-
open exr in AE, choosing 32 bit or 16 bit float
Is there a big disadvantage choosing 16 bit float instead of 32 but float for exr.
We need to render a 3d max animation at 1080 p in open exr with render elements.
We need to decide between 32 or 16 bit float. Thing is we are going to use a render farm and the transferring of the files may take time depending on the file size.
-
4 Replies
-
Half-float (16b) is smaller, but less precise. Depending on what you are doing, the precision difference may be imperceptible and not worth the added weight during compositing.
I’d suggest rendering a few frames as a test to help you figure out which is more important to you.
In general: for visual passes (color, lighting, etc.), half-float is fine. For data passes (motion vectors, world position, etc.), full-float is significantly better.
Walter Soyka
Designer & Mad Scientist at Keen Live [link]
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
@keenlive | RenderBreak [blog] | Profile [LinkedIn] -
Thank you for the help.
I am exporting some architectural animations for editing in After effects.
will do some render elements such as reflection and refraction light, and couple more.
Render is Vray and we have a day and a night scene with lots of lights.
I love the linear editing and float especially you can bring down peaked highlights with simple exposure effect.
Would 16 bit be more then good for this… I am using render farm and need to transfer the files back to my computer of course.
-
[Ali Quintana] “Would 16 bit be more then good for this… I am using render farm and need to transfer the files back to my computer of course.”
My rule of thumb is full-float for data passes and that half-float is ok for visual passes, but don’t trust some random guy on the internet — do your own quick test to confirm for yourself that you’ll be happy with the result.
Walter Soyka
Designer & Mad Scientist at Keen Live [link]
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
@keenlive | RenderBreak [blog] | Profile [LinkedIn] -
Thanks, I tried it, I am happy with the 16 bit float result, so far… now I have to pay a renderfarm a lot of money to render a heavy 5 min animation.
I just hope I made the right decision because the real thing will be found out when deep into the render….
But your rule is not just a random internet guy rule, I have seen on the net many other pros giving the same advice. 16 bit float for normal renders and for data passes 32 bit.
Reply to this Discussion! Login or Sign Up