Digital BW, The Print

Just a quick update on my use of Paul's QTR Eboni-3 curves for the 
R1800. First let me thank Paul for giving us R1800 owners another 
option to play with. 
Up until recently I have been printing with NK7 Splitone/QTR on HPR 
188, 8 x 10s. Very happy with this setup and the prints to my eye are 
fantastic (nice variable slightly warm tone and very smooth creamy 
tones (ie lack of grain)). The NK7 inks however are expensive and I am 
always on the look out for a cheaper alternative if I can maintain 
print quality. So the idea of using bulk Eboni ink and filling the 
cartridges my self to reduce cost was very tempting. 
Following MIS and Paul's instructions the filling of cartridges and 
installation in the R1800 printer was easy and I got a clean nozzle 
check after only one cleaning cycle - I am very impressed already!
The image I am printing is a landscape (tmax 4x5 readyload, developed 
in PMK, scanned with Epson 4990) with a dramatic dark sky and some 
delicate tones in foreground trees and rocks.
So what were the results?
The Eboni-3 setup with Paul's curve on HPR is tonally very pleasing 
(slightly warm) and prints slightly lighter overall than the NK7 setup. 
To the naked eye the dmax is comparable. Out of the box, with no 
tweaking the printed image is very close to what I see on my LCD Eizo 
monitor - again very impressed!
The downside that I am struggling with (and Paul explains this in his 
notes)is the increased grain in the mid-tones. The NK7 setup wins 
handsdown in this area but it is using 7 inks while the Eboni setup 
only uses 3. With my eye 1 foot from the 8x10 print I can clearly see 
the extra grain. However, if I stand 3 ft back the increased grain is 
not a significant problem. (Note: my wife, who does not have much 
experience in comparing B/W prints, could not detect the difference in 
grain)
I am now wondering if I try increasing the number of Eboni inks to 4 or 
more will this help to reduce the grain while keeping the current dmax? 
Or will this only work with lighter inks in the mix (like NK7)?
Adding color to smooth out the print is not something I want to try as 
the whole BO approach really appeals to me. 
Any comments  would be greatly appreciated.
Jim

Hi Jim,

>Just a quick update on my use of Paul's QTR Eboni-3 curves 
>for the R1800. ...


> The downside ... is the increased grain in the mid-tones...
> With my eye 1 foot from the 8x10 print I can clearly 
> see the extra grain. ...

> (Note: my wife, who does not have much experience in 
> comparing B/W prints, could not detect the difference 
> in grain)

That's the same reaction I'm getting.  We can see the difference, but most
viewers can't.  I even took samples to a group of fine art B&W printers at a
B&W gallery.  None of them saw the problems I was worried about.

So, I stopped worrying about it.  The fine grained structure is there in the
midtones, but, fortunately, it is not like the large, visible dots in the
highlights that I did find irritating with the older printers.  It looks to
me like a print from a fine-grained film, and as many silver printers found,
a sharp fine grain actually gives the illusion of more sharpness to a print.

>I am now wondering if I try increasing the number of Eboni inks 
>to 4 or more will this help to reduce the grain while keeping 
>the current dmax?

I played briefly with up to 6 inks and thought the optimum might be 4.  The
3-ink version was actually smoother than the 6 ink version.  I did not see a
significant difference in the grain structure with any of the versions
between 3 and 6.

Since I rarely print color, what I may do is put in 4 Eboni carts and 4
cleaning carts.  The idea of being able to simply swap positions (and
re-load) the minute there is a bad nozzle, thus avoiding cleaning cycles
(except for the re-loading/swapping) has some appeal to me.  I might even
swap positions with each re-load so that the channels stay cleaner and don't
wear out as fast.  Obviously, I'll have to keep track of which channels are
loaded and include in the profile name which channels are being used.
However, with QTR this is easy.  It should make a printer that will last a
long time with a minimum amount of maintenance.  The idea of only 2 bulk
bottles to mess with has appeal.

