10 Replies Latest reply on Jan 22, 2009 9:38 AM by Panoholic

    Camera Raw - C1Pro Conflict

      I shoot tethered with a Phase One p45+ camera back using C1Pro version 3.7.9. I evaluate my exposure using a Macbeth color chart.
      With all my exposure settings zeroed out and a destination profile of ProPhoto RGB I get a RGB value of the white patch of 252-252-2
      51. When I open the raw file up in Camera Raw v 4.6 using the default camera profile (which is version 4.4) and measure the same white patch I get a value of 243-240-233 needing a plus one stop exposure correction. I am trying to figure out which set of values to trust as an indication of the true exposure. Operating System: Mac OS X (v10.4). I have an open case with Adobe tech support (Case #0180543397) that dates back to November 13, 2008. They have told me that my best hope for an answer is to post here. Any insight would be greatly appreciated.
        • 1. Re: Camera Raw - C1Pro Conflict
          Jao vdL Adobe Community Professional & MVP
          >I am trying to figure out which set of values to trust as an indication of the true exposure.

          Neither. This is purely a function of the raw processor's internal operation and the profiles that are applied to the RAW to render it. There is no standard for this. ACR for example, is known to apply hidden exposure corrections that you have no control over. It is also going to depend very strongly on the tone curve where this value wil end up. Your best bet at really figuring this out is opening the raw file in a program like rawnalyze, but then again, there is not much point to it. If you target correct exposure in ACR, then trust the ACR values. If you target C1, trust that one.
          • 2. Re: Camera Raw - C1Pro Conflict
            Panoholic Level 2
            Which ISO was this shot made with?
            • 3. Re: Camera Raw - C1Pro Conflict
              Level 1
              Jao

              Thanks for the reply. While I understand that different RAW processors will render the same file differently, my real concern is trying to assess the highlight threshold of my camera sensor. The linear exposure nature of digital capture strongly suggests an exposure that stops just short of highlight clipping to maximize the data set. The readout in C1Pro indicates that I am reaching that threshold, while ACR indicates that I am leaving a great deal of data on the table. I have no problem correcting the tonal range in either processor, but I want to be correcting the maximum amount of data my sensor is capable of gathering. Any ideas?

              John
              • 4. Re: Camera Raw - C1Pro Conflict
                Level 1
                G Sch

                I was shooting with an ISO of 100.

                John
                • 5. Re: Camera Raw - C1Pro Conflict
                  Panoholic Level 2
                  > I was shooting with an ISO of 100

                  Then it's all right. Have you tried it @ 400? You would see it by 1 EV too high exposed, and @ 800 by 2 EV.

                  LR/ACR is fooling you by applying an "exposure" correction of -1 EV @ ISO 100, +1 EV @ ISO 400 and +2 EV @ ISO 800. I guess there is no adjustment @ ISO 200, but I can not verify it, for I don't have any such raw file.

                  In order to come closer to the real exposure, you should counteradjust it (always on the "exposure" slider, this is a linear adjustment).

                  The P45 Plus is a special case. The P45 was (is) an ISOless camera, the effect of different ISO settings has been faked by software adjustment. Apparently Adobe neglected to make a test with the P45 Plus (it is expensive even for renting), and the P45 profile is being applied. However, the P45 Plus does have sensor supported ISOs.

                  Thus a properly exposed ISO 800 shot *must* appear in ACR horrendeously overexposed.

                  Btw, neither C1 nor ACR shows the "real exposure", i.e. the raw channel histograms.
                  • 6. Re: Camera Raw - C1Pro Conflict
                    Level 1
                    Thanks so much for the reply, I finally feel like I am getting somewhere. Not to labor the point, but if I understand you correctly, the profiles in ACR are expecting me to use an ISO of 200 with my P45+ back, and that any deviation from that will result in an apparent under or over exposure when in reality the exposure is correct. The only way the readouts in ACR (with the exposure slider set to 0) will be close to the ones in C1Pro is if I use an ISO of 200. Is that correct? If that is the case, is seems like Adobe needs to revisit this. I am not sure how the ACR engine works, but would a counter adjustment made with the exposure slider nullify the mistake in how ACR presents the tonal information, or will there be a hit to the final outputs tonal range. In other words will I have a more detailed tonal range shooting with an ISO of 200 and the exposure slider set to 0 then an ISO of 100 with the exposure slider set to +1 ?

