>I have a scanned image in RGB TIF format. I have converted it to a CMYK TIF
>and it looks great and prints great in Photoshop. Appearantly all colors
>are within the print gamut.
Obviously you have been following the calibration procedure in the User
Guide and you have the Photoshop "calibrated". If you compare the shadows
between the monitor and the print I bet you will find considerable
differences. But that is another thing.
Quote:>When I place the image in PageMaker or open it in any other graphics
>application it is deeply saturated and prints with the same unacceptable
This is because one gigantic mis-information from Adobe.
In the Photoshop (win) the "Monitor Parameters, Gamma" -setting in the
File/ColorSetting/MonitorSetup -dialog is *not* a monitor parameter at all.
When the calibration is done according to the User Guide then the result is
that the image-files that are created on Photoshop will have that value of
gamma compensation buried into them that is put into the "Monitor
Parameters, Gamma" input box.
The title of the box should be like: "Gamma Compensation That is Buried into
the Images Files Automatically".
On top of that Adobe recommends that one should enter a gamma value of 1.8
into that box. So, doing this results that the images will have a
compensation for gamma 1.8 buried into them. This happens when You believe
that you are only enhancing the images. Actually you are doing two things,
enhancing *and* compensating the image because the display is not linearly
When such images are then loaded into another application that is properly
(linearly) calibrated then of course the images looks very badly saturated
due to the fact that there is gamma compensation 1.8 buried into it.
Adobe recommends the gamma 1.8 as follows: "A target gamma 1.8 is
recommended for printing CMYK images, because it closely matches printer dot
gain", see User Guide, page 86.
So they must believe that there is CRT monitor tube inside printers because
the Gamma is the Transfer Property of such monitor tube. The gamma 1.8 (or
the inverse of it) is not close to anything. Typical monitor gamma is 2.5.
The printers have dot gain. The effect of dot gain is rather small (when
compared with all the other problems, bugs and mis-information) someting in
the range 5% to 15%. But printer manufacturers often make modifications to
the printer driver software so that the overall transfer curve of the
printer looks more like the transfer curve of a CRT monitor (but nowhere do
they matches it nor even come close). This is a rather desperate move from
them but printing with un-calibrated systems is eased by doing this, so
there will be less complains.
If you then have proceeded and made a full match (hue and intensity) between
the printer and display, including the dot gain adjustment in Photoshop then
the images will indeed be so twisted and curled as you reported. They are
good only for printing from Photoshop. It is easy to understand for example
that a dot gain adjustment of Photoshop is not the proper one for
de-compensating the transfer curve modification that the printer
BTW: Full monitor to print match is not good either. If you have done this
you can see (in separations) that you are using only some 50 to 80 levels on
the green channel on the monitor and equally low on blue. Now as the image
file still does have all the 256 levels for each of the three color, what
will happen? The 256 levels in the image file will be compressed for viewing
(bu Photshop). In case of green (that is the color that the eye is the most
sensitive for) there will be huge compression (256 levels into 50 or so) so
subtle hue- or intensity changes will be invisible, instead on the monitor
you will see here and there large areas in the image that have no hue- or
intensity changes at all. It would be much better to do the hue match and
then expand the intensity as much as the particular mapping allows. The eye
is very good in adapting for different intensity ranges (just think e.g.
about a print and a projected slide showing the same scene). But I think
this can not be done on Photoshop if it is calibrated according to the User
Guide, since it will end up with a non-linear "calibration" for Photoshop
and there are big troubles with hues on a system that has a non-linear
display + non-linear but not similarly non-linear printer + that use image
files that have some unknown mixture of both display- and printer
compensations buried into them.