Re: Panther, sRGB, web browsers
Re: Panther, sRGB, web browsers
- Subject: Re: Panther, sRGB, web browsers
- From: Chris Murphy <email@hidden>
- Date: Sun, 14 Dec 2003 14:00:41 -0700
On Dec 14, 2003, at 1:37 PM, Jim Sims wrote:
If a system were to always assume sRGB...
what effect would there be on images which contain colors outside of
sRGB displayed on a device capable of displaying ECI-RGB or greater
range.
What display device exists like that? Even the new NEC-Mitsubishi that
supposedly has the Adobe RGB gamut, I would not want to assume Adobe
RGB for untagged images. I'd still want sRGB assumed. If I don't like
that, then I should embed some other profile into the image.
And as for output, I think we have a tendency to believe that the only
gamut mapping options are clipping or compression. Printer drivers that
assume sRGB do gamut expansion, and at least in one vendors case they
are using aspects of CIE CAM (97 version last I'd heard) to do some of
their gamut mapping.
What is the image creators intention when he is serving an untagged
image with a greater range of color than exists in sRGB.
Lunacy? If they want something better than sRGB, they need to specify
it and embed it into the image. We cannot get into a situation where we
are guessing what other people intended. Oops - we are already in that
situation. I think it's better to relegate such images to sRGB unless
otherwise specified.
If one assumes display technologies will improve in the future, and
not wish to restrict an image's display quality to current
technologies --what is the best methodology now...
Assuming sRGB for untagged images in no way limits display quality now
or in the future. All it does is it says "if this sucker has not been
defined any other way, I'm going to assume sRGB." And if at some point
in the future there is data to support most untagged images behave some
other way, it's a very simple thing to set that as the assumed source
space. The underlying architecture would make such a change a much
easier one than what we are currently contending with - huge numbers of
untagged RGB images, and an acceptance of it.
would it be to bring the image into a large space (ECI-RGB), turn off
color management, and "monitor proof"...
then save to a format that allows for including gamma information
(PNG)
I don't think I fully understand the question.
Anyone sophisticated enough to be contemplating ECI RGB as a source
space is sophisticated enough to find a suitable source profile first,
before converting it into ECI RGB for further editing and then ensuring
that profile gets embedded when the image is saved.
These kinds of issues get to be even more important when we consider
other file formats like PNG, PDF, and other media like video. Simply
having a bunch of untagged images assumed to be in the monitor's space
doesn't strike me has encouraging consistent color. It quite literally
guarantees inconsistent color. In the above case of new displays with
substantially different primaries and gamuts, the idea of assuming
monitor RGB for images very clearly is a problem.
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.