My COLORSYNC new year's wish
My COLORSYNC new year's wish
- Subject: My COLORSYNC new year's wish
- From: Roberto Michelena <email@hidden>
- Date: Wed, 02 Jan 2002 11:04:54 -0500
I do wish to see Color Management grow in 2002 ; and I feel there has been
no significant advance in the last years. A revision of the ICC standard,
finally, but not bringing a much needed push forward to the technology.
A rogue idea: the next ICC standard should make a profile into something
like a javabean; a portable cross-platform thing able to contain much more
than just data. I see that maybe Java is still considered somewhat slow, but
is the most viable option for anything portable and cross-platform. And
speed gets less of an issue with each iteration of processor wars.
Can you imagine that? No more finger-pointing between CMMs and profiles. The
CMM would disappear, the 'jProfile' being a snippet of code that receives
and outputs data. Allowed data formats could include spectral, since a
'jProfile' that handles it would take advantage of it, while one that
doesn't only need to convert it to LabD50 or whatever it consumes.
There would be much more room for improvement and differentiation between
profiling software, since the profile would contain their own CMM and gamut
mapping. In fact, a profile could just be a mathematical model with a few
data points, or be a zillion-point matrix not limited by the current ICC
spec limitations (such as even dimensional spacing).
And the OS's color management engine would be so simple, just a linker for
coordinating the interaction between profiles, that there would be no OS
without it. What's more, your result would be totally consistent regardless
of platform and software choice, dependant exclusively on the profiles
themselves.
We would lose one thing: the ability to peek into profiles and edit them
with software that is not the one which built them. But who knows, maybe an
'edit' API could be part of the spec so editing softwares could be built to
send edit commands to the jProfiles.
Sounds too crazy? All currently propietary Color Management softwares such
as Aurelon, Barco, etc. would have a standard to which they can adhere,
while current ICC-based offerings would have much more room to grow while
preserving propietary value. Profiles themselves would have much more value
than the data they were built from.
-- Roberto Michelena
EOS S.A.
Lima, Peru