Digital BW, The Print

Dirk,

>With regard to the UT80-C+3 curve that I adjusted for the Epson Pro Glossy
>paper, why does the blue channel (which controls the lightest grey on the
>1280) behave as it does? ...
> I understand why it comes down abruptly in the highlight area,

Yes, in the highlights the light gray ink is the main ink and, due to it's
lightness, must be pulled in steeply to have the necessary contrast.

>and then why it rises again to hold back light grey ink,

Once the cyan main gray inks take over the main image-forming function, the
light gray can be removed (reverse slope on the curve).  Not only are they
not necessary, but keeping them in there involves other issues.  First, the
light ink will dampen the paper more than is justified by the small amount
of density they'd contribute.

Second, since the drive imposes an absolute ink limit, if there is too much
light ink in there, the darker cyan ink will be limited in the extent to
which the driver will allow it to be put in.  Once the ink limit is hit,
putting in more light ink actually might lower the total density because the
driver will remove the dark ink to compensate for the greater amount of
light ink (and the driver is dumb; it doesn't know what the inks are).
Likewise, once the ink limit is hit, pulling light ink out may well cause
the driver to put in more of the dark cyan ink even though that curve has
not been touched, thus darkening the image.  This also means that in the
dark midtone areas, where the cyan is the main ink, use the red (cyan
controlling) curve to adjust density.  Use of the light ink (blue curve on
hextones) will be a frustrating exercise.  Once the cyan ink is pouring in,
use a few dots/points as possible to just stabilize the blue curve on it's
way back up.  You need to conserve the limited number of points on that
curve.

Finally, the blue curve (light gray hextone control) is pulled down in the
deep shadows.  Often the cyan and even the toner are almost full-on there.
As such, the blue curve becomes, a de facto black ink control curve.  While
the light ink does not contribute much to shadow density, the driver looks
at all three curves to calculate when to turn on the black ink.  So, if the
toner and cyan are all the way down, when the driver sees the blue curve
coming toward the black point it starts to pour in the black ink.

If the blue curve had just been left at 100% on after the highlights, it
would limit the cyan ink due to the total ink limit, and it would cause the
black ink to be started sooner, causing rougher shadow tones. On the best
printers this might not make much difference, but on the older ones the
black inks look rather rough in the image if they are started too soon.


>... ink loading seems to be an issue with the glossy papers,
>and the area with the most problems was between 60% and 80%.

The cyan ink is dark enough to carry the image there.  Of course, the toner
is needed if you are trying to get a neutral image.  On some RC papers,
getting enough toner in the shadows will be a problem.  As we've discussed
before, I'm not sure Pictorico can be tamed with the current UT toner mix.
If there is ever a really great RC paper, and it has this warm shadow area
problem I might darken the toners in a second UT ink setup.  However, it's
too soon to worry about that, and the current UT ink densities will, I'm
sure, be continued for most papers/printers.

By the way, the UT inks will be pre-loaded into 1160 and 1280 carts very
soon -- maybe this week.

Good luck manipulating the curves.  I wish more people would learn how to do
this.  If more of us customized the curves and made the results available to
all, the system would be that much more useful.

Paul
http://www.PaulRoark.com

_____________________________________
-----Original Message-----
From: Paul Roark [mailto:paul.roark@...]
Sent: Thursday, July 17, 2003 10:29 AM
To: dhobman@...
Subject: 1280, UT & Pro Glossy (was 1160 curves on the Pictorico (was
Ultra Tone Feedback))


Dirk,

>I concentrated on using the Ultratones on the 1280 with Epson Pro Glossy
>paper....

Good, I think that combination has a brighter future than the Pictorico and
1160 with respect to the UT inks.

>After much curve tweaking I was able to get rid of the ink puddling
>problem (very good news) and corrected the problem of muddy shadows.

Great.  I'd like to see the curve and give it a try.

> Now I can definitely see that the 1280 has much smoother gradations
>than the 1160.

