If I were a RIP designer who could work from scratch with all
the input over the years of lists like this I would look for a
printer calibration method that would consider the native
behaviour of the printer and what realistic standard was
possible with that behaviour plus at the same time check what
fitted color management on top of that standard best. If all
the perceptual curves of different papers + BPC etc result in
a space that isn't a too complex shape it must be possible to
find the average 3D curve within that shape and use that as
the calibration standard. Building the profiles on that should
theoretically allow a better fit from the native behaviour of
the printer up to the color management.
To give an analogy, if you have a metal cloth hanger with a
shape you like to alter you could first start to make the wire
frame straight and after that bend it to the shape you want.
Measuring is easier then. You could also look whether the
original shape has elements that come close to the desired
shape and just alter the differences. Less bending needed but
a more complex shape model to build on. Depends on the
original shape and the desired shape which method is best.
If I were a RIP designer who had a RIP working as nice as QTR
I would take care not to go on paths unknown. And Roy may even
know that those paths are actually dead ends. For me it is an
abstract model, Roy will now the numbers.
--
Ernst Dinkla
www.pigment-print.com
( unvollendet )Message
Re: [Digital BW] ICC v. Transfer Function in Epson driver
2005-10-24 by Ernst Dinkla
Attachments
- No local attachments were found for this message.