Announcement

Collapse
No announcement yet.

Color mapping - Clamp output

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

  • Color mapping - Clamp output

    Hi dears.
    I am wondering if there is a way to save more than 8 bits color depth but with advantages of antialiasing if you use clam output in color mapping option. Sometimes, especially in interior scenes with some reflective materials clamp output drastically decreases noise. So..sometimes I have to use clamp output but this way, I will loose HDR output which I would like to use for composition. Oh...I really don't know the right way.
    Also, Am I loosing hdr in other channels this way? I mean Z channel or velocity...and so one...
    Any ideas?
    Thank you

  • #2
    you can still save to a hdr format even with clamp output checked. even if you clamp any value above one, you'll still have a lot of colour information that will lost if you save to a ldr file format. plus, if you plan to do any compositing that involves recreating the final output from the render elements, working linear with hdr files is the proper way to get consistent results.

    Comment


    • #3
      You do not need to clamp at 1.0 we typically clamp at 2-3. While it still gives enough range in float, it certainly removes undesirable grain in reflections. Always render in exr, they are optimized for smallest possible float size, you may choose exr 16 bit half float, which is plenty of range typically. For the zdepth however, see if 16 bit half float gives you the result you need. Often we have to render separate zdepth at 32 bit full float.
      Dmitry Vinnik
      Silhouette Images Inc.
      ShowReel:
      https://www.youtube.com/watch?v=qxSJlvSwAhA
      https://www.linkedin.com/in/dmitry-v...-identity-name

      Comment


      • #4
        Originally posted by Morbid Angel View Post
        You do not need to clamp at 1.0 we typically clamp at 2-3. While it still gives enough range in float, it certainly removes undesirable grain in reflections. Always render in exr, they are optimized for smallest possible float size, you may choose exr 16 bit half float, which is plenty of range typically. For the zdepth however, see if 16 bit half float gives you the result you need. Often we have to render separate zdepth at 32 bit full float.
        What difference do you see exactly? I am rarely very happy with post DOF because the Zdepth mask seems to create odd edges that show up in hi-res renders. I always use 1/2 float though, maybe that is why?

        Thanks for any insights. /b
        Brett Simms

        www.heavyartillery.com
        e: brett@heavyartillery.com

        Comment


        • #5
          its quite complicated actually. I have seen amazing dof work out of nuke. You can see it here in my latest demoreel http://vimeo.com/27501594 The sequences with the house is done with macro photography style which is purely achieved in nuke. You may also notice DOF in some tron shots, which were also all done in nuke.
          We typically have to render non clamped non aliased zdepth, along with a few other passes and they are put together in nuke through a series of custom nodes.
          Dmitry Vinnik
          Silhouette Images Inc.
          ShowReel:
          https://www.youtube.com/watch?v=qxSJlvSwAhA
          https://www.linkedin.com/in/dmitry-v...-identity-name

          Comment


          • #6
            Killer reel!

            Thanks for the info. I don't use Nuke so it's probably going to be less helpful for me, but I'll give it a look next time it comes up on a project.

            b
            Brett Simms

            www.heavyartillery.com
            e: brett@heavyartillery.com

            Comment

            Working...
            X