hi Igor,
yeah, sounds like somebody on the other end is a bit confused
and knows just enough to be dangerous.
as you say, saving EXR with log encoding hardly makes any sense
and only adds possibilities for errors. so you have two options:
try to explain to them that linear (float) EXR are just fine, or
just accept their premise, but make sure that you have the same
log decoding as their encoding since there are quite a few
variants by now (otherwise you'll risk negative black values and
other unpleasant stuff).
btw, applying a (cineon) log curve to linear data still wouldn't
necessarily get everything in the 0-1 range since extreme
highlights would still be over 1 value in float (or get clipped
to 1 in DPX).
bests
chris
On 3/20/17 at 5:26 PM, subscripti...@badgerfx.com (Igor
Majdandzic) wrote:
Hey,
A company wants that we supply exrs in log because it would
have more latitude to work with. Does that make sense? Using
the curve for log brings everything in 0-1 range, so nothing
would have been gained using exrs, right? I know both are
containers, but for me linear exrs or dpx in log would be the
way to go. Everything else seems weird and disorienting. Is
there one argument for log exrs, that I just don't know about.
Cheers,
Igor
_______________________________________________
Nuke-users mailing list
Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
