Re: CMY vs. K only Neutrality
Re: CMY vs. K only Neutrality
- Subject: Re: CMY vs. K only Neutrality
- From: Graeme Gill <email@hidden>
- Date: Mon, 03 Jun 2002 14:33:46 +1000
Scott Griswold wrote:
>
I am working with a simulation profile on the Bestcolor RIP printing to an
>
Epson 10000 and have been able to edit the profile so that the CMY values
>
are neutral, however, the areas that contain large amounts of K separate and
>
appear to be too Magenta. If I continue to pull Magenta out of the profile
>
then the CMY areas start to look too green. Is there a way to dictate how
>
the K separates?
Generally a profiling system that regenerates the black (typical of most)
have some sort of control labeled "UCR curve", "black start" or some
such. Any color profiling system that works properly should give
an output that is the correct color, irrespective of the level of
black used to compose the output. If the output color changes with the
black generation rule, then the profiling system isn't functioning
accurately.
Some packages that link two CMYK profiles together directly, allow the
black intent to be passed through the link, rather than regenerating
the black using a black separation rule.
Generally for more stable neutrals (ie. less dependent on
exact calibration, illuminant, observer, print aging etc.), then
maximizing the level of black helps (and once again, an accurate
profiling package will maintain control over the hue even when
maximum black is used), but on many inkjet printers the results are
not that attractive, giving "dotty" looking results in the highlights.
Higher resolution printers (such as the 10000) give less objectionable
results, and something like the new Epson X600 series can give really
good looking results due to their use of "light black".
If you're editing a profile by hand, then what is going on depends a lot
on what approach is being used to adjust the profile. A package could
let you alter the PCS (Profile Connection Space) values going into
the B2A table, giving you basically 3 number to play with at each point,
and no way to influence the amount of device black used.
Alternatively it could let you alter the device values out of
the B2A table, giving you 4 device numbers to play with at each point,
allowing total control (and responsibility) for setting the black values.
It might only let use alter the per-device value curves, giving you very
limited scope for altering the color.
>
What determines how K separates into neutral CMY? Is there a component
>
within a profile which handles this part of the separation?
The B2A table in a CMYK ICC profile is simply a table - conceptually for
any possible input color, the device values are simply looked up, so
the answer is that "the contents of the B2A table determine the K separation".
[In practice the table is sparse, and the in-between values are interpolated.]
>
How can I edit how K separates apart from how the CMY values are rendered.
>
Since K needs to become CMYK, this creates an interesting dilemma for
>
keeping things neutral. Has anyone else found a need for this?
Sorry, I can't help with such an application specific question. In my
experience it is better not to edit profiles by hand, but rather to
use a profiling package that works correctly in the first place. Such
a package will give accurate neutrals without any need for adjustment.
Software can juggle the number far faster and more accurately than a human
being can, and given that there are often between 5000 and 36000 entries in
a B2A table, it's not surprising that hand tweaking color profiles is often a
frustrating exercise.
Graeme Gill.
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.