Digital BW, The Print

I use an Epson 2100 with Epson UC inks and have been experimenting 
with QTR to see if I can optimise the print quality. I have ordered 
some Piezography inks so while I am awaiting their arrival I want to 
establish a benchmark of what the Epson inks can achieve if 
optimised. 
I like the warm tones I get by just using the Mk and Lk inks but like 
the smoothness I get when I add a little LM and LC to cool the print 
a little. To see if I can improve smoothness in warm prints I have 
been trying printing at 2880 dpi. I built some linearised curves 
using my newly acquired Printfix Pro 2 (I love that little toy) for 
2880 printing (ink limit ended up at 37!). I used Epson Archival 
Matte paper (called Enhanced Matt I believe in most countries) for 
the experiments discussed above together with Epson Mk and Lk.

After a lot of time, ink and paper expended I have to report that I 
cannot see any appreciable difference compared to the prints made at 
1440 dpi.

Also, to validate the linearity of my linearisations, I printed a 
radial gradient. Again I was a little disappointed to see some 
distinct steps that make my radial gradient look a little like a 
bullseye target.

My questions of the group are:

1) Are my findings regarding 1440 vs 2880 printing consistent with 
the experience of others or should I persist?

2) Can I reasonably expect a smooth radial gradient from just K and 
LK on an Epson 2100?

3) If so how do I "tweak" the linearisation to achieve the smooth 
radial gradient? Bear in mind I have already linearised using the the 
tool in QTR.

Thanks and regards

Alistair Owens
owens@...

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "alistair_owens" <owens@...> 
wrote:
>
> I use an Epson 2100 with Epson UC inks and have been experimenting 
> with QTR to see if I can optimise the print quality. I have ordered 
> some Piezography inks so while I am awaiting their arrival I want to 
> establish a benchmark of what the Epson inks can achieve if 
> optimised. 
> I like the warm tones I get by just using the Mk and Lk inks but like 
> the smoothness I get when I add a little LM and LC to cool the print 
> a little. To see if I can improve smoothness in warm prints I have 
> been trying printing at 2880 dpi. I built some linearised curves 
> using my newly acquired Printfix Pro 2 (I love that little toy) for 
> 2880 printing (ink limit ended up at 37!). I used Epson Archival 
> Matte paper (called Enhanced Matt I believe in most countries) for 
> the experiments discussed above together with Epson Mk and Lk.
> 
> After a lot of time, ink and paper expended I have to report that I 
> cannot see any appreciable difference compared to the prints made at 
> 1440 dpi.
> 
> Also, to validate the linearity of my linearisations, I printed a 
> radial gradient. Again I was a little disappointed to see some 
> distinct steps that make my radial gradient look a little like a 
> bullseye target.
> 
> My questions of the group are:
> 
> 1) Are my findings regarding 1440 vs 2880 printing consistent with 
> the experience of others or should I persist?
> 
> 2) Can I reasonably expect a smooth radial gradient from just K and 
> LK on an Epson 2100?
> 
> 3) If so how do I "tweak" the linearisation to achieve the smooth 
> radial gradient? Bear in mind I have already linearised using the the 
> tool in QTR.
> 

Hello Alistair, I don't print with a 2100 so I can't comment on the 1440/2880 issue but I 
have a comment on the radial gradient and this has come up before in other forums.  What 
is the source of the radial gradient?  I have made them in PS and they will show banding 
on their own.  I have also moved a radial gradient from one monitor to another and 
watched the banding change.  Everyone assumes the gradient is perfectly smooth and any 
banding is the result of the printer but I am not convinced of that.  Many think the 
algorithm used by PS for gradients is the culprit.  Others suggest different blurring 
techniques when making the gradient.

Somewhere inside QTR you can find a document: Ideal Densities for Given dMax Values for 
a 21 step wedge.  I think this is a more useful tool.  You can also add your own curves 
before linearization to smooth the gradient allowing the linearize function to have to do 
only the fine tuning.  Creating the Curves, I believe, differs between Mac and QTR GUI.  I 
have only a tenuous grasp of the concept myself for Macs creating them in text mode, but 
the info is out there and perhaps others can give you a better direction.

Robert

You might also wish to create your gradient in 16 bit instead of 8 
bit.  16 bit has so many levels of gray that they tend to look much 
smoother on print out.  

FWIW, I find 2880 give me a better Dmax and smoother output.  That is 
what I use for all my QTR profiles.  This is on a 2200 and a 4000.  

Lou

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "robert49brake" 
<robert49brake@...> wrote:
>
> --- In 
DigitalBlackandWhiteThePrint@yahoogroups.com, "alistair_owens" 
<owens