[argyllcms] Re: Profile generation with perceptual intent with compensation for paper white
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Fri, 22 Sep 2017 00:38:19 +1000
Henrik Olsen wrote:
On 21 Sep 2017, at 14.34, Graeme Gill <graeme@xxxxxxxxxxxxx> wrote:
sorry, that makes no sense. By default all the intents except absolute
colorimetric
make the paper perfectly white and neutral. You can easily check this by
feeding
L*a*b* 100,0,0 into a profiles B2A table, and you will get CMYK 0,0,0,0 out.
Can see the wording can be a little tricky. “Perfectly white and neutral” to
you means
no ink on paper, it seems. To me neutral means a* and b* is 0.
There's no such thing as absolute a* b* = 0, since L*a*b* space is
always relative to some white point. In ICC PCS space that is sometimes
defined to be D50 though, and in that context can be used as a pseudo
absolute measure.
But the usual situation with any viewer looking at a print is that
they are assumed to be, and in practice are very largely white point
adapted to the paper white. So no ink = white.
Only if you contrive some sort of viewing situation in which the
observer is not so adapted (partial or mixed adaptation), might
paper not look white. This is unlikely to correspond to how
an observer will see a print.
I print on inkjets with RGB interfaces, by the way.
But underneath it's printing CMYK. It's clearer what I mean when I can
talk to what inks are actually being laid down.
I don’t agree, or understand. If the paper is bluish, then the final print
has all my
colors skewed towards that bluish paper white, as in relative col.
It will only look blue at all if you introduce some other white point,
other than the paper into your field of view. This is not typical - usually
a print will be seen in isolation, or along side other images printed
on the same paper (i.e. a book).
Relative intent gives me relative colors, as in all colors are skewed to the
papers
whitepoint.
As it should be, since an observer will usually see all the colors relative to
the paper white.
Most available workflows are not intended to deal with a mixed white
point adapation situation, since they are not common.
Cheers,
Graeme Gill.
Other related posts:
