Datacolor User to User Support Group.

-----Original Message-----
From: Tom <ttrostel@...>

>>Ok .... you have my deepest respect on handling Gamut mapping.  I've
gotten through making a color reference model simplistic handling of
in-gamut colors.  Handling out of gamut colors and actually making
them *look good* is a serious pain.  Sure you can just clip things or
slide them inward somehow but photographs look horrible if you do it
wrong.

Yes, its a challenging area, and no two users ever agree on just what 
the "right" result is...

>>Here's a question though.  Is CV working on / with the new Microsoft
color model?

It won't be available to users until 2007 (with more delays possible) 
and the feature set of the initial release is still an open question. 
Another question will be adoption; will Adobe choose to integrate OS 
level color management thats not cross-platform, and not compatable 
with Win2000 and XP? They don't choose to use OS level color management 
now, preferring to use their own Adobe Color Engine within their 
applications instead. If Adobe doesn't adopt it, then it will be 
seriously lacking in "critical mass"...

>>I heard that it is compatible with ICC profiles but they
have their own de-coupled engine (ICC profiles have some
pre-calculated parameters).

Windows Color System will use a "late binding" approach, where factors 
like illuminent will not be built into the profile as simplifying (but 
limiting) factors, but instead will apply these features later, more 
flexibly, but will more liklihood of varying results and user 
confusion. It will certainly be interesting to see how a more 
complicated color management system works out for end users, since the 
current ICC process, with its built in simplifications, is generally 
considered to be too complex already...

>>Since that gamut mapping is such a pain
in the butt and the CV routines do a great job with my photographs, if
we use the new de-coupled windows color system we loose that carefuly
calcuated mapping don't we?


If it is done entirely at the OS level, using their algorithms, then 
all flavors from current products will be lost, including ColorVision's 
mapping that you note. If it is done by color management companies, 
using their own tools in conjunction with MCS elements, then perhaps 
there would still be room for some of these unique factors... the 
answers are not clear yet. And of course, we are assured that running 
our current software with its current functions will be possible as 
well, though Microsoft would very much like to see everyone on their 
WCS bandwagon...


>>If the answer is under the Microsoft NDA I'll understand if no answer
is possible.

There's always the tightrope act between the NDA materials, and the 
announced stuff... what we are discussing here is, to the best of my 
knowledge public information;  I've seen Microsoft Color Team members 
talk about it in front of public audiences on multiple occasions, which 
would indicate it to be public to me...

>>Honestly I'm looking to preserve the quality of the
profiles I have now and am unsure about the Microsoft direction.

You are not alone in this... there is general confusion in the industry 
about this approach, its proprietary nature, its success, and the 
advantages it may or many not offer over advanced ICC version 4 
functions. Only time will tell.

C. David Tobie
Product Technology Manager
ColorVision, Inc.
CDTobie@...
www.colorvision.com