Re: Profiling Cinema
Re: Profiling Cinema
- Subject: Re: Profiling Cinema
- From: Steve Upton <email@hidden>
- Date: Sun, 11 Nov 2001 13:49:58 -0800
At 8:03 AM -0500 11/11/01, email@hidden wrote:
In a message dated 11/9/01 3:29:48 AM, email@hidden writes:
Is it necessary to manage the curves in such a way? I appreciate that
the LCD's respond differently than a CRT but if anything are they
probably not fairly linear? Would that not be something we would want
>to change to 2.2?
How liniar they are depends on your definition of that term.
Yes it would. I meant linear from a input->output constant variance
sense. Like when you linearize an image setter so when you ask for
50% you get 50%. Or from a CRT calibrated to 1.0 so when you send
RGB=128, you get 50 light output (not, probably 50% gray as we see
light intensity through a logarithmic correction)
They seem to
respond well to an R=G=B gray balance if thats what you mean...
No, not really. You can get gray balance from linear or
curve-calibration if they are done well.
LCDs do not, however, have nice smooth gamma shaped curves,
varying from slightly off shape curves to ones with zig zag angles and near
reversals in them.
OK cool, this is what I meant. It seems that unless we are using
LUT-based profiles to correct for this non-linearity, then we'd
better let the calibration curves in the graphics card smooth these
out and bring them into the 1.8 or 2.2 gamma we desire.
So for all LCDs except Apple's you do indeed want to bring
the curves to 2.2 (or 1.8 or some such) but its still interesting to see what
the native gamma is, and occasionally useful to match it, for testing
purposes.
Sure, I'm all for testing and learning. It seems best though, for
normal workflow, to calibrate the display system to the desired
gamma, whether or not you choose to try to modify the white point or
leave it alone. (I support leaving it alone too)
Regards,
Steve
________________________________________________________________________
o Steve Upton CHROMiX www.chromix.com
o (hueman) 866.CHROMiX
o email@hidden 206.985.6837
________________________________________________________________________
--