View Full Version : rgb half vs short

06 June 2010, 02:02 PM
anyone know the main difference between rendering
rbg 16 short and rgb 16 half?
I know , half is a floating point with half precision:
16bit short is 16bit integer, means the maximum of color variations per rgb channel is 2^16 = 65536

My question is:
when it comes to production which one to pick for after effects or nuke?

06 June 2010, 03:03 PM
half is better for Nuke because it's floating point (Nuke does everything in 32-bit anyway). I don't know how AE handles 16-bit float but Photoshop opens them as 32-bit because it doesn't support half.

06 June 2010, 07:55 PM
yeah after effects reads it the same way as photoshop-- you have to switch to 32 bpc for it to read floating point properly. it's also incredibly slow, at least in CS4. if you have the option to use nuke, do it, it's way faster with float files.

06 June 2010, 10:14 AM
what exactly means "floating point with half precision" ?

06 June 2010, 10:28 AM

There is a standard for representing floating point numbers in computing, this wikipedia page should explain it.

06 June 2010, 07:21 AM
cheers, looks complicated.

06 June 2010, 12:56 PM
thanks guys
Nuke it is :)

06 June 2010, 02:09 PM
what exactly means "floating point with half precision" ?

Someone correct me if I'm wrong but I think this just means fewer decimal places for the float value, meaning it's not as precise as a full 32-bit float value.

CGTalk Moderation
06 June 2010, 02:09 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.