11 Replies Latest reply on Oct 5, 2012 8:31 PM by b2martin_a

    DNG Profile Editor 1.0.4

    b2martin_a Level 2

      Upgrade for the DNG Profile Editor is availalbe on the Adobe site.  It works with v4 of the D700 profiles - I am checking it out now. 

        • 2. Re: DNG Profile Editor 1.0.4
          Yammer Level 4

          Help!

           

          The link takes me to the Labs website, which then sends me back to the page I started from.

          • 3. Re: DNG Profile Editor 1.0.4
            Yammer Level 4

            I found it using Google and adding 1 to the Mac ID for Windows:

            http://www.adobe.com/support/downloads/detail.jsp?ftpID=5494

             

            And it's also listed here:

            http://www.adobe.com/support/downloads/new.jsp

            • 4. Re: DNG Profile Editor 1.0.4
              Yammer Level 4

              A quick question for those knowledgeable in these things:

               

              Does one shoot the ColorChecker 24-patch chart 6500K image in full sunlight, part-cloudy, overcast, or artificial lighting?

               

              Or does one shoot in all these conditions to produce different profiles for different conditions?

               

              I ask because I always seem to get quite different profiles depending on the natural light conditions.

               

              I have just used the new DNGPE with a previously-used chart image and got a significantly different result, with darker blues and several more saturated colours. Has anyone noticed the new version giving different results to the old?

              • 5. Re: DNG Profile Editor 1.0.4
                ssprengel Adobe Community Professional & MVP

                A DNG Profile has two color tables, so:

                 

                You either shoot two images, one 2850K and one 6500K, and create a dual-illuminant chart used for a wide-range of light-sources where the actual profile used is interpolated or extrapolated from the dual-illuminant profile numbers depending on the WB of the photo it is applied to, OR you shoot a single image under a particular lighting condition and make a profile with that image used for both color tables and use it only for that light source or ones close to it.  The documentation PDF describes all of this.

                 

                What I usually do is have a general purpose dual-illuminant chart created from one tungsten shot and one hazy-sunny shot (choosing one from several different that is closes to a WB of 6500K) and then a few others, where either the tint is significantly off like fluorescent lights or the lighting is outside of the range of 2850K and 6500K, such as sodium vapor or twilight or night shots that are very orange or very blue.  I also make ones for different studio setups or important shoots just to be sure I have a choice if the main ones I use aren’t quite what I want.  None of these user-created profiles have the hue-twists that the Adobe-generated ones do, so sometimes, depending on the mix of colors, I use Adobe’s, too, either Adobe Standard or the Camera-match ones.

                • 6. Re: DNG Profile Editor 1.0.4
                  Yammer Level 4

                  Yes, I am aware of the theory outlined in the documentation and elsewhere, but in the absence of controlled 2850K and 6500K lighting setups, I usually go for the natural single-illuminant variety. Most of my photography is in outdoor daylight anyway.

                   

                  I don't know what sort of lighting conditions would be classed as equivalent to 6500K, but your suggestion of hazy sunshine certainly helps. I don't know if this has any precise relationship to the CR colour temperature slider. Most of my sunny day shots usually come out at 5200-5800K in CR. Anything over 6000K is usually very late in the day, heading towards twilight.

                   

                  My best homemade profile so far was made from a chart shot in overcast weather. Strangley though, the same image generates a much more saturated profile result with the latest version of DNG Profile Editor, which I wasn't expecting.

                  • 7. Re: DNG Profile Editor 1.0.4
                    ssprengel Adobe Community Professional & MVP

                    You get a 6500K image either by shooting at a D65 workstation (what Adobe does, I’m sure) or by trial-and-error outdoors, shooting the CC in various degrees of clouds in front of the sun until one has an eye-droppered WB close to 6500K and a Tint in the low positive numbers.   Depending on the weather this can take days or weeks to have the right mix.  You obviously want to keep reflected colorations to a minimum too like no trees or red-brick buildings or non-neutral clothing nearby when shooting the CC.  Northlight is usually bluer than pointing more toward the sun so your experiments can contain shots with the CC angled in several different compass directions.

