1 2 Previous Next 59 Replies Latest reply on Apr 23, 2011 8:16 AM by Andrew_Hart

    ACR Camera Profiles and Camera Manufacturing Tolerances

    loareg Level 1

      For every camera model there is one set of model specific ACR camera profiles. Although those profiles are model specific, they are not camera specific. I'd suspect that not all cameras of the same model will produce exactly the same colors due to manufacturing variations. I'm wondering if differences in color rendition between cameras of the same model are visible or neglectable? Any experiences?

       

      A similar question: Taking the "same" photo with a Canon 550D and a Nikon D90 camera and processing the RAWs in LR with Adobe Standard profile can result in quite different colors (even when setting WB manually). Shouldn’t the processed photos look almost the same?

       

      Thanks in advance for any help!

        • 1. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
          ssprengel Adobe Community Professional & MVP

          Profile creation is part science and part art, with the latter due to the fact that not all color error can be minimized so some colors are optimized for and some are left to drift.  There are also hue twists built into the Adobe profiles that change hue based on luminance to make photos look "better" some someone's opinion. 

           

          Both the illuminate Kelvin temperature and the particular spectral response of the illuminate affect how colors look, with daylight-type lighting in the 5000K to 6500K having the least problems, while fluorescent and other high-pressure gas lighting with its uneven spectrum and incandescent lighting with its severe reddish bias cause the most problems.  Of course something like sodium vapor streetlights have worse issues but typically you don't photograph things in this lighting or at least don't expect the photos to look right.

           

          If you're wanting to make images from more than one camera be similar then you can create your own profiles for each camera in the lighting your photos are in using something like the X-Rite color-checker passport:

           

          http://xritephoto.com/ph_product_overview.aspx?ID=1257

           

          You can also further optimize the color profile by computing values for the sliders underneath the profile using a photoshop script:

          http://21stcenturyshoebox.com/tools/ACRcalibrator/

          1 person found this helpful
          • 2. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
            Hudechrome Level 2

            That's like expecting Kodachrome and Fujichrome to look the same.

             

            They won't.

            • 3. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
              loareg Level 1

              Thanks for your answer ssprengel. If I understand correctly, you would argue that the differences I see between the 550D and D90 photos (both RAWs processed in LR with Adobe Standard and identical WB) are due to the profile creation not being fully accurate an not due to manufacturing variations? I expected Adobe to use a similar process as the X-Rite color-checker to create the same Adobe Standard color rendition for all camera models. Essentially, you're arguing that such a process is not really accurate and cannot be done accurately at all?

              • 4. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                loareg Level 1

                Hudechrome, I don't think so. I don't expect the Camera Standard profile of a Canon to look the same as the Camera Standard profile of a Nikon. I expect pictures from different cameras processed with the Adobe Standard profile available for all camera models in ACR/LR to look the same. They don't and I'm wondering where the differences come from.

                • 5. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                  areohbee Level 5

                  Hudechrome wrote:

                   

                  That's like expecting Kodachrome and Fujichrome to look the same.

                   

                  They won't.

                   

                  I don't think so. Expecting Kodachrome and Fujichrome to look the same is more like expecting Canon Camera Standard to look like Nikon Camera Standard.

                   

                  Adobe Standard "should" look the same for all cameras. I realize thats not as easy as it sounds, but your analogy is not correct in this case.

                   

                  Summary: I *think* the idea was for Adobe Standard to look very close to the same for all cameras.

                   

                  Rob

                  1 person found this helpful
                  • 6. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                    Hudechrome Level 2

                    My Nikon D80 and D90 don't look the same and I have run comparisons between the Canon 7D and the Nikon D90. Taken together, they all different from each other.

                     

                    The biggest difference between the D80 and the D90 seems to lie with the much larger dynamic range of the D90. Compared to the D80 at first glance, the D90 seems washed out at the lower values. This is easily overcome in ACR, but even with that, the subjectivity of the reproduction sometimes gives a nod to one over the other.

                     

                    The closest film comparison is Fuji Astia vs Provia. The D90 at default Nikon Camera Standard resembles Astia, while the D80 is a cross between Provia and Velvia. All this is controlable. One slider I use to enrich the D90 presentation is the black slider

                     

                    The Canon has other undefined differences which I have simply noted by viewing. I haven't engaged in any tweaking of that camera's images.