Back to the issue of smoothness, note how much smaller the highlight dots
are than the midtone graininess.  (See the magnified scans at page 2 of
http://www.paulroark.com/BW-Info/R1800.pdf )  I think, like with silver
grain, we're not seeing the individual dots, but a pattern -- perhaps close
to random -- of how the ink is distributed.  It makes me wonder if there is
a way to optimize the curves to avoid some of it.  So far, I have not found
such.  As with silver grain, there may be a random pattern that is
unavoidable.

At any rate, I must say I'm back to producing 11x14 prints for the gallery
(I sell matted, un-framed 11x14 versions in clear bags) that please me more
than any I've made recently, in part because I know they are the most stable
I can make -- and it's just plain easy.

Paul
www.PaulRoark.com

Paul Roark wrote:
> Hi Jim,

>> The downside ... is the increased grain in the mid-tones...
>> With my eye 1 foot from the 8x10 print I can clearly 
>> see the extra grain. ...
> 
>> (Note: my wife, who does not have much experience in 
>> comparing B/W prints, could not detect the difference 
>> in grain)
> 
> That's the same reaction I'm getting.  We can see the difference, but most
> viewers can't.  I even took samples to a group of fine art B&W printers at a
> B&W gallery.  None of them saw the problems I was worried about.
> 
> So, I stopped worrying about it.  The fine grained structure is there in the
> midtones, but, fortunately, it is not like the large, visible dots in the
> highlights that I did find irritating with the older printers.  It looks to
> me like a print from a fine-grained film, and as many silver printers found,
> a sharp fine grain actually gives the illusion of more sharpness to a print.
> 
>> I am now wondering if I try increasing the number of Eboni inks 
>> to 4 or more will this help to reduce the grain while keeping 
>> the current dmax?
> 
> I played briefly with up to 6 inks and thought the optimum might be 4.  The
> 3-ink version was actually smoother than the 6 ink version.  I did not see a
> significant difference in the grain structure with any of the versions
> between 3 and 6.

> Back to the issue of smoothness, note how much smaller the highlight dots
> are than the midtone graininess.  (See the magnified scans at page 2 of
> http://www.paulroark.com/BW-Info/R1800.pdf )  I think, like with silver
> grain, we're not seeing the individual dots, but a pattern -- perhaps close
> to random -- of how the ink is distributed.  It makes me wonder if there is
> a way to optimize the curves to avoid some of it.  So far, I have not found
> such.  As with silver grain, there may be a random pattern that is
> unavoidable.

Paul,


I still have some worries about the dotgain control with 
Multihead BO and the small dots. Your observation of the 
pattern fits that. With BO the midtone is 50% black dot on 
50% paper white, exactly where dotgain with conventional 
fine screens in offset have a problem. In that case at bit 
of dotgain influences the checker board formation at say 45 
% where it should happen at 50%. By using ellipse shaped 
dots the pattern formation can be regulated over a range 
where say at 40 % linear chains are formed at 45 degr. and 
at 60 % the the rest follows. That made it smoother and 
better to control. Later on stochastic screens followed but 
in a sense there still is that point where the dot pattern 
flips from black dots on white paper to white paper holes in 
a black layer. Paper coating consistency, humidity, speed of 
printing (smaller prints on a big machine) all influence 
that point and with black ink it is then yes or no.. There 
must be some literature on how to regulate that pattern 
formation but the consistency issue remains.


Met vriendelijke groeten,  Ernst


|  Dinkla Grafische Techniek  |
|     www.pigment-print.com    |
|             ( unvollendet )            |