                    Also I was under the impression that the sensitivity of a sensor was fixed, and that all of them achieved a different ISO via software amplification (plus or minus). Would you explain what you mean when you say that the P45 Plus does have sensor supported ISOs. It seems you are suggesting that the sensors sensitivity is actually changing instead of the signal from it being adjusted by software. Thanks again.
                    • 7. Re: Camera Raw - C1Pro Conflict
                      Panoholic Level 2
                      > The only way the readouts in ACR (with the exposure slider set to 0) will be close to the ones in C1Pro is if I use an ISO of 200. Is that correct?

                      Correct, except for the term "readout".

                      Regard this as if the "exposure" slider would be set initially to -1, +1 or +2, depending on the ISO. This is namely, what is happening, only that it this initial setting is not displayed.

                      > would a counter adjustment made with the exposure slider nullify the mistake in how ACR presents the tonal information, or will there be a hit to the final outputs tonal range

                      There is no downside in counteradjusting the exposure.

                      > I was under the impression that the sensitivity of a sensor was fixed, and that all of them achieved a different ISO via software amplification (plus or minus)

                      The
                      i sensitivity
                      of the sensor is fixed (that is the
                      i quantum efficiency).
                      The "real" ISO settings affect the
                      i analog amplification,
                      changing the gain (the voltage created from a certain amount of captured photons) which in turn affects the digital result.

                      The sensors of most MFDBs are working with a fixed gain; the ISO setting is only metainformation, instructing the raw processor to boost the pixel intensities.

                      Following captures show the raw channel histograms of P25+ images, shot at ISO 100, 200, 400 and 800, with decreasing shutter time. These histograms should look very similar, for the analog amplification would make up for the lower exposure; instead, the pixel intensities are getting lower and lower. Here the initial adjustments are correct (although they should be visible):

                      http://www.panopeeper.com/Demo/PhaseOne_P25Plus_ISO100_CF000251_Hist.GIF
                      http://www.panopeeper.com/Demo/PhaseOne_P25Plus_ISO200_CF000252_Hist.GIF
                      http://www.panopeeper.com/Demo/PhaseOne_P25Plus_ISO400_CF000253_Hist.GIF
                      http://www.panopeeper.com/Demo/PhaseOne_P25Plus_ISO800_CF000254_Hist.GIF

                      However, some models do apply analog gain; the P45+ is one of those. Thus the analog gain together with the boost in raw processing blow the pixels.

                      The consequences for you (beside the required adjustment in ACR) are:

                      1. the ISO selection does have an influence on the noise. While the noise with a P45 or P25+ depends alone on the exposure, the noise of the P45+ depends on the ISO setting as well;

                      2. the ISO selection can cause pixel saturation. Again, while the P45 records the ISO selection only as metadata but never causes increasing the pixel intensity in the camera, the P45+ creates higher pixel values. Nominal clipping, caused by the pseudo-ISO can be reversed perfectly in raw processing, but the saturation caused by higher ISO with the P45+ is not reversable, just like saturation caused by too high exposure.
                      • 8. Re: Camera Raw - C1Pro Conflict
                        Panoholic Level 2
                        Cont. consequences:

                        3. if you can not afford the required exposure due to whatever constraints, it does make a difference of using higher ISO with the P45+, as opposed to the P45 etc. Higher analog gain is not a perfect substitute for exposure, but it is better than underexposure.
                        • 9. Re: Camera Raw - C1Pro Conflict
                          Level 1
                          Thank you so much.

                          John
                          • 10. Re: Camera Raw - C1Pro Conflict
                            Panoholic Level 2
                            Welcome, John. If you see banding in some P45+ images, just email me through the forum.