Re: ScannerRGB to WorkingSpaceRGB (rendering intents)
Re: ScannerRGB to WorkingSpaceRGB (rendering intents)
- Subject: Re: ScannerRGB to WorkingSpaceRGB (rendering intents)
- From: Joel <email@hidden>
- Date: Wed, 11 Jul 2001 09:25:27 -0500
Fred Bunting wrote:
Perhaps ... although one has to be careful.
Currently, the amount of gamut compression (from perceptual rendering)
depends entirely on the relative sizes of the source and destination
spaces.
If the amount of gamut compression varies depending on the image
content, this can lead to unpredictable behavior. One solitary pixel
could concievably change the amount of gamut compression *dramatically*.
I believe Terry wrote:
The thing that irks me...
However, I agree it is a valuable consideration in any discussion
where one party recommends one intent over another to spell out the
faults and graces of both. Users are entering larger and larger
workspaces destined for smaller destinations and, in many cases, this
is where perceptual has its strengths. That one solitary pixal can
re-map all others is a fault, albeit an editable one.
For those users assigned the task to match colors and maintain
saturation and maintain brightness, relative situates colors in more
predictable fashion. That this situation is at the cost of out of
gamut colors is a fault, but one where choices can be made and
editing done. For images, the addition of BPC has added considerable
advantages to using relative which, in turn, leads one more step
towards a single intent for vector and image based projects. Another
ideal.
But I like ideals...
--
joel johnstone - designtype
Winnipeg Manitoba Canada
(Fuzzy in my logic. SCSI in Bus 0)