                     

                    So I'll use both the D80 and the D90 according to what I am wanting to happen. Of course, there are times where the differences simply inform the operator of what may be doable, and then one is tweaked to look much like the other.

                     

                    I checked out sprengel's links to the calibrator software. They have stopped at CS3, it seems. How does it perform with CS5? I may want to at least run a calibration of both cameras and look again.

                     

                    And, of course, Adobe Standard and Nikon Standard do not agree. At all. So, when is a standard not a standard?

                     

                    When there is more than one.

                     

                    Looking back at your post, I should specify that the profile I used when making the comparisons have been  the Camera standards, not Adobe Standard.

                     

                    Message was edited by: Hudechrome

                    • 7. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                      Noel Carboni Level 7

                      Y'know, it sure seems like someone ought to be able to create a tool where you take a picture of a test chart with a continuous tone (i.e., a HUGE number of different colors), then create a profile from it.  It's pretty clear the swatch-based test charts aren't up to the task.

                       

                      Don't I recall Vit Novak is already doing this?  Though I *think* he's been taking photos of a display monitor, which sounds to me like it would inject its own problems.

                       

                      -Noel

                      • 8. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                        ssprengel Adobe Community Professional & MVP

                        Having many more swatches would be much better, I think.  Using only 18 + 6 neutral is a problem.  I assume Adobe uses more, internally.

                         

                        And monitor-based colors would be ok as long as you have a spectrophotometer-based mathmatical model to compare against.

                         

                        Using printed swatches you're limited the reflectance gamut.

                        • 9. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                          Hudechrome Level 2

                          Using reflected color is best, imo, because that's what we photograph. With a monitor as a source, daym, the problems with repeatability, drift, and ??? bring up too many unanswered questions. At the very least, I would insist on a feedback loop controlling the output. I might even consider having a readout of the loop values at the moment of exposure.

                          • 10. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                            Vit Novak Level 3

                            Noel Carboni wrote:

                             

                            Y'know, it sure seems like someone ought to be able to create a tool where you take a picture of a test chart with a continuous tone (i.e., a HUGE number of different colors), then create a profile from it.  It's pretty clear the swatch-based test charts aren't up to the task.

                             

                            Don't I recall Vit Novak is already doing this?  Though I *think* he's been taking photos of a display monitor, which sounds to me like it would inject its own problems.

                             

                            -Noel

                             

                            As about camera profiles, I tried 2 methods

                             

                            First  one was really using the monitor. It created some problems due to  limited precision of jpegs from camera and moire, but since I was  comparing raw and jpeg from the camera (and not some absolute measured  values), result was quite good. But procedure was quite cumbersome, so I made only one camera profile this way (400D standard)

                             

                            My current method is software based - I'm generating a raw file containing test chart by the software and developping it by DPP or NX2 (or the camera itself in case of Canon compacts), then calculating the profile on the difference between raw and 16 bit tiff. So it returns the same result as that software / camera, with precision limited only by size of the profile and interpolation error (when profile is used by ACR) connected with it. I suppose Adobe team is doing tha same (at least for the latest generation of profiles), as results are very similar

                            • 11. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                              loareg Level 1

                              Vit Novak wrote:

                               

                              My current method is software based - I'm generating a raw file containing test chart by the software and developping it by DPP or NX2 (or the camera itself in case of Canon compacts), then calculating the profile on the difference between raw and 16 bit tiff. So it returns the same result as that software / camera, with precision limited only by size of the profile and interpolation error (when profile is used by ACR) connected with it. I suppose Adobe team is doing tha same (at least for the latest generation of profiles), as results are very similar

                               

                              This makes sense to me. You're essentially "reverse engineering" the "special" color processing by Nikon or Canon and don't have to deal with the cameras at all. Still, did you ever compare pictures taken with different cameras of the same model and processed with the same camera profile in ACR/LR?

                               

                              A question about the camera profiles in ACR/LR: In another thread you mentioned that one could replace the camera model string in the profile file of camera model A with the string of camera model B and then apply the modified profile to camera B in ACR/LR in order to mimick the color rendition of model A with model B. Wouldn't this only work if both model A and B used the same color primaries (color filters) and therefore had the same starting point?

                              • 12. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                Vit Novak Level 3

                                Exactly, it's a kind of inverse ingineering in both methods.

