Announcement

Collapse
No announcement yet.

EXR - VRIMG and color depth

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

  • EXR - VRIMG and color depth

    Hi all
    Just little bit confused. Why Eyon Fusion 6 is telling me, that exr that was saved by vray is just only 16 bit? I used VRIMG saving option, rewrited the extension from vrimg to exr and save the file. I din't use clamp ouput to 1.0 at color management. We are using Vray 1.5 sp5 and max 2010. At max 2011 is option to save every channel of exr with different depth and vray channels, but not in max 2010.
    So my question is. What is real color depth of vray exr? Every channels have the same color depth?

  • #2
    by default vray saves exrs as 16 bit half precision floating point, which is not, however, 16 bit integers. even if half float, those exrs are still hdr files, with no clamping of any kind going on.
    half precision is usually more than enough for any comping, post processing operation. if you need full float outputs though, you have to specify that via command line.
    here's a thread you may want to have a look at:

    http://www.chaosgroup.com/forums/vbu...half+precision

    Comment


    • #3
      vray2.0 now has an option in the saving dailog to save 32bit, but like rivoli says, 16bit is fine for most things. You wont notice a difference between 16bit and 32bit with your monitor and eyes, except for a much larger file size
      Kind Regards,
      Morne

      Comment


      • #4
        ehm...could anybody pleas explain easily the difference between INTEGR and FLOAT? I have found a lots of threads about that for programmers, but not so much good information about graphics. Hard to understand:
        int type is used to store positive or negative integers.size of this datatype is 2 bytes.range of int is from -32768 to +32767.

        float datatype stores a single precision floating point(real)number.size of this datatype is 4 bytes.range of float is 3.4*(10**-38 )to 3.4*(10**+38 ).

        Does it mean that in 16 bit integr format can be 65535 variation of b/w color. But how many color variations can be at 16 bit float picture, where i would like have gradient from black to white color? If I can say black and white in 16 bit meaning ...

        As for me...When I saved my velocity channel to the PNG, which is 16 bit integr format, my motion blur wast so much dynamic. When I render the same render to 16bit exr float, the motion blur was nice, looks weight better then by PNG. So....there must be some different. Maybe better don't wonder about that, and render everything to the EXR that is default 16 float by Vray.

        Comment

        Working...
        X