Yahoo Groups archive

Digital BW, The Print

Index last updated: 2026-04-28 22:56 UTC

Message

Re: [Digital BW] Questions about Using Color Settings and Workspace Profiles for Grayscale Printing

2006-02-24 by Steve Kale

> From: G Guhan Gunaratnam <guruguhan@...>
> Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Date: Fri, 24 Feb 2006 11:27:55 -0500
> To: "DigitalBlackandWhiteThePrint@yahoogroups.com"
> <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Conversation: Questions about Using Color Settings and Workspace Profiles for
> Grayscale Printing
> Subject: [Digital BW] Questions about Using Color Settings and Workspace
> Profiles for Grayscale Printing
> 
> Hello,
> 
> I've just read Clayton Jones' article Using Color Settings
> and Workspace Profiles for Grayscale Printing, and I'm not getting it.  I
> understand the points about dot gain, gamma, using curves, etc.  What I
> don't understand is the WYSIWYG thing.  If my monitor is calibrated (lets
> say with an EyeOne Display), and I have a profile then for it (this would be
> considered the front end...correct?).  And lets say I'm using EEM on an
> R2400, epson driver.  I have WYSIWYG don't I?  Doesn't the epson driver take
> care of that?

The strict answer to each of your two questions here is no.  The last one
first:  the Epson driver does not participate in the management of the look
of your image on your display.  This of course is what determines whether
you achieve WYSIWYG.

You are right that the starting point for good soft proofing begins with a
well calibrated display.  If that's not up to par then who knows what you're
looking at.  Photoshop provides a mechanism for soft proofing printer output
on our well-calibrated displays by using the ICC output profile which
"profiles" our printer's output.  This is all well and good for colour work
when we have an ICC profile of the printer's colour performance.

An issue arises when we turn to B&W workflows which have often not used
colour management.  Many B&W workflows reject colour management by using
"Same as Source" or "No Color Management" in Photoshop's print menu.  At
that point the ICC profile embedded in your image file - that dictionary
which gives all those pixel values meaning - get tossed out the window.
With such workflows there is no direct management of colour or luminance
from the file's document space (Adobe RGB, GG 2.2 or whatever) to the gamut
of the printer (for given settings).  The file's pixel values are sent
unadjusted for the response of the printer to each of those numbers.  So any
given pixel value for, say, GG2.2 your printer may churn out a colour on
paper that is nothing like what that pixel is in the GG2.2 space.  Also, if
you haven't been able to profile the output of your printer in B&W mode then
you can't set up a soft proof of that output in Photoshop.

When faced with this many people go to all sorts of lengths to try to remedy
the situation.  For example, they try different workspaces.  Remember that a
colour or shade of grey that looks identical in a new workspace has a
different pixel value than before and the hope is that when that new pixel
value is sent to the printer the printer generates a colour closer to the
one viewed on screen.  Other people try messing with their display -
effectively undoing any colour management from file to display.

It would be far simpler and sensible if we could deploy colour management
rather than abandoning it.  It's quite a helpful tool - not perfect, but not
enormously wrong either.  To deploy colour management you need to be able to
profile the response of the printer to possible stimulus, ie to possible
pixel values that are sent to it.  This involves, as you likely know (and
just as you do in profiling your display), sending it a set of sample
stimuli, measuring the response and from those samples building a picture or
profile of its output.  Up until recently this has been difficult for B&W
because the profiling packages out there that do this sort of thing are
designed for colour work.  In B&W we often work with single channel images,
greyscale without colour information, and use settings in the driver (eg the
hue picker in Epson Adv B&W) or RIP (eg ink curve selection in QTR) to
manage hue.  Profiling packages are for colour and colour files.  You might
imagine the profiling system "wondering" why neutral grey was printed, say,
with a warm tone and in effect trying to find a colour number (it only works
with numbers) which actually produces a neutral grey instead of a warm one.
If the entire printer response is warm there exists no number that can be
sent which prints neutral!  So basically it all breaks down.

Roy Harrington, though, wrote a nifty little piece of software which fixes
this situation.  All ICC profiles contain two sets of data for each
rendering intent.  One set is used "outbound" to remap the pixel values in a
file to new numbers which then cause the printer to print the right colour
(subject to the treatment of out-of-gamut colours).  The other "inbound" set
is used for soft proofing to the monitor by, say PS's soft proofing
mechanism.  With B&W workflows which use printer or RIP settings to
determine colour, we want an ICC profile which is just single channel.  We
want to manage luminance/density but leave the colour alone.  In other words
we want to use colour management technology to manage the inevitable
compression of luminance from a perfect scenario (the file) to an imperfect
scenario of ink black and paper white.  But we'd also like full colour soft
proofing and not just a soft proof of this luminance management.  Roy's QTR
Create ICC is used to write an ICC profile (not really to official specs)
which only has luminance management in the outbound leg but full colour
information in the inbound soft proofing leg.  In essence we get the best of
both worlds.  We can use a more B&W oriented driver (eg Adv B&W) or RIP but
still use colour management for luminance compression and soft proofing.

The problem, of course, is that you need to be able to measure a test target
printed by your printer which requires either a densitometer (in which case
you'll be able to measure the information required for luminance management
but not the information required for colour soft proofing) or a
spectrophotometer, or you need to get a profile from someone who has does
this work already.  And of course you have to pay Roy's token $50 shareware
fee.  There are enough people out there with the necessary equipment to make
these profiles that we should be able to set up some sort of library for
those who have paid the shareware fee but who don't have the equipment to
make the profiles. Unfortunately I don't have a 2400.  I'm certainly happy,
though, to share my 4800 Adv B&W ICC profiles (made with QTR Create ICC) but
of course these are only valid for the 4800 (and I run Eboni ink instead of
Epson MK).

> 
> Is this for times when one can't get a pre-made profile?  If I use a
> commercial RIP, is this issue also a null point?  (I thought that
> calibrating my monitor and using a RIP was to get WYSIWYG).  If lets say Im
> using ImagePrint, I have my image as I want in Photoshop, use ImagePrint to
> select my ink, paper combo, and print.  If I change my paper type, I make
> the change in ImagePrint, and the image will print the same on the new
> paper.  If I'm unsatisfied with contrast/brightness/etc I can adjust in
> photoshop, see the effect on screen, print again through ImagePrint, and see
> the same changes as on screen.  Is this not how it works?
> 
> Sorry, for the stupid questions, just a bit confused.

No question is ever stupid.  Not asking then is stupid.
> 
> 
> Thanks!

Attachments

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.