Yes, it's a better machine.  I think those who want good RC prints may have
to go to a modern printer.

>The curve I started with for the adjustments is the UT80-C+3 curve.

>... Comparatively, papers like
>Epson Premium Luster, Premium Semigloss, Pro Glossy Paper, the Pictorico
>paper, etc. all have a base that tends towards a more yellowish-green tint.
>The extent of the color cast depends on the paper. ...

It does seem like there are problems with some of the bases, at least with
the UT inks.  I hope that the Epson UC inks react the same way as the UT
inks, because we depend on that large market to get the goods.  No one will
tailor their products to MIS UT inks.  That is one reason I've made a UT
version that uses mostly Epson inks -- just in case.  I might add that which
of the versions is best depends on the paper.  Both have bronzing and other
issues with RC papers.  The Epson inks are not clearly better than the MIS
inks.

>The UT80-C+3 curve made for a good starting point on the Epson Pro Glossy
>Paper. The extra amount of toner compensates for the warmer paper base cast
>of the Pro Glossy Paper. The result is a curve that looks like the UT80-N
>curve on EAM.

Sounds good.  With the 1160 and the Pictorico, even with the toner full on
the warmth was too much for me.  The 1280, of course, is such a better
printer that fixing the problem may be possible on that machine.  It can
pump more toner than the 1160 to neutralize the cyan.  Also, the black dots
are so much better that with the 1280 we could run the black ink up the
shadows a lot, and it might be less warm than the cyan.

>Another impression: getting these inks to work well on semigloss papers is
>going to be trickier than on matte paper...

Yes, I agree.  Add to that the differences among them, and it turns into a
real zoo.  If there was one clear RC winner, I'd be more inclined to put a
lot of resources into the problem, but I'm not sure there is one.  I'll have
to try the Pro Glossy with the 1280 and see how it looks.

> I could email you the curve ...

Please do.  I would like to try it.

>One thing you mentioned earlier - you found the Epson Pro Glossy Paper to
>have a weak dmax. I'm not seeing this at all. ...

I'll have to see how it looks.  The 1280 has a better dmax than the 1160, so
that could be part of the difference.  I'll just give it a try and see what
I get.

> I could eventually send you a print if I can get the 1280 to
>work with the Pro Glossy Paper to my satisfaction. I'm optimistic ...

I would like to see the results at some point.  I'll try it myself and see
what happens.

How is the bronzing?  That is another problem that needs to be fixed.  I'm
not sure but what that is up to the manufacturers, however.

>... it's still too warm.

If the results look promising, I'll take a shot at it also.

>I'm curious to try the new Oriental fiber based paper to be released. ...

Yes, I am too.


>Let me know what you think of all this, and whether it makes sense in terms
>of your experiences.

I'm very happy to see you getting into the curves.  When I started this
project my view was that lots of people would be making curves for different
papers.  Although the variable-tone approach and MIS's involvement have been
great, I have been disappointed at how reluctant most are to tweak those
curves.  If we could encourage more users to contribute to the effort, the
system would be much better.

So, thanks for your efforts.

Send the curve and I'll see if I can do anything to improve it.

Paul
___________________________






-----Original Message-----
From: Paul Roark [mailto:paul.roark@...]
Sent: Wednesday, July 16, 2003 12:40 PM
To: dhobman@...
Subject: RE: 1160 curves on the Pictorico (was Ultra Tone Feedback)


Dirk,

>You asked for a reminder about printing with the 1160 on the Pictorico
paper ...

Yes.  I had a little time before I left to do some experimenting.  One thing
I found is that the Pictorico paper seems to turn very warm in the shadows.
I'll take another look at it, but it appears to me there will be a
cross-over to warm there that is beyond what the toner can handle.  I might
be able to pull the black ink (which is not as warm) in sooner, but that
will cause dots in the shadows.  I might have some time this afternoon to
see what happens, but my initial conclusion is that the paper and the UT
inkset are not meant for each other.

