Re: saturation rendering
Re: saturation rendering
- Subject: Re: saturation rendering
- From: Graeme Gill <email@hidden>
- Date: Sun, 17 Sep 2006 12:26:23 +1000
eugene appert wrote:
Am I understanding you correctly to be saying that , target colours
will be more saturated than source colours if this "gamut expansion" is
provided and the target space is larger ( even only in some areas) than
the source space?
That's a minimalist way of doing it. Colors could also be altered
in hue & lightness, to provide more room for expansion (trading off
hue/lightness accuracy for a better "shape" fit of the gamut hulls),
or it is also possible to change the saturation of colors within the
gamut, effectively applying an overall saturation enhancement at the
same time.
What I have gathered from your answers is that in theory, that is
according to the specifications, there are no specific provisions made
for increasing saturation, but only to "preserve" ( ie. maintain )
saturation levels as they existed in the source space. However it seems
that there is a consensus that, in practise saturation rendering does
increase saturation. The explanation given is that, the vendor, is
deciding to shape his product to increase chances of pleasing his
customer. like adding sugar to tomato sauce.
The specification is very general. As long as the intent favours saturation
over other attributes, it's a saturation intent. The four intents chosen
by the ICC are broad categories, and within those categories different
users under different circumstances will each have a specific idea of
what is optimal. Clearly having broad categories doesn't cater for
what is perfectly optimal in every situation.
What puzzles me is why graphic artists would rely on rendering intents
to correct their imagery , certainly editing software is a far more
precise tool to insure colours have been saturated to the edge of gamut,
and at the same time insuring that colours that have been left
deliberately less saturated would remain so, even for bar graphs I would
think that this would be important.
Right, ICC saturation intents are the wrong tool for the job in this
situation. ICC intents work rather well though at the "office user" level,
when they want to print something out on the corporate color printer,
like reports with charts, etc.
One of the problems really is that like most human senses, vision
is very sensitive to variations, not to absolute levels. Anyone
who's ever played around with enhancing a photograph knows
that if you compare a photo to that photo with some slight
contrast enhancement, the one that's contrast enhanced always
looks better. Taken to it's illogical conclusion, one ends up
with bitonal images. The same applies to saturation. The more
colorful image always looks better, to the point where they
look like posters painted in fluorescent paint. A result is
that a manufacturer can end up in an "enhancement" race, always
striving to make the images out of their device look "punchier"
that the competitors.
Graeme Gill.
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden