Re: use of sRGB as a default
Re: use of sRGB as a default
- Subject: Re: use of sRGB as a default
- From: John Zimmerer <email@hidden>
- Date: Mon, 21 Jun 2004 12:59:06 -0700
Chris,
You're happy with images that look consistently bad, we're happy with
images that look consistently good. Stalemate.
JZ
On Jun 21, 2004, at 7:44 AM, Chris Murphy wrote:
On Jun 21, 2004, at 12:14 AM, John Zimmerer wrote:
Chris,
The issue isn't gamut size, it's where certain hues land inside the
gamuts. Find an image, like the red bridge reference image, with a
vivid blue sky in it, duplicate the image, then assign Generic RGB to
one and sRGB to the other. You'll notice the sky is much more purple
in sRGB than in Generic RGB.
All you've demonstrated is that Generic RGB is a more appropriate
source profile for this particular image than sRGB. I'm sure I could
dig up a whole bunch of ColorMatch RGB images, perform the same "test"
and come up with the same "results." How was this image originally
created? What platform? What application? Was the display calibrated
and profiled? Is it by nature a tagged or an untagged image?
I will bet you money that image, untagged, displayed on 10 machines
with a display gamma of 2.2 (linear LUT gamma), you'd see an image
with a sky that is on the purple side. You are attempting to achieve
color correction through the reassigning of profiles.
I still fail to find any of the explanations I've read even remotely
compelling as to why Generic RGB is better for untagged images than
sRGB - except to compensate for the antiquated legacy display gamma on
the Macintosh. Certainly it would not be amusing for end users working
in a non-color managed application to see an image (with display gamma
1.8, uncompensated), and the same image in Preview assuming sRGB as
the source profile. There'd be a mismatch. It also explains the
obsession with wanted tagged images, for how else could a Macintosh
know how to display the image correctly? In a world where sRGB can be
safely assumed as source for both external images and locally created
images, there is only one real valid assumption. But the legacy
display gamma on the Macintosh creates a major quandary because there
are two possibilities for assumption - if it came from a Macintosh
then chances are Generic RGB is a better source to assume. But if it
came from the vast outside world, then chances are sRGB is a better
source.
Therefore I stand by my current opinion, open to change at any point I
hear a compelling alternative explanation, that the fallout of Apple's
strategy to date encourages arbitrary practice of color management.
There is nothing Apple can do to resolve the ambiguity of image gamma
for untagged images without eliminating one of the possibilities such
that there is only one rational possibility. It's clear that
eliminating a gamma 2.2 world isn't going to happen, and would be a
whole lot easier to get rid of the gamma 1.8 world. Clutching onto
that relic aids no one, and hurts many.
Suffice to say, for Safari to assume sRGB as source for untagged
images would be a good thing if not well overdue considering that's
how web browsers are expected to behave per the W3C.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
---------------------------------------------------------
Co-author "Real World Color Management"
Published by PeachPit Press (ISBN 0-201-77340-6)
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.