Digital BW, The Print

Steve, you wrote:

> > > Hmmm... I'm confused, don't know Profile Prism at all. Its
a
> > color
> > > profiling app right? You're suggesting he make that capable
of
> > > generating greyscale ICCs like the new Cone system, and
then
> > using
> > > them in Qimage? Sounds interesting!
> > >
> > > Steve Karafyllakis
> >
> > Steve,
> >
> > But is most likely not delivering what you want. So far even
the
> > cheaper Spectrometers (SpectroCam etc) can't measure beyond
1.7 D
> > accurately and people involved in B&W printing get secondhand
> > densitometers instead to measure their greyscales for curve
> > control. I'm afraid scanners are even more limited on that
> > aspect.
> >
> > The best thing for quad and hexatone printing with Qimage
would
> > be a port of Gimp-print (+ CUPS ?) to Windows. That gives
better
> > custom paper setting controls and curves. When greyscale
profiles
> > can be created they can be used in Qimage, till then
greyscale
> > curves can be applied in Gimp-print. Linux and OS X have an
> > advantage now in B&W printing (if it has to be done cheap:-).
> >
> > Ernst
>
> I agree that tryimg to take it all the way to creating ICCs is
a bit
> of a leap; I'm looking for relatively easy additions to make
Qimage
> useful enough to more frequently take advantage of its
remarkable
> interpolation and sharpening abilities.
>
> So far I have on my list-
>
> A) full .psd layered file compatibility
>
> B) Setting up Profile Prism to generate greyscale ICCs which we
> could then use in Qimage
>
> C)Ability to load Photoshop .acv curves in the curves dialogue
box;
> that would allow partitioned RGB workflow without having to
flatten
> and save a copy.
>
> d) The ability to save an interpolated/ sharpened image to use
with
> other workflows/RIPS; To me this alone would justify the price-
this
> is much faster and more effective than that over-rated piece of
> *##*## Genuine Fractals.
>
> I have to confess at this point that I don't know the program
well
> enough to know if there's some not too obvious way to already
do
> some of these things, such as saving a processed image. If
anyone
> does know it well enough, please jump in.
>
> Any further ideas?

Possible already:
Qimage could be used with quad or hexatone RGB curves that
already exist for PS or Gimp-print use (there may be more than
that meanwhile, so much is happening already). The curves have to
be converted if the curve tool is different between those
applications.
MIS www.inksupply.com
have several in their download section.

That will deliver similar quality to what Photoshop does: Epson
driver paper settings + resolution + and in PS the curves to get
the grey inks in the CMYK/CcMmYK inklines in line with one
another. Qimage can of course add its nesting and imposition
features etc
but I'm not sure what the Interpolation routines will do on B&W,
think that will be trial and error time. On the other hand it
might need a special interpolation and it is more likely to
appear in Qimage than in another application  ;-)

Next step:
What some people already do and what isn't so widely known on the
B&W list is that after the individual RGB curves are applied a
general stepwedge is printed from 0-100 % and the steps measured
with a densitometer, measurements converted to PS for the master
curve and by that the unevenness in the greyscale caused by the
rough individual RGB curves + the inconsistency of the printer +
the
points where the separate grey inks overlap are compensated to
bring it in one line. More or less a linearisation on top instead
of a linearisation at the bottom as found in RIP software. It
would be nice if a tool was added to Qimage that would allow the
conversion of densitometer data to the master curve or a
conversion from an Excel files that contain that data. AFAIK
Excel is used now for the PS curves. Maybe 17 nodes in the curve
tool of Qimage could be upgraded to more if that helps in
refining the tone corrections, don't know whether that is
necessary.

Another step:
The same people that are using densitometers seem to use Doctor
Pro in PhotoShop to convert PS curves to RGB ICC printer
profiles,
possibly CMYK printer profiles. It is a profile editor using the
colour engine + tools of Photoshop and the action feature.
Profile Prism doesn't have an editor as far as I can see on the
site. I don't know whether an addition to Profile Prism to make
that possible is better than a standalone program that converts
densitometer readings directly into an ICC profile for
linearisation or through a curves stage. The last would be nicer
as it could incorporate the greyscale "separation" curves as well
in the profile, those curves can't be made with an instrument but
are the result of trial and error work. The "greyscale" profiles
don't need the extra bagage of colour info.

Yet another step:
This workflow still relies on the Epson (Canon is another
candidate) driver. That driver is RGB based, has usually limited
paper settings and doesn't allow ink limitation and/or
linearisation of the individual CMYK channels, custom paper
settings can't be added and after 2 to 3 years the driver updates
for new Window OSses are not appearing anymore. As often older
printers are converted to B&W use this can be a problem.
Linearisation of the CMYK channels isn't that important in B&W
printing if a general linearisation on top is possible. I would
be nice however to get more control of the printer driver in all
aspects mentioned here. That's what is done in RIP's (Wasatch
etc). To big a task to undertake for the many models available
and appearing + the Windows changes along the road. The only
option that may be possible is a port of Gimp-print parts etc to
Windows. What is done on that port now seems to progress slowly.
On the other hand there are already B&W adaptions of Gimp-print,
CUPS running on Linux and OS X so part of the work mentioned
above is already done on the B&W aspect.
Qimage with Gimp-print as a driver comes very near to a RIP in
almost any aspect but its bias to RGB profiling. There's a trend
in RIPs to add RGB profiling on top of CMYK profiling so whether
this is con or a pro has to be seen. That it could have B&W
greyscale management on top would make it very attractive.

Much of what I've written is common knowledge on the B&W list and
will be familiar to you but not for other readers (if there are
any).

Implanting the ideas is the hard work and I can't be of much help
then so writing this gives me more a feeling of guilt than of
satisfaction.

Ernst