Author Topic: 32bit or 16bit half float exr?  (Read 1623 times)

Offline AC5LT43R

  • Member
  • *
  • Posts: 40
32bit or 16bit half float exr?
« on: June 23, 2014, 02:00:01 PM »
I'm rendering all my passes out as EXR's and when I load them in to Photoshop, it says they are 32bit images. However, when they are in Nuke, it's telling me they are 16bit half float. This is causing an issue specifially with the point pass outputs.

Any idea?

Offline Oshyan

  • Planetside Staff
  • *
  • Posts: 13022
  • Holy snagging ducks!
Re: 32bit or 16bit half float exr?
« Reply #1 on: June 23, 2014, 09:07:31 PM »
They are 16bit half float currently. We are hoping to add the option for full 32 bit/channel images in the near future.

- Oshyan

Offline AC5LT43R

  • Member
  • *
  • Posts: 40
Re: 32bit or 16bit half float exr?
« Reply #2 on: June 24, 2014, 10:47:10 AM »
Ah right, thanks Oshyan.
« Last Edit: June 24, 2014, 01:00:16 PM by AC5LT43R »

Offline digitalguru

  • Member
  • *
  • Posts: 419
Re: 32bit or 16bit half float exr?
« Reply #3 on: June 27, 2016, 03:14:51 PM »
Just thinking about that - especially when exporting a vector map to use as displacement in an external 3d program

Would be very useful as an option in the Render node, but also as a selectable option per Render element for the example AC5LT43R mentioned.

Since the elements are grouped under Lighting, Data etc, each group could have a 32bit switch? Then your data render elements could be 32 bit and your color beauty elements 16bit float?
« Last Edit: June 27, 2016, 03:16:56 PM by digitalguru »