                                 

                                I didn't compare pictures taken with different cameras of same model, as I have only one camera of each model. But I presume that color profile built into camera is equal for all cameras of that model, so it's the same when you develop the raw with ACR, using the same profile every time

                                 

                                As about using the profile for a different camera, it's safe only when the camera has the same sensor, like for instance Nikon D7000 and D5100, or Canon G9 and A650. If sensors are different, their spectral responses are also different. Depending how big these differences are, you will get more or less acceptable or totally unacceptable result.

                                • 13. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                  Hudechrome Level 2

                                  DXO Pro appears to be doing that, at least in some fashion as one of the tweaks available is to mimic other camera models, and the list is rather large.Trying to mimic a larger dynamic range with the output from a device with a smaller one sends red flags to me. Is it by design or simply a SWAG?

                                  • 14. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                    Hudechrome Level 2

                                    That's what I would expect, which makes DXO process suspicious. I haven't investigated it much because I mostly use their lens and noise correction, leaving RAW otherwise alone.

                                    • 15. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                      Noel Carboni Level 7

                                      Interesting.  So, Vit, as noted your technique models the processing while avoiding the physical camera characteristics.  I can see how this can be useful.  But couldn't it be extended to where you're shooting a photo of a rich test pattern, though (e.g., like the one in your avatar)?  Only thing would be you'd have to have very good registration so the software would be profiling the proper colors in the proper pixels.  It would also be a good idea to take multiple photos and stack them, to average out noise.

                                       

                                      If you have two cameras, not only could you create profiles for each, but I imagine in cases where they're the same make and model you might be able to use the same profile but perhaps add your own tweaks to the default Camera Raw settings then store them as defaults.  Remember that the settings, which include choice of profile, can be stored independently by camera serial number with a change of one of the prefs.  I'll bet you could get everything very close and repeatable between multiple bodies.  It's surprising this is so far from reality at the moment.

                                       

                                      I wish I had Vit's capabilities; I'd love to experiment with this more.  Seems like a "camera calibration system" could be created and marketed.

                                       

                                      Or that Camera Raw itself could provide such a feature.

                                       

                                      -Noel

                                      • 16. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                        Hudechrome Level 2

                                        Are you finding two cameras of the same model noticeably different?

                                        • 17. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                          Noel Carboni Level 7

                                          No, to be honest I have never had two cameras of the same model.  There's only one of me, and the lenses ARE interchangeable after all... 

                                           

                                          However, I have had dealings with people who do have multiple bodies, and even the same model body with the same model lens can yield slightly different color and exposure.  It's an observed fact.

                                           

                                          For people who want absolute consistency, one should think that a profiling process should be available that will get them more than just "close".

                                           

                                          -Noel

                                          • 18. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                            Hudechrome Level 2

                                            Close has always been a pain, as it is obvious that the slight variations in lens characteristics will affect the output. This cannot be helped as the essence of manufacturability is tolerance. So differences are going to be there. And even in profiling, absolute perfection and absolute matching is nowhere's near guaranteed. Not only that but I also suspect differences between viewers perceptions exist. In audio this is true, which is why the concept of "Golden Ears" exists. Also, some people also have perfect pitch.

                                             

                                            So has anyone come across individuals with "Golden Eyes" or capable of looking at a color labelled a certain way like burnt Umber and say "that isn't burnt umber. It's too.... and be right? My SO can approximate that but I have been able to deceive her. In which case, we get into an argument!

                                             

                                            While I stand behind such explorations and definitions and profiling as being pursued here, It sort of reminds me of tuning a piano with a microphone attached to a nulling device. It is never as good as a human tuner with a golden ear. But then, measuring and qualifying sound is different than light.

                                            • 19. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                              loareg Level 1

                                              Noel Carboni wrote:

                                              ...

                                              However, I have had dealings with people who do have multiple bodies, and even the same model body with the same model lens can yield slightly different color and exposure.  It's an observed fact.

                                              ...

                                              Thanks for your contribution Noel. That was my initial question :-) It will probably be hard to find someone who wants to object, because most people simply don't own more than one camera body of the same model.

                                              • 20. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                loareg Level 1

                                                Vit Novak wrote:

                                                 

                                                ...

                                                But I presume that color profile built into camera is equal for all cameras of that model

                                                ...

