10 Replies Latest reply on Oct 5, 2010 5:35 AM by Lou Dina

    Is the Kodak Gray card "spectrally neutral"?

    Lou Dina Level 3

      Does anyone know if the Kodak R-27 18% gray card is spectrally neutral? How about when lit with different illuminants? Would it be appropriate for setting WB in ACR?

       

      I'm more interested in actual measurements or well-documented articles than opinions.

       

      Thanks, Lou

        • 1. Re: Is the Kodak Gray card "spectrally neutral"?
          Jeff Schewe Level 5

          No, the Kodak 18% graycard is not really designed for color. It's designed for producing an 18% reflectance used for metering and exposure not really color. Besides, it makes for a poor target for white balancing because the tone is too dark. For white balance, you want a lighter, non-spectral white. The second from the brightest patch of the ColorChecker is what Thomas Knoll used to develop white balance tool in Camera Raw. So, I would suggest that would be a much better target than a graycard.

          • 2. Re: Is the Kodak Gray card "spectrally neutral"?
            Lou Dina Level 3

            Thanks, Jeff. Your understanding mirrors my own.

             

            I am having a discussion with another photographer who claims that the Kodak 18% gray card is spectrally neutral and is a suitable WB tool. Since I don't own one (I did 20 years ago), I cannot confirm or deny his assertion. It did get me to thinking it was something I should know, so I am hoping to find definitive information (measurements, Kodak white papers, specifications, in-depth articles) either confirming or refuting that it is spectrally neutral.

             

            Kodak has produced various "gray" targets over the years, so perhaps some are spectrally neutral and others are not. Twenty years ago, I used a Kodak 18% gray card for exposure purposes, but that was with film. I didn't worry about gray balance at the time, since that was left up to film choice and the processing lab. My gray card has long since vanished into the ozone, so I do not have it available to measure with my spectro.

             

            BTW, I am about to build some profiles with DNG Profile Editor. I took a photo of my CC target in sunlight and under some Solux lamps. Looking forward to the results.

             

            Lou

            • 3. Re: Is the Kodak Gray card "spectrally neutral"?
              Jeff Schewe Level 5

              Lou Dina wrote:

               

              I am having a discussion with another photographer who claims that the Kodak 18% gray card is spectrally neutral and is a suitable WB tool.

               

               

              He's wrong...simple as that. An 18% graycard is too dark to be used optimally for white balance. There's a reason why it's called "white balance". You want to have the balance determine much further up the tone scale than a middle gray. The second to the lightest (or even the lightest if the exposure is down a bit) patch is what was used to develop the Camera Raw White Balance tool. It's foolish to ignore that...

              • 4. Re: Is the Kodak Gray card "spectrally neutral"?
                Noel Carboni Level 7

                VERY interesting!  I know a lot of folks who treat that card as THE indisputable reference for white balance.

                 

                I've always just pointed my cameras at someone's white shirt or some other white object, myself, and (as long as nothing's overexposed) they've always come up with dandy results.

                 

                -Noel

                • 5. Re: Is the Kodak Gray card "spectrally neutral"?
                  RASouthworth Level 3

                  To answer Lou's original question, the Kodak gray card is actually pretty good spectrum wise - did a quick check with my I1 spectro, results attached, the bluest value is 380nm and the increment is 10nm.  In tristimulus lab terms, L=49, a=-.8, b=.7; fairly close to neutral, and I only took one measurement.  The card does have significant directionality, i.e. it's not all that diffuse.  The spectro is a 45 degree type, so it pretty much only measures the diffuse component.

                   

                  It's main disadvantage for white balance is it's too far down in luminosity, and therefore errors get multiplied.  Single point white balance, either in a camera or in external software, is a simple minded operation, usually one channel is left alone (normally green) and the other two have their slope changed so that the target object will measure R=G=B in the selected colorspace.  It's better to be up in the high luminosity region rather than down in the darker grays, to ensure the lighter neutrals are more likely to appear so.  A slight error in the adjusted channels at lower luminosity is magnified for bright neutrals.

                   

                  Richard Southworth

                   

                  KodakGrayCardSpectrum.jpg

                  • 6. Re: Is the Kodak Gray card "spectrally neutral"?
                    Lou Dina Level 3

                    Hi Richard.

                     

                    It's been awhile. Thanks for taking some readings and posting the information. My Kodak gray card vanished years ago. The neutrality you measured is reasonably close to the measurements I get of the gray patches on my 24-patch color checker. I agree on using a lighter luminance to establish WB.

                     

                    Hope all is well.

