Announcement

Collapse
No announcement yet.

ETTR and Middle Gray

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #76
    Frank, so far it does look like ETTR at just under clipping for the scene is definitely a viable option. The way we both have been shooting has yielded great results, and this test could indicate that the BMCC is very forgiving with overexposure (as we already know).

    That being said, there is an analytical part of me that wants to know more to give me more confidence in where I place exposure (and to be able to explain the way I expose to producers in a way that's not offensive). It seems obvious this test was performed with a BMCC on old firmware. I want to perform my own test (without the magenta highlights).

    Comment


    • #77
      Originally posted by Frank Glencairn View Post
      It's more like learning to shifting gears in a car.
      At the beginning it's an ordeal, after a while it's a no brainer - you just do it.
      As with anything to learn i guess.

      @Jared - definitely think it's worth doing your own testing. I'd like to as well at some stage, to see what works for how i like to shoot. I always learn more if i can do my own comparisons.
      Blackmagic Design
      My BMD LUTs.

      **Any post by me prior to Aug 2014 was before i started working for Blackmagic**

      Comment


      • #78
        Originally posted by Jared Caldwell View Post
        Frank, so far it does look like ETTR at just under clipping for the scene is definitely a viable option. The way we both have been shooting has yielded great results, and this test could indicate that the BMCC is very forgiving with overexposure (as we already know).

        That being said, there is an analytical part of me that wants to know more to give me more confidence in where I place exposure (and to be able to explain the way I expose to producers in a way that's not offensive). It seems obvious this test was performed with a BMCC on old firmware. I want to perform my own test (without the magenta highlights).
        Okay, I'm far from a math genius and my deeper math knowledge in debayering and signal processing is limited, but from what I can see, if you work with raw, and you use the "ISO" slider in the raw(camera) tab, you just remapping the values without doing any damage to the signal. So as long as you have a clean signal to begin with, there should be no color shifting or anything. But if you use a LUT, remapped values can read different of course.
        That's why I always try to get a signal that is clean as possible.

        When you look at the tests, you can see, what lower exposure does with the shadows of the girls hair (and yes, Vimeo compression didn't make it any better).
        Blog: http://frankglencairn.wordpress.com/

        Comment


        • #79
          Wow, you guys put me to shame. I just tend to play with a camera till I am happy with the footage in a few different lighting scenario's then just go with my gut, which, touch wood, has not failed me yet. I will not heavily ETTR though, this has given me some trouble in the past, I usually try to keep it at most a stop over and try to stay away from under exposing. Maybe I need to do a bit more reserch and testing.

          Comment


          • #80
            It depends what I am shooting really. Like tonight I went into the city centre and shot some christmas stuff, mostly on the Industar 69 pancake lens. I shot it all at 400ASA, f4. Pulling it into resolve and do a luminance adjustment and threw a basic REC701 LUT on it, and it looks great. If I knew that I was going to really be doing something serious with the footage, then I would probably pay much more attention. I am not going to be so critical for everything I am shooting. No point.

            What the OP is talking about is all the more important when it comes to matching sequences with different cameras, and get the best image possible. I stand by that, but you have to separate different shooting modes.

            Comment


            • #81
              Well here's some plugin specific evidence that ETTR may not be ideal. This was graded with standard FilmConvert Vision3 settings (no grain) and a cooled white balance as it was too warm at 5600K.

              The first is shot at approximately the right exposure (no adjustment in FilmConvert).



              The second shot is ETTR with probably about 1.5-2 stops under exposure in post. The exposure slider in FilmConvert doesn't display in stops. No other settings, like saturation were changed.



              The saturation is greatly increased when underexposing an overexposed image in FilmConvert. I'll do the same with Nick Shaw's Rec709 LUT plugin and update.

              Granted this is A) ProRes, and B) not a raw converter suite like ACR which would handle things differently.

              UPDATE

              I compared the normal and ETTR shots corrected for with the LUT plugin and that one handles it just fine. Within 5% of the same image, only a different aperture and different depth of field.

              FilmConvert may be doing the right thing for the intent of what it does to emulate film. I'm not sure.
              Last edited by nickjbedford; 11-29-2013, 03:03 PM.

              Comment


              • #82
                Could you post the 2 DNG's Nick? Would be interested to see this in Resolve too.
                Blackmagic Design
                My BMD LUTs.

                **Any post by me prior to Aug 2014 was before i started working for Blackmagic**

                Comment


                • #83
                  Originally posted by CaptainHook View Post
                  Could you post the 2 DNG's Nick? Would be interested to see this in Resolve too.
                  I actually shot these in ProRes, but here's two frames of the ProRes 422 HQ files.

                  http://www.nickbedford.com/uploads/e...l-exposure.mov
                  http://www.nickbedford.com/uploads/e...r-exposure.mov

                  Comment


                  • #84
                    As I said, ProRes reacts a bit different than raw, since you not just remapping the values during debayer.
                    Blog: http://frankglencairn.wordpress.com/

                    Comment


                    • #85
                      Originally posted by nickjbedford View Post
                      I actually shot these in ProRes, but here's two frames of the ProRes 422 HQ files.

                      http://www.nickbedford.com/uploads/e...l-exposure.mov
                      http://www.nickbedford.com/uploads/e...r-exposure.mov
                      Files don't work?
                      Blackmagic Design
                      My BMD LUTs.

                      **Any post by me prior to Aug 2014 was before i started working for Blackmagic**

                      Comment


                      • #86
                        For those that want to know.....

                        Middle grey in BMD film is 38.4%....

                        jb

                        Comment


                        • #87
                          Thanks John !
                          Blackmagic Design
                          My BMD LUTs.

                          **Any post by me prior to Aug 2014 was before i started working for Blackmagic**

                          Comment


                          • #88
                            Originally posted by Jared Caldwell View Post
                            Thanks for the response, Kholi. I was just quoting something from JB when he said he felt like the LOG curve is more linear in nature (maybe that is a misquote) and that he felt that placing your subject at a higher exposure didn't rob you of color information.
                            This was only in regards to RAW not ProRes

                            JB.

                            Comment


                            • #89
                              Originally posted by John Brawley View Post
                              For those that want to know.....

                              Middle grey in BMD film is 38.4%....

                              jb
                              we can check this in waveform, right?
                              Dustin Uy-Filmmaker

                              Comment


                              • #90
                                Originally posted by dcloud View Post
                                we can check this in waveform, right?
                                That's the point. That's where it sits.

                                JB.

                                Comment

                                Working...
                                X