected (usually RC or Perc, with BPC). This way, when I convert to CMYK, I am less likely to have a mess on my hands. I then edit the image in CMYK to make it sing the best I can and take advantage of whatever the press space can deliver.

                               

                               

                               

                               

                               

                               

                               

                              I am  uncertain how many designers use the simulate paper option when soft  proofing for CMYK. I am also uncertain how much the color temperature of  the monitor white point affects the soft proof accuracy.

                               

                              I can't speak for others, but I rarely use the simulate paper white option. But to be fair, nearly all my work goes to a sheetfed press using #3 gloss coated stock or better, usually not a dingy, yellow, uncoated, absorbent paper grade. I typically try the Black Ink option, though I usually find that the final press sheet ends up somewhere between having Black Ink checked and unchecked. Larry and I have had this discussion before, and it could be that my monitor just doesn't show it accurately, (or perhaps limitations of the algorithm, technology, differences between reflected CMYK and RGB light, etc). If I did a lot of printing on uncoated stock, newsprint, etc, I am sure I would use the Paper White option a lot more. I suspect, but cannot confirm, that most designers don't use paper white.

                               

                              In my own fine art inkjet printing, I tend to follow the same guidelines. When printing onto glossy, luster or photo type papers, I soft proof with both black ink and paper white unchecked, and get the best match, when viewing under controlled, 5000K lighting of the proper intensity. When printing on matte fine art papers, I find the final print usually falls between having black ink checked and unchecked. (I only use the paper white option for soft proofing if printing on a very warm paper).

                               

                              BTW, my monitor is calibrated for 5200K, 2.2 gamma, 85 cd/m2, and has a black luminance of about 0.25 cd/m2.

                               

                              Lou

                              • 12. Re: Why are Gamut Warnings Inconsistent?
                                Printer_Rick Level 4

                                Thanks for the information Lou. I'm certainly no expert on monitor calibration. My white point is daylight, I'm not sure about the cd/m2 values.

                                • 13. Re: Why are Gamut Warnings Inconsistent?
                                  Larry Tseng Level 1

                                  Rick & Lou,

                                   

                                  Thanks for sharing your experiences.

                                   

                                  About being able to check for monitor gamut clipping while the image is still source RGB: there is sort of a way to do that without (a) submitting a new feature request and waiting forever for it to not happen, or (b) figuring out how to create a non-standard ICC profile that has the combined characteristic of CMYK + monitor that Photoshop won't reject outright.

                                   

                                  The idea is to first convert the source RGB into a smart object, then convert the space that the smart object is in into CMYK. Inside the smart object, you'd check for CMYK gamut clipping as you adjust the RGB image for optimal conversion to CMYK. Outside the smart object, you'd check for monitor gamut clipping to find out what CMYK colors can't be displayed as you make final tweaks in CMYK to make the image "sing" (as Lou put it). Automate the soft proofing setup by assigning the action to different function keys.

                                   

                                  BTW, I just noticed that necdisplay.com has a special closeout sale for the wide gamut NEC LCD2180WG-LED-SV. This unit, and others like it, should reduce the monitor clipping issue significantly. (I happen to have the NEC and viewing the entire gamut of GRACoL2006_Coated1v2.icc is not a problem.)

                                   

                                  Cheers,

                                   

                                  Larry