                     

                    Where I live, full sun seems to be 4500-5500K and to get a higher Kelvin number you need more blue, which is where the degree of cloudiness comes into the equation, blocking the direct-sun and allowing more scattered blue into the mix.

                     

                    The X-Rite color-checker comes with software that also make dual-illuminant profiles and it doesn’t appear to have the restriction of needing to match the 2850K and 6500K, which makes things seem easier.  It all depends on whether the DNG PE creates profiles that label the color tables with the actual WB of the two images or if they contain hardcoded numbers of 2850K and 6500K, and if the usage of the profiles assumes hardcoded numbers or uses whatever the WB of the two images was.  I don’t know enough about how to read camera profile data to confirm one way or the other.  I also don’t know if X-rite has produced software that works with the newest DNG specification or not.

                     

                    It would seem reasonable to assume the actual WB of the two images is used in a dual-illuminant profile and you could make dual-illuminant profiles for mixed lighting in an indoor arena where there are different types of halogen or glowing-gas-type lighting based on the age of the bulbs, in which case getting a shot that is exactly 6500K is less important than just something in daylight. 

                     

                    The way to know for sure would be to make profiles in two different ways then use a program like ImaTest to compute the color error of CC shots in various lighting situations and see which have lower error.

                    • 8. Re: DNG Profile Editor 1.0.4
                      deejjjaaaa Level 2

                      > A quick question

                       

                      just read that thread noting comments from Eric Chan :

                       

                      http://www.luminous-landscape.com/forum/index.php?topic=43733.msg366772#msg366772

                      • 9. Re: DNG Profile Editor 1.0.4
                        Yammer Level 4

                        ssprengel wrote:

                         

                        You get a 6500K image either by shooting at a D65 workstation (what Adobe does, I’m sure) or by trial-and-error outdoors, shooting the CC in various degrees of clouds in front of the sun until one has an eye-droppered WB close to 6500K and a Tint in the low positive numbers.   Depending on the weather this can take days or weeks to have the right mix.  You obviously want to keep reflected colorations to a minimum too like no trees or red-brick buildings or non-neutral clothing nearby when shooting the CC.  Northlight is usually bluer than pointing more toward the sun so your experiments can contain shots with the CC angled in several different compass directions.

                        Thanks. That helps a bit, although I have since read that a 6500K reading can be obtained by varying spectrums of light. It makes sense, since calibrating a monitor can produce the same temperature reading with different combinations of R, G, and B.

                         

                        The thing which makes me laugh is the fact that lots of people say it doesn't matter in what daylight conditions you shoot the chart, when every time I have shot the chart I get a different result. Clearly it does matter. Hell, apparently it even matters which version of the DNG Profile Editor you use to make the profile.

                        • 10. Re: DNG Profile Editor 1.0.4
                          Yammer Level 4

                          deejjjaaaa wrote:

                           

                          just read that thread noting comments from Eric Chan :

                           

                          http://www.luminous-landscape.com/forum/index.php?topic=43733.msg36677 2#msg366772

                          Excellent reading. Thanks!

                           

                          I'm still only on page 2, but already I've learned a lot from this. It seems that an overcast day is not ideal, even though my best general profile was obtained from this situation. Lots of people seem to recommend midday direct sunlight, at 45 degrees, facing blue sky and an absence of buildings and foliage, shot in neutral clothing. I guess the blue sky will push the colour temperature higher and the high direct sunlight will provide the broadest spectrum.

                           

                          I had created a profile from conditions like this in the past, but was unhappy with the result. I had to boost saturation quite a lot, and the greens/blues never seemed correctly balanced. Maybe I should revisit the old charts with the new Profile Editor, seeing as it seems to produce much more saturated results with my overcast chart.

                          • 11. Re: DNG Profile Editor 1.0.4
                            b2martin_a Level 2

                            The Red/Greem/Blue Primary Saturation sliders under the Color Matrices tab do not work correctly for a D700 v4 profile.  I wanted to reduce the saturation by 5 units, so I set each of the Primary Satruation slicers to -5.  The Red Primary Saturation slider appeared to work, but when I changed the Green and/or Blue the Hue changed. 

                             

                            Please fix this.