I have been seriously considering moving to Linux but find myself 
stymied by the primitive state of color management support.  Both at 
the system level and in the image acquisition, editing and printing 
tools.

I have a good color management workflow on Windows and once properly 
setup it is mostly automatic.  Even with all of this built in support 
for color management it is still a very large task to get this set up.  
Getting good device profiles being the hardest part - particularly 
printers.  The only semi-manual step in my Windows CM workflow is 
selecting the correct printer profile in Photoshop before sending the 
image to the print driver.  Since I almost always use the same printer 
setup and paper this setting will default to the correct profile most 
of the time.  I am using profiling tools other than LCMS (Profile 
Prism).   

I would like to understand the CM workflow for Linux with the tools 
currently available.  I have seen some information on creating and 
checking profiles but nothing on an over all workflow.   For example, 
how do I get the Xserver to use my monitor profiles?   In Windows I 
set this up with Adobe Gamma.   Where and how do I convert the image 
to my printer color space?  In Windows this is part of the Photoshop 
print with preview dialog.  Where and how do I setup my scanner 
drivers to imbed the scanner profile?  In Windows my scanner software 
(VueScan and SilverFast) do this automatically once it is setup in the 
software with a simple GUI dialog.  How do I setup a working color 
space in The GIMP? Again in Phototshop I simply go to the preferences 
dialog and select my working color space profile.

An example of my frustration.  When using The GIMP I can apply a ICC 
profile preview to the image much like in Photoshop. Presumably to see 
what the image will look like when printed.  But if Xserver is not 
using my custom monitor profile then this "preview" is meaningless.  

In addition, The GIMP, other than the preview filter, knows nothing 
about color management.   This means, at the very least, that an 
effective color management workflow is much more complex since color 
management operations must all be manually handled outside of the 
image editing application.  So there must be preprocessing steps to 
embed the input device profile in the image (since Xsane does not 
handle this - VueScan will do this if your device is supported), then 
convert the image to the display color space and then post processing 
steps to convert to the printer color space before printing.

How are you handling these in your workflow?  Have you automated any of 
these steps?  Is there a document somewhere that covers this?
-- 
Hal V. Engel

Attachment: pgptbZ0akosun.pgp
Description: signature

Reply via email to