11 Replies Latest reply on Jun 12, 2013 11:11 AM by ffredrikk

    Fujifilm X100S color rendition with ACR 8.1

    ffredrikk

      I just downloaded the Lightroom 5 beta (ACR 8.1) and here are my impressions/feedback on color rendition when working with RAW files from the Fujifilm X100S (firmware 1.03).

       

      There is an orange cast to RAW images (especially evident in skin tones). This can be somewhat fixed with the red “Hue” and “Saturation” sliders under “Camera calibration” (I always start out by setting Saturation to +25 in here), but I hope this is something that Adobe will adress in an upcoming Adobe Camera RAW update for this camera so that images will get better treatment of reds (again, especially in skin tones).

       

      Other RAW developers such as Apple Digital Camera RAW (Compatibility Update 4.05) or Iridient Developer is treating the reds similar to each other – and much more naturally than what Lightroom 5 (with ACR 8.1) is doing. When you perform a comparison of the same RAW in these different software packages, you can see that it looks like Lightroom has de-saturated the skin tones which makes them loose the fine reds and richness and are left with a de-saturated, flat and orange version of what the non-Adobe packages offer.

       

      I know that this is a widely discussed matter throughout Internet forums and I know a lot of people will be very happy, including myself of course, if this could get addressed by the excellent ACR team.

       

      Question to the Adobe Camera RAW team members: Do you think this is something you will be able to look into?

      Please let me know if you would like me to give examples.

       

      Regards,

      Fredrik

        • 1. Re: Fujifilm X100S color rendition with ACR 8.1
          JimHess Adobe Community Professional & MVP

          In the meantime, one thing you could do is adjust the saturation and any other adjustments that you think are required consistently and then save new camera defaults. Then those adjustments would be applied whenever you import images into Lightroom.  Or, you could create a preset that makes the necessary adjustments and apply that preset during import. Either way works very well.

           

          Everyone here is just a user you like yourself. Occasionally someone from Adobe will comment about upcoming changes. But most of the time they are constrained by a nondisclosure agreement that prevents them from being able to discuss things like your question. It's probably good that you mentioned the saturation problem. There is a possibility that if they discover that this is a consistent anomaly there could be updated profiles sometime in the future. But the only way to find out is to simply wait and see.

          • 2. Re: Fujifilm X100S color rendition with ACR 8.1
            ffredrikk Level 1

            Hi Jim,

             

            If this forum is not the way to get in touch with Adobe, but instead with its customers, how would you recommend I proceed to get this information to Adobe?

             

            Regards,

            Fredrik

            • 3. Re: Fujifilm X100S color rendition with ACR 8.1
              ffredrikk Level 1

              Here are a couple of samples images to show what I mean... First the Adobe Camera RAW conversion followed by the Apple Digital Camera RAW conversion. No settings or sliders have been touched, of course. The Adobe RAW conversion produces pale skin tones with a somewhat orange cast, probably because the lack of reds. Apple's RAW conversion produces far much deeper and complex color rendition in especially skin tones.

               

              DSCF1961.jpgDSCF1961-Apple.jpg

               

              DSCF2137.jpgDSCF2137-Apple.jpg

               

              DSCF3971.jpgDSCF3971-Apple.jpg

               

              Unfortunately, it is not possible to create one single Lightroom preset which will work for all different scenarios.

               

              All RAW files and Apple-exported TIF files are available here for inspection (~450 MB) as it is very hard to see these details from the thumbnails posted above:

              https://mega.co.nz/#!KtYSjCRD!KnHU4V8XT7zXE-HjtAXUW1W1cItZTqEA_vSp0sKZ-2g

               

               

              Regards,

              Fredrik

              • 4. Re: Fujifilm X100S color rendition with ACR 8.1
                b2martin_a Level 2

                This is not a problem with the RAW converter (Lightroom 5 Beta or ACR 8.1).  Adobe only has one camera profile for this camera (Adobe Standard) and it gives different results than the camera profile used by Apple software.  Adobe supplies camera matching profiles for Nikon, Canon, and a few other cameras, but not for this camera.  You can modify the Adobe Standard camera profile using the free DNG Profile Editor software, but it will be very difficult to generate a camera profile that matches what you get using the Apple software. 

                 

                I use a Nikon camera and Adobe Standard, Camera Standard, Camera Neutral, Camera Portrait, Camera Landscape, Camera Vivid, Camera D2X Mode 1, Camera D2X Mode 2, and Camera D2X Mode 3 camera profiles are available for RAW conversion.  The camera profiles with names of Camera xxxxx are profiles that match the Nikon profiles. 