>On another note, I've been experimenting with the UT inks on a 1280 as
well.
>I was wondering if my observations match your own and others, or if I am
>experiencing something different. On matte paper, a greyscale wedge prints
>perfectly. However, when I compare an image printed on the 1280 with the
>same one printed on the 1160, the 1160 print seems to be the superior
print.

That is the opposite of what I see with the 1160 and 1280 I have.  It may be
that there is just that much difference between the individual printers
within a model number.

>The primary problem I am seeing on the 1280 is a lack of separation in the
>deeper shadow values. Where shadow areas from an 1160 printer contain
>excellent detail and separation, the same areas on the print from the 1280
>become featureless patches of dark ink. This is quite visually disruptive.

The curves were written to the same standards using the X-Rite to measure
the densities.  With my 1160 and 1280, the tones and densities are the same
in the shadows.  The 1280, however, is smoother.  So, again, I suspect
individual printer differences account for this.  Where our individual
printers lie on the bell curves or performance is unknown to me.  The
feedback I have received from other 1280 owners seems to indicate that mine
is not that far off the norm, but I have not seen test strips from others to
be certain.

>The problem is particularly noticeable when one moves from matte paper to
>semigloss paper. I tried several semigloss papers, and all exhibited the
>same problem.

I have yet to find an RC paper that I think is up to the job.  From what I
can tell, the Epson UltraChrome printers are not free from the bronzing
problems either.  Once I hear that there is a paper -- perhaps the
Oriental -- that is really good, then I'll spend the time to try and write
curves that work with it.

For now, the cheap Epson Glossy Photo Paper seems to be the most free from
these surface artifacts, although its dmax at about 1.8 is not up to the 2+
that we'd like to see from these types of paper.  I also suspect it is far
from archival.

>The other noticeable problem I am seeing is that the 1280 prints on
>semigloss papers are exhibiting the ink puddling problem I was orginally
>seeing on the 1160 and that you corrected with the new curves. The problem
>is evident in the exact same part of the greyscale as was occurring with
the
>old curves on the 1160.

On the Epson Glossy Photo Paper, I don't see this with my 1280.  I have not
tried enough other RC papers to know.

In general, I would not be surprised if hextone printers are over-saturating
the paper.  They are putting a lot of light ink on the paper.  As I've noted
to MIS, if there is a great RC paper and I can't solve this problem with
curves, a new UT mix just for the newest 6 or 7 ink printers may be
justified.  However, the older hextones, including the 7000/9000 families
need the light inks, and with the current UT inks, these families of
printers make almost perfect fine art B&W display printers.  So, I do not
recommend any changes in the current UT mix.

If the puddling problem is too bad, it might be fixable with curves.  Since
I have not seen it yet and don't like any of the RC papers enough to put
much time into the problem, I'm just going to wait and see.

>It seems that prints coming off the 1280 should be technically superior to
>those from an 1160, but this does not seem to be my experience. ...

It may be that your 1280 is not working properly.  With the 1280 I have, the
prints are clearly superior to the 1160.  The puddling and too-dark shadows
may indicate that your machine is pumping out excess ink from one or more
jets.

Could you send me some samples -- matte (EEM) and RC -- so I could see what
they look like?  I'll try to see if I can get samples for some others also
to see if this is a common problem or one that is unique to your printer.

Thanks for the feedback.  I don't think anyone believes that today's
pigments and RC papers are really 100% compatible.  We're close, but
compared to what color dyes can do, the 2200 is still weak in some respects.
I've mixed "UT" inks from Epson UC inks, and the results are mixed.  On some
papers the UC inks are better, on some the MIS inks are better.  I think
pigs on glossy/RC have some problems that the industry has just not totally
solved.  As with color tints and shifts, bright colors may help hide the
artifacts, whereas the neutral B&Ws make them obvious.  For me and for now,
I'm sticking to matte paper.

Paul
________________________