2005-01-30 by Paul A. Yesnosky
> Because I've always done it this way, I don't really understand what
> QTR does with profiles (since I don't attach one). If an image has a
> profile attached, what does QTR do with the profile? If Mike got
> around to supporting grayscale, and we could tag the output file with
> an appropriate profile, what difference would this make? If I
> attached a colour profile to my RGB images, would QTR do anything
> with it?
Okay, here is the way I think about it.
When we "convert to profile" gray-matte, we are changing the color values of
the image to match what is required by QTR in order to get output that is
close to what we see on the screen.
But, we have no way of making sure these color values of the image stay
intact through Qimage. I have no problem with Qimage converting to RGB if
that is what is necessary within its color space, but what does it convert
the image color values to when it does this? And, after processing the
image, because the images is saved to a file as an untagged RGB, print
drivers and RIPs downstream don't know how to interpret the color values of
the image. They must just take whatever values are there and operate on
them.
If Qimage allowed at a minimum the ability to convert from its RGB color
work space back to gray-matte at a printer icc profile, then it would take
the color values it converted to when the file was loaded and convert it
back to "gray-matte", tag the image and save it to a file. This would be
the same a PS using gray-matte as its output profile.
Even if QTR does nothing with the attached profile itself, if the color
values of the image have been converted back to gray-matte by Qimage, the
color values would be appropriate for QTR (assuming the downstream system
was calibrated and linearized to this profile).
If this is not possible, then the real way to use the new gray-lab working
space in a PC environment is to load the image, work on it in gray-lab,
convert to gray-matte, process through Qimage to get the untagged RGB. Then
you need to send your image or reference file to QTR, look at or measure the
output and then tweak gray-matte until you get an output profile that is
calibrated to whatever Qimage does when it converts to untagged RGB and how
QTR processes that untagged RGB.
If Roy created gray-matte (and gray-photo) assuming you would take output
directly from PS using gray-matte as your printer ICC and fed that directly
to QTR (a la how you do on a MAC), they the gray-matte and gray-photo are
not exactly what us PC users who use Qimage need. It's fine for PC users
who do all processing in PS and save their TIF at exact print resolution in
PS because when they convert to gray-matte and save the file, they are doing
exactly what is done on the MAC. But, PC users who use Qimage I think need
a modified version of gray-matte/gray-photo or a change in Qimage to allow
it to convert back to gray-matte on saving of the TIF file.
(At least this is how I think it works...)
Paul