                     

                    Lou

                    • 7. Re: Is the Kodak Gray card "spectrally neutral"?
                      Bill_Janes Level 2

                      Jeff Schewe wrote:

                      He's wrong...simple as that. An 18% graycard is too dark to be used optimally for white balance. There's a reason why it's called "white balance". You want to have the balance determine much further up the tone scale than a middle gray. The second to the lightest (or even the lightest if the exposure is down a bit) patch is what was used to develop the Camera Raw White Balance tool. It's foolish to ignore that...

                      You shouldn't use the N1 patch (brighest patch) of the ColorChecker, since it is not spectrally neutral. It is difficult to produce paint that is spectrally neutral at that reflection density. See Bruce Lindbloom's Color Checker Calculator for typical values. The remaining patches are pretty neutral. N2 is best for white balance, since the camera will collect more photons and the signal:noise will be better. If the camera has good noise characteristics, a good white balance at base ISO can be obtained from the N4 square, which is midgray (L* = 51). However, at high ISO or with cameras with worse S:N the WB may not be optimal.

                       

                      For example, with the Nikon D3 at base ISO (200), here is the white balance from N4 (patch 22). The RGB values for the neutral patches are shown in the ACR readout at the top.

                       

                      02_WB_N4.png

                       

                      Here are the results for N2, patch 20:

                       

                      02_WB_N2.png

                       

                      N1 (patch 19):

                       

                      02_WB_N1.png

                       

                       

                       

                      And N6 (patch 24):

                       

                       

                      02_WB_N1.png

                       

                      With the exception of WB with N1 (which is not specrtally neutral), the RBG values are very close at all levels when WB is performed with the other patches.

                       

                      I don't have a spectrophotometer to read my Kodak gray card, but gray balancing on this card produces suboptimal results as shown here:

                       

                      02_WB_Kodak.png

                      • 8. Re: Is the Kodak Gray card "spectrally neutral"?
                        Lou Dina Level 3

                        Bill,

                         

                        Thanks for the post and the images.

                         

                        It appears there is some inconsistency between some of Kodak's gray cards. Maybe they issued different products, or perhaps it is a quality control issue (or age/wear of the card). Richard Southworth's card reads very neutral, but yours does not. I have seen cases of both before, but this inconsistency alone is enough to keep me using the Color Checker, which has shown great consistency over the years.

                         

                        Lou

                        • 9. Re: Is the Kodak Gray card "spectrally neutral"?
                          Bill_Janes Level 2

                          Lou Dina wrote:

                           

                          Thanks for the post and the images.

                           

                          Your'e welcome. Unfortunately, the images are now worthless as they have been resized such that the numbers are no longer readable in the magnified view. I have not seen this before. To prevent this in the future, one should use a link to the image rather than imbed the images.The new forum software sucks.

                           

                          It appears there is some inconsistency between some of Kodak's gray cards. Maybe they issued different products, or perhaps it is a quality control issue (or age/wear of the card). Richard Southworth's card reads very. neutral, but yours does not. I have seen cases of both before, but this inconsistency alone is enough to keep me using the Color Checker, which has shown great consistency over the years.

                          My gray cards are quite old, dating from the film era. However, they have been stored in their envelope and appear to be in good shape. As Jeff Schewe commented, they were originally used for exposure determination. Perhaps Kodak has given more attention to spectral neutrality in the digital era. For field use, I much prefer a durable plastic card designed for white balance (I use the WhiBal). The new Passport is another option, but I see little need to calibrate in the field, and white balance is usually sufficient for my needs.

                          • 10. Re: Is the Kodak Gray card "spectrally neutral"?
                            Lou Dina Level 3

                            Bill,

                             

                            Just left-clicking on your embedded images enlarged them and made them easily readible, so your post served its purpose fine.

                             

                            I'm with you on calibration. I see no need to recalibrate for each lighting condition (I might consider it for a big, important photo session under constant lighting conditions). I will say, however, that creating a dual-illuminant custom profile with DNG Profile Editor and CC target for my 5DmkII has made a HUGE difference. That's now my "standard profile". Maybe it's just my particular camera, but I am seeing a vast improvement in my images by using a custom profile. My colors snap into place very quickly, and I have to do a lot less jiggering around with the various Camera RAW sliders. I still set the WB with the WB tool and/or color/tint sliders, but I get to where I want to go quickly now.

                             

                            I'm also finding, in general, that the closer I set the contrast slider to zero, the more natural my skin tones appear. As I increase contrast, the colors get a bit pumped up and start looking less natural. I'd love to see an option to control luminosity and color independently in Camera RAW.

                             

                            Thanks again,

                             

                            Lou