                                                Thanks for your answer Vit. For Canon cameras the color profile seems to be tweakable in camera. A friend of mine had a Canon 550D that took slightly reddish pictures. He wasn't able to solve the problem by simply adjusting WB. So he sent the camera to a Canon service center. Canon replaced the focusing screen (would fix WB problems, if it had been tinted somehow) and "adjusted the color matrix", as they wrote. His 550D now features neutral (perhaps even a bit muted) colors and the reddisch color cast is gone. So it seems there was something wrong with his camera and it could be fixed "in software". I don't know if this is Canon specific or not.

                                                • 21. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                  MadManChan2000 Adobe Employee

                                                  Individual units of a given camera do have different responses (for both color and noise). For example, the exact gains applied to each color channel are slightly different from unit to unit. They can be (and often are) normalized through calibration data, but usually not perfectly.

                                                   

                                                  To make matters more complicated, there are differences among lenses. For example, shooting a fixed subject under a stable, controlled lighting condition using two different lenses (e.g., Canon 50 f/1.4, or Canon 100 f/2.8 macro) from the same position (with the same camera & exposure) will result in slightly different colors.

                                                   

                                                  Many of these differences are effectively "normalized out" in practice by applying white balance, especially doing a click-WB on an image area. Hence they go unnoticed. But sometimes the differences can be seen (e.g., they result in slightly different white balance numbers).

                                                  • 22. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                    Hudechrome Level 2

                                                    This is what I would expect, and would have said as much except I don't do bench testing!

                                                     

                                                    And while I agree that clicking on a specific area for WB will tend to normalize between lenses, it shifts the color to normalize the WB, and colors off that point may still not match. I've seen this shooting architecture with different lenses. It's not enough to necessarily write home about, but some architects....!

                                                     

                                                    I set this test up some years back when I obtained my D90 and ran tests for distortion between lenses. That's when I saw the off color between frames that were WB with the eyedropper on the same place on the building deliberately picked to allow this. (White trim). Change lens brands and you get to do all this again!

                                                     

                                                    Message was edited by: Hudechrome

                                                    • 23. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                      Vit Novak Level 3

                                                      loareg wrote:


                                                      Thanks for your answer Vit. For Canon cameras the color profile seems to be tweakable in camera. A friend of mine had a Canon 550D that took slightly reddish pictures. He wasn't able to solve the problem by simply adjusting WB. So he sent the camera to a Canon service center. Canon replaced the focusing screen (would fix WB problems, if it had been tinted somehow) and "adjusted the color matrix", as they wrote. His 550D now features neutral (perhaps even a bit muted) colors and the reddisch color cast is gone. So it seems there was something wrong with his camera and it could be fixed "in software". I don't know if this is Canon specific or not.

                                                       

                                                      I don't work for any camera manufacturer so I really don't know is it possible to calibrate colors of individual camera by some diagnostic software.

                                                       

                                                      However, it's a bit strange that focusing screen was affecting colors, because when you press the shutter, light from the object goes through the lens and falls directly to the sensor. And I presume that WB in the camera is calculated only from raw data from the sensor (except in some cameras with additional WB sensor), the same way as when you select Auto WB in ACR (just that every manufacturer has different algorithm for this, so AWB of the camera generally won't produce the same result as Auto WB in ACR). So, just in case, check that the problem wasn't "solved" by applying WB shift in the menu 

                                                       

                                                      But maybe they replaced something else, like antialiasing / IR filer, which is placed in front of the sensor ,,,

                                                      • 24. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                        ssprengel Adobe Community Professional & MVP

                                                        The focus screen if it’s the one for viewfinder image wouldn't have anything to do with the white-balance so could have been an independent repair item.

                                                         

                                                        Loareg can you verify whether the friend's camera pictures that were reddish and then corrected by a service-center visit were RAW files or in-camera JPGs?

                                                         

                                                        There'd obviously be a color-profile and color-matrix associated with converting RAW to JPG in-camera, but that wouldn't affect RAW file data, but it may be possible to adjust the individual gain of the RGB channels coming off the sensor to make up for differences in the Bayer filter transparency and saturation or the pre-Bayer anti-moire filter transparency or the UV-filter transparency or it's alignment, or even for variance in the amplification of the signals coming off the sensor due to electronic component variability.  We see maze patterns for differences between the two green channels so it's possible for other sensor signals to be less than optimal, too.

