Digital BW, The Print

With QTR you can just have pure black at 100%.  You can also control the ink
limit at that level.  I do think there is an issue with Gimp's dither not
really working at 100% ink load (at least on my 2100) but you need to be
careful making a comparison with the Epson driver.  You don't know what %
ink the Epson driver lays down at 100% black - it is not visible to you nor
definable by you.  It could be inherently regulated at 90%, for example.

In any case, I posted a while ago the results of my QTR ICC workflow, Epson
"colour" driver and Epson BO driver comparison.  They can be found here:

http://homepage.mac.com/stevekale/stevekale2/FileSharing37.html

I was using the canned QTR profiles for Epson UC inks, an Epson 2100 and
matte paper with MIS' UC equivalent inks (so eboni in the K slot).  The
Epson colour profile was my Eye One-generated EEM profile which was profiled
after spraying the test chart with Lascaux fixitive.  The QTR curve was
linearised after spraying with Lascaux fixative. I did not spray for this
test - all results are without spray.  QTR and the Epson colour driver
achieve identical dMax (for all intents and purposes) and BO exceeds them by
a small margin.  BO has the poorest UNADJUSTED (no print curve) reproduction
of the step wedge image, especially when the gamma is set at 1.8 which is
not surprising as it is way off the gamma of Lab (the step wedge) which is
best approximated by 2.2.  What I was a bit surprised about was the extent
of difference between the Epson colour profile and QTR - remember we are
just measuring density or L(uminance) here.  Both QTR and Epson colour used
perceptual rendering intent to map to a print space with the same paper
white and ink black.  If the Epson profile were linear I would expect the
two to be right on top of each other - although I was using the canned QTR
profile and not one linearised for my printer and that will be a
contributing factor.  Note that QTR produced the most accurate
representation of the midtones in the test.


> From: Clayton Jones <cj@...>

> 
> The blackest possible print area can only be achieved when there is
> 100% coverage by only the K ink.  Many systems that blend inks often
> mix something else along with the K, even in the lowest zones.  This
> effectively reduces the dmax.  I see this with UT7 using sliders,
> where it can't match BO dmax.  So you need to adjust the curves to
> ensure there is only K down there.  However, I think there is more to
> it than just eliminating other inks.
> 
> In a BO print the visible paper white specks disappear at around 95%,
> yet the Epson driver is still able to achieve gradations between 95
> and 100%.  I assume it's because even though the dots are overlapping
> at that point, the dot centers are continually spaced closer together
> as it approaches 100%.  This means that the amount of ink laid down
> increases and affects visual density, even though there is full
> coverage of the paper at 95%.
> 
> So when adjusting blending curves, seems like the challenge would be
> not only when to eliminate other inks, but also how to achieve the
> continuous gradation from that point on with only K ink.
> 
> As to where to cut out the other inks, seems to me that any dilution
> of pure K is going to weaken the visual affect, so eliminate other
> inks at the earliest possible place.  Since the grainy look of a BO
> print fully disappears around 80 or 85%, that should be the ideal
> place. That would equal BO's dmax and beautiful intensity in the lower
> zones, one of its strong points, yet still have a non-grainy print -
> the best of both worlds.  This of course assumes the ability to get
> the pure K gradation like the Epson driver does.
> 
> I would be very interested to hear from QTR/IJC users if this is
> possible.  Is anyone doing this?
> 
> Regards,
> Clayton
>