On Aug 5, 2004, at 5:21 PM, Belinda Beckinsale wrote:
If an imposition contains some pages that have been
set using
traditional gray balance (eg 75 65 65 90 etc) and some pages that have
been done using a profile such as the ISO web coated that uses a very
different gray balance, with a lot less yellow, then how is the
printer meant to handle balancing two very different gray balances in
the same sheet?
The problem is being overstated. Downloading the "ISOwebcoated.icc"
profile, Paper type 3, gloss coated web (LWC), 150 lpi (60/cm),
FOGRA28L, from the ECI web site, I get the following conversions for
L*50
SWOPv2, RelCol+BPC, 55, 46, 46, 11
ISOWeb, RelCol+BPC, 44, 35, 34, 28
ISOWeb, Perceptual 41, 32, 32, 23
For L*25 I get:
SWOPv2, RelCol+BPC, 68, 62, 61, 51
ISOWeb, RelCol+BPC, 60, 49, 46, 62
ISOWeb, Perceptual 56, 45, 43, 52
Now when I take 100% yellow, and ask the SWOP v2 profile, and ISO Web
Coated profile what the LAB values are (AbsCol),
I get:
SWOP v2 84, -6, 83
ISO Web 86, -3, 93
That is a huge difference numerically. Simple delta E is 10. Delta E
2000 predicts delta E 3 because it does a better job of understanding
humans aren't that sensitive to saturated colors, in particular yellow.
Still a real delta E of 3 is noticeable, therefore it makes sense the
separation is going to be different because an integral primary is
different. The other primaries are no where near this different.
So the answers to the question:
1. There is no traditional gray balance.
2. If a printer is receiving jobs with different gray balance, one of
them is wrong, and he will have to either make the right one look right
(good), the wrong one look good (bad), or compromise (maybe good or bad
depending on context).
Moral of the story is that your separations, including gray balance
need to be targeted for the final output process. That means one set of
separations for U.S. publications and another set for Euro. You need
them anyway because the TVI difference is more radical than the 4 point
discrepancy in shadow gray balance for yellow compared to magenta.
This may not be a problem with ISO or SWOP profiles
that have GCR and
aren't really relying on CMY to make neutrals, but what about when the
idea of tailor making your own profiles to suit your press really
takes off, and different separations and black generations get used?
If the measurement data is correct for the printing conditions being
used, even irrespective of paper white, then GCR is not going to
dramatically affect this if the profile is built correctly. The profile
building application should build the proper CMY balance to produce a
neutral regardless of paper white.
Is there perhaps a case for a gray balanced CMYK
'working space' like
there is in RGB? (ie work is done in this space and then converted to
the appropriate press space). This would probably cause less arguments
with scanner operators. Lets face it, the ideal RGB workflow is still
not a workable reality for a lot of pre-press.
I don't see how this could work. There are standard printing conditions
already, and profiles based on their measurement data are gray
balanced. That each one has different gray balance is due to different
inksets being used, primarily, as in this example.
Also, there was a comment earlier that you don't
want to set your scan
range in RGB from 0 to 255 for the same reason that you would not set
it to 0 to 100 for CMYK. I'm just wondering if this is unecessarily
limiting the range of the scan. You don't actually print the RGB scan,
it gets converted to a CMYK profile which would then handle what can
actually be printed from that data (if that makes sense). Maybe I
haven't fully thought through the logic of this, but any yays or nays
would be appreciated.
There are two possibilities. A dumb scanner setup should be set
conservatively so you aren't clipping shadow or highlight detail. This
is the one size fits all approach, constantly scanning with the same
settings. Yes you do lose a few levels with this strategy. Someone is
going to have to set white and black later, in Photoshop for example.
So there is no free lunch here unless you have more skilled Photoshop
operators than you do scanner operators (that is the trend).
The other approach is to set black point and white point in the scanner
software on an image by image basis. This maximizes the ability of the
scanner better, but clearly is more time consuming up front.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
---------------------------------------------------------
Co-author "Real World Color Management"
Published by PeachPit Press (ISBN 0-201-77340-6)