                                                        • 25. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                          Vit Novak Level 3

                                                          There is another possibility for Canon cameras (I forgot that because I never used it, although even my 400D has this possibility)

                                                           

                                                          http://www.canon.co.jp/imaging/picturestyle/index.html

                                                           

                                                          So, Canon provided a tool similar to DNG profile editor - you can make your own picture styles (similar to "recipee" in DNG PE) and use them with Canon development software (DPP ...) or upload to the camera. So check also that problem wasn't "solved" just by selecting some custom picture style in the camera

                                                           

                                                          But as ssprengel said, all these things affect only jpeg from camera and not raw

                                                          • 26. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                            loareg Level 1

                                                            Vit Novak wrote:

                                                             

                                                            However, it's a bit strange that focusing screen was affecting colors, because when you press the shutter, light from the object goes through the lens and falls directly to the sensor.

                                                            That was my first thought, too. But on second thought it seems to make sense. A DSLR camera's light meter usually sits somewhere in the viewfinder part of the camera and sees the image of the scene through the focusing screen. At least for Nikon cameras auto white balance also takes into account what the light metering sensor sees. This could be valid for Canon cameras, too (considering what Canon service did to fix the color cast).

                                                            • 27. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                              deejjjaaaa Level 2

                                                              > The focus screen if it’s the one for viewfinder image wouldn't have anything to do with the white-balance

                                                               

                                                              focusing screen lays in the path of the light before the WB/metering sensor(s) for typical dSLRs... so if it has some color tint it might affect the wb measurement

                                                              • 28. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                loareg Level 1

                                                                Loareg can you verify whether the friend's camera pictures that were reddish and then corrected by a service-center visit were RAW files or in-camera JPGs?

                                                                I tested the 550D myself for a day before it was sent to the Canon service center. The reddish cast was in the JPGs *and* the RAWs. After the service center visit the cast was gone in both the JPGs and RAWs. Now the colors are very neutral, maybe even a bit muted.

                                                                 

                                                                We see maze patterns for differences between the two green channels so it's possible for other sensor signals to be less than optimal, too.

                                                                What *two* green channels and waht maze pattern? I know in a Bayer filter there usually are two green pixels and just one red and one blue pixel. Do you speak of those two pixels as two green channels?

                                                                • 29. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                  deejjjaaaa Level 2

                                                                  > What *two* green channels and waht maze pattern?

                                                                   

                                                                  well known problem for Oly cameras for example - CFA is R-G1-G2-B, where G1 != G2 ... color dyes are different for 2 green pixels, hence w/o some special preprocessing some demosaick methods that assume G1 = G2 will generate so called maze patterns.

                                                                  • 30. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                    loareg Level 1

                                                                    Vit Novak wrote:

                                                                     

                                                                    There is another possibility for Canon cameras (I forgot that because I never used it, although even my 400D has this possibility)

                                                                     

                                                                    http://www.canon.co.jp/imaging/picturestyle/index.html

                                                                     

                                                                    So, Canon provided a tool similar to DNG profile editor - you can make your own picture styles (similar to "recipee" in DNG PE) and use them with Canon development software (DPP ...) or upload to the camera. So check also that problem wasn't "solved" just by selecting some custom picture style in the camera

                                                                     

                                                                    But as ssprengel said, all these things affect only jpeg from camera and not raw

                                                                    Those custom picture profiles and custom curves were my friend's first attempt to fix the color cast himself. It did work to some extent, but he wasn't able to get consistently neutral colors. As ssprengel said in this thread "Profile creation is part science and part art"

                                                                    • 31. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                      loareg Level 1

                                                                      deejjjaaaa wrote:

                                                                       

                                                                      > What *two* green channels and waht maze pattern?

                                                                       

                                                                      well known problem for Oly cameras for example - CFA is R-G1-G2-B, where G1 != G2 ... color dyes are different for 2 green pixels, hence w/o some special preprocessing some demosaick methods that assume G1 = G2 will generate so called maze patterns.

                                                                      I see. Thanks!