                • 5. Re: Fujifilm X100S color rendition with ACR 8.1
                  RASouthworth Level 3

                  I believe the primary difference between the images is white balance - raw converters may well differ in their wb interpretations, IMO the proper way to compare images from various conversions is after manual white balance has been applied.

                   

                  All three of your example images appear to me to be difficult lighting situations, easy for either a camera or raw converter to not arrive at the most pleasing rendering.  The first image shows the typical "blue sky" shadows tint, easily corrected by using the ACR wb dropper on a neutral surface also in the shade, as demonstrated by my wb only adjustment of the jpeg image in ACR ( I did not download the raw images).  The Apple wb algorithm may arrive at images more pleasing to your eyes, does not necessarily imply an ACR defect.

                   

                  Richard Southworth

                   

                  image wb.jpg

                  • 6. Re: Fujifilm X100S color rendition with ACR 8.1
                    Vit Novak Level 3

                    Yes, there can be slight differences in WB interpretation among different programs if manual WB is used (setting the temperature and tint). To get the same WB for comparison, as shot WB should be used. Different programs will show different reading in K, but whitebalancing should be the same - as measured by the camera

                     

                    However, first pair of photos has very similar WB. Can't say for other photos, as there is no near-white color to compare

                     

                    Adobe standard profiles have generally less saturated skin/brown colors than other programs (one of the differences). From my findings, these differences are mostly due to applying of tone curve in photo pro color space in ACR workflow, while many other programs or cameras are applying it in sRGB or Adobe Rgb, which, in my opinion, produces better results.

                    • 7. Re: Fujifilm X100S color rendition with ACR 8.1
                      Level 5

                      Vit Novak wrote:

                       

                       

                      Adobe standard profiles have generally less saturated skin/brown colors than other programs (one of the differences). From my findings, these differences are mostly due to applying of tone curve in photo pro color space in ACR workflow, while many other programs or cameras are applying it in sRGB or Adobe Rgb, which, in my opinion, produces better results.

                       

                      A very enlightening comment, Vit!  Thank you.

                       

                      I often find myself lowering the saturation of skin colors generated by many digital cameras.

                      • 8. Re: Fujifilm X100S color rendition with ACR 8.1
                        RASouthworth Level 3

                        Vit,

                         

                        In my experience wb will vary from converter to converter, even if all are set to reflect the camera settings.  I "see" a difference in wb and saturation in the first pair of images, I still hold that a valid comparison will require some sort of manual wb adjustment on a common image area.

                         

                        Richard Southworth

                         

                        Added by edit - the wb shift is most apparent in the last pair of images, looking at the ceiling above the man's head.

                        • 9. Re: Fujifilm X100S color rendition with ACR 8.1
                          RASouthworth Level 3

                          To carry my argument further, I downloaded the last pair of images (jpegs), brought them into ACR, wb'd with the dropper on the top of the open door to the right of the man's neck, and then saved them out.  Now I believe it is valid to compare saturation.

                           

                          Richard Southworth

                           

                          image_jpeg ACR_wb.jpg

                           

                          image_jpeg CPP_wb.jpg

                          • 10. Re: Fujifilm X100S color rendition with ACR 8.1
                            Vit Novak Level 3

                            Yes, actually it's possible to get different WB even with as shot setting. There are generally 2 whitebalancing methods, as described in dng specification and also mentioned by Eric a while ago on this forum. ACR supports both, depending on the profile (whether forward matrices are defined or not - from my remembering - however all profiles have those matrices defined, so the second one is always used). If some program or camera is using the first method, result won't be exactly the same. However, if the second method is used (with RGB color multipliers that are recorded in raw metadata), white balance should be the same, unless, of course, something else went wrong ...

                             

                            If manual WB is used, there can be differences among programs in algorithms used to convert color temperature/tint into RGB multipliers. It also depends on the dng profile, so with slightly different color matrices in the profile for the same camera, white balance will be different for the same manual setting (but will not be different if as shot used - in that case temperature and tint displayed in ACR will be different)

                             

                            Whitebalancing of jpeg, like in your sample, is less accurate, because image is already rendered

                            • 11. Re: Fujifilm X100S color rendition with ACR 8.1
                              ffredrikk Level 1

                              @b2martin_a – many thanks for suggesting the DNG Profile Editor. I got an X-Rite ColorChecker Passport and created an X100S camera profile. The color rendition is now perfect!