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Ernst Dinkla <E.Dinkla@...> 
wrote:
>
> Paul Roark wrote:
> > Hi Jim,
> 
> >> The downside ... is the increased grain in the mid-tones...
> >> With my eye 1 foot from the 8x10 print I can clearly 
> >> see the extra grain. ...
> > 
> >> (Note: my wife, who does not have much experience in 
> >> comparing B/W prints, could not detect the difference 
> >> in grain)
> > 
> > That's the same reaction I'm getting.  We can see the difference, but most
> > viewers can't.  I even took samples to a group of fine art B&W printers at a
> > B&W gallery.  None of them saw the problems I was worried about.
> > 
> > So, I stopped worrying about it.  The fine grained structure is there in the
> > midtones, but, fortunately, it is not like the large, visible dots in the
> > highlights that I did find irritating with the older printers.  It looks to
> > me like a print from a fine-grained film, and as many silver printers found,
> > a sharp fine grain actually gives the illusion of more sharpness to a print.
> > 
> >> I am now wondering if I try increasing the number of Eboni inks 
> >> to 4 or more will this help to reduce the grain while keeping 
> >> the current dmax?
> > 
> > I played briefly with up to 6 inks and thought the optimum might be 4.  The
> > 3-ink version was actually smoother than the 6 ink version.  I did not see a
> > significant difference in the grain structure with any of the versions
> > between 3 and 6.
> 
> > Back to the issue of smoothness, note how much smaller the highlight dots
> > are than the midtone graininess.  (See the magnified scans at page 2 of
> > http://www.paulroark.com/BW-Info/R1800.pdf )  I think, like with silver
> > grain, we're not seeing the individual dots, but a pattern -- perhaps close
> > to random -- of how the ink is distributed.  It makes me wonder if there is
> > a way to optimize the curves to avoid some of it.  So far, I have not found
> > such.  As with silver grain, there may be a random pattern that is
> > unavoidable.
> 
> Paul,
> 
> 
> I still have some worries about the dotgain control with 
> Multihead BO and the small dots. Your observation of the 
> pattern fits that. With BO the midtone is 50% black dot on 
> 50% paper white, exactly where dotgain with conventional 
> fine screens in offset have a problem. In that case at bit 
> of dotgain influences the checker board formation at say 45 
> % where it should happen at 50%. By using ellipse shaped 
> dots the pattern formation can be regulated over a range 
> where say at 40 % linear chains are formed at 45 degr. and 
> at 60 % the the rest follows. That made it smoother and 
> better to control. Later on stochastic screens followed but 
> in a sense there still is that point where the dot pattern 
> flips from black dots on white paper to white paper holes in 
> a black layer. Paper coating consistency, humidity, speed of 
> printing (smaller prints on a big machine) all influence 
> that point and with black ink it is then yes or no.. There 
> must be some literature on how to regulate that pattern 
> formation but the consistency issue remains.
> 
> Met vriendelijke groeten,  Ernst
>

Ernst,

I'm not sure I understand your comments.  With inkjets there's no notion 
of dot shape -- they are just drops out of the piezo jet.  I agree that smaller
dots have slightly more dotgain but the linearization and/or gray ICC profile
measure the actual densities and compensate for any dotgain differences.

I have some of Paul prints and my feeling with them is more of an overall
sandpaper feeling if you get up close to it.   One effect that may be happening
in the midtones is that there are two pseudo random dithering patterns from
two inks overlaid.  The interaction is more truly random giving the possibility
of dot doubling/clumping on a minute scale.  (it's a little like moire patterns but 
there's no overall regularity).   It may be this randomness however that helps 
get rid of banding side effects.

Roy

Hi Paul, thanks for the reply.
I take it from your following comment: 

"I played briefly with up to 6 inks and thought the optimum might be 
4.  The 3-ink version was actually smoother than the 6 ink version.  
I did not see a significant difference in the grain structure with 
any of the versions between 3 and 6."

That I could spend alot of time playing with this and not see any 
major gains. Anyway, like you I am almost at the stage of excepting 
the grain as something I can live with. But before I give up persuing 
this, what about paper selection? I was not clear from your notes if 
you saw any major difference in smoothness depending upon paper 
selection. If the answer is yes, could you provide a list of paper 
types you have tested in order of smoothness rather than tone. 

I have ordered some Arches Watercolor Paper HP 140lb Bright White to 
give that a go. (mainly to compare uncoated paper with HPR)

Once again thanks for all your efforts.

Jim

> Ernst,
> 
> I'm not sure I understand your comments.  With inkjets there's no notion 
> of dot shape -- they are just drops out of the piezo jet.  I agree that smaller
> dots have slightly more dotgain but the linearization and/or gray ICC profile
> measure the actual densities and compensate for any dotgain differences.
> 
> I have some of Paul prints and my feeling with them is more of an overall
> sandpaper feeling if you get up close to it.   One effect that may be happening
> in the midtones is that there are two pseudo random dithering patterns from
> two inks overlaid.  The interaction is more truly random giving the possibility
> of dot doubling/clumping on a minute scale.  (it's a little like moire patterns but 
> there's no overall regularity).   It may be this randomness however that helps 
> get rid of banding side effects.
> 
> Roy

Roy,

whatever dot distribution you select random or ordered 
there's a point with BO only where half the paper is black 
and half is white. To reduce the dotgain effect at that 
point to a minimum one would prefer a line screen of black 
and white lines and the lines as thick as possible. Horrible 
for detail and a very dominant pattern. Any compromise to 
get more detail and less visibility in the pattern results 
in smaller dots and  more complicated patterns including 
random dot distribution. That will increase the influence of 
dotgain whatever pattern is selected and around 50% the 
patterns are sensitive to that influence, little dotgain has 
  a lot of influence then. Profiling should deal with that 
but climate control in the list members houses still gives 
variables. Another variable is the paper coating per batch, 
the printer's condition and printing smaller prints on an A3 
printer speeds up the printing while the drying time isn't 
increased. Individual printers vary so a general profile 
will lead to different  results. In quad ink sets much of 
this tone effect is dampened by the diluted inks, it will 
not be like that in BO printing whether with one or multi heads.

I think that with BO more attention has to be given to 
regular linearising of the ink set. Multihead BO suppresses 
the effects of  banding, nozzle defects etc but isn't an 
answer to dotgain, it actually will increase the influence 
of dotgain as it forces the heads to the use of the smallest 
droplet for the resolution selected. Expect a lot of 
different opinions on what list members get from their 
printers on BO, not just a matter of taste for BO but also 
real differences in the prints with everything else being equal.

It isn't that I'm against the method but when I wrote about 
it 4 years ago I immediately realised that it could have 
that negative effect too. I was also a bit worried about the 
"CMYK" dot distribution that might not be the best KKK dot 
distribution. Related to that dotgain effect. I'm curious 
what practice will tell us, few users right now.

Met vriendelijke groeten,  Ernst


|  Dinkla Grafische Techniek  |
|     www.pigment-print.com    |
|             ( unvollendet )            |

Hi Jim,

>... what about paper selection? I was not clear from your
> notes if you saw any major difference in smoothness depending 
>upon paper selection. If the answer is yes, could you 
>provide a list of paper types you have tested in order 
>of smoothness rather than tone. 

Yes, paper affects smoothness.  Start with Epson Premier Art Scrapbook
(great deal at Atlex), or the Premier Art 205 (essentially if not actually
the same paper).

The Moab papers seem good.  190 may be better than 300.

The 325 Premier Art Hot Press is too thick.  The 270 might be the optimum
combination of weight and good dmax/tone. (Premier Hot Press papers have
higher dmax and more even tones above the 205 thickness, but the 325 did not
print as smoothly.)

In general, (with the possible exception of the PA 325, graph which I should
replace with the 270 chart) the papers I listed printed very well, at least
with the caveat that the printer does better if it has enough paper margin
top and bottom to grab onto.

Many papers/coatings print with that cool-warm split tone.  That's fine for
some images, but for a "cool" or more neutral look I either like an even
distribution.

My Moab papers are a bit old, but if their coating has not changed (and if
they have finally solved the flaking problems) that paper/coating is going
to be near the top in terms of evenness of tones.

Let me know how your experiments work out.

Paul
www.PaulRoark.com