                                                                      • 32. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                        Vit Novak Level 3

                                                                        Yes, light meter in some new Canon cameras, including 550D, seems to be sensitive to colors (like in Nikon), so this may be included into WB calculation. Did you manually set the same WB temperature and tint in ACR when comparing results from raw before and after this repair? In case they really fixed the problem, there should be a difference in that case also (and not only with As shot WB selected)

                                                                        • 33. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                          loareg Level 1

                                                                          Individual units of a given camera do have different responses (for both color and noise). For example, the exact gains applied to each color channel are slightly different from unit to unit. They can be (and often are) normalized through calibration data, but usually not perfectly.

                                                                          Thanks! So for the 550D with the reddish color cast those color channel gains may have been a little bit off and the Canon service center probably fixed the problem by calibrating those gains then. Sounds reasonable

                                                                          • 34. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                            loareg Level 1

                                                                            Vit Novak wrote:

                                                                             

                                                                            Did you manually set the same WB temperature and tint in ACR when comparing results from raw before and after this repair? In case they really fixed the problem, there should be a difference in that case also (and not only with As shot WB selected)

                                                                            No, I did not, as it isn't my camera and the camera is on holiday now Because the color cast could not be eleminated by simply adjusting white balance and now seems to be gone, I would argue that the Canon service center did more than just adjust white balance somehow. But again, I did not do all the tests myself (not my camera and I'm Nikon).

                                                                            • 35. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                              Vit Novak Level 3

                                                                              Ok. I hope that the problem was actually fixed.

                                                                               

                                                                              Was just curious about this (in case something similar happens to me when I buy the new camera), because I had some bad experiences with the Pentax service several years ago

                                                                              • 36. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                                loareg Level 1

                                                                                Vit Novak wrote:

                                                                                 

                                                                                Was just curious about this (in case something similar happens to me when I buy the new camera), because I had some bad experiences with the Pentax service several years ago

                                                                                Well, if you really would like to "reverse engineer" what the Canon service center did to fix the problem, I could provide you ColorChecker Passport images shot with the 550D (RAW+JPG). At the moment I do only have images from the before state. When the camera is back from holiday I could take additional images for the after repair state. But I'm not sure if the shooting conditions were "controlled" enough to get meaningful results. The images were taken rather quickly in blue sky daylight outdoors. Right now I don't have the time to analyse the said 550D any further for myself.

                                                                                • 37. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                                  ssprengel Adobe Community Professional & MVP

                                                                                  The "focusing SCREEN" is in front of the viewfinder and is used to judge focus with your eye when the lens projects the image, there, with the help of the half-silvered mirror and pentaprism. 

                                                                                   

                                                                                  The auto-focus sensor array and exposure metering sensor array are in the floor of the lens box area and could also have sensors that detect white-balance along with judging exposure for various colors.

                                                                                   

                                                                                  For a reminder of what a "focusing screen" is, here are instructions how to change one with pictures of said item:

                                                                                  http://www.focusingscreen.com/work/500den.htm

                                                                                   

                                                                                  The "focusing screen" is not the same as the "focusing sensor" and I wouldn't expect a Canon service tech to use that term incorrectly on a workorder.

                                                                                  • 38. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                                    Hudechrome Level 2

                                                                                    That's a film camera, is it not? I don't expect that the screen in the dslr's to be the same nor  a user capable change to be available. At least, there is no instructions to do so in the Nikon manuals.

                                                                                     

                                                                                    I also would take the nomenclature on a repair sheet with a grain of salt. FAPP, there is no focusing screen on a dslr only a viewing screen with little focus capability!

                                                                                    • 39. Re: ACR Camera Profiles and Camera Manufacturing Tolerances
                                                                                      ssprengel Adobe Community Professional & MVP

                                                                                      The focus-screen changing instructions show a Canon 500D, not a film camera.  They are changing the focusing screen to be one more like a film camera has with a split circle useful for manual focusing.  I actually changed the focusing screen on my original Digital Rebel with this type when the secondary mirror hinge pin broke (a common problem with this original DR model) to see if I could use it with manual focus.  My eyes are old enough this wasn't feasible, either, so I put the original screen back and sent it in and waited 6 weeks for it to be returned with a new mirror box and shutter.

                                                                                       

                                                                                      The viewfinder has a focusing screen used to verify AF and for manual focus feedback.  I use it this way all the time when taking close-ups by moving by body back and forth very slightly until the right part of the subject is in focus.

                                                                                      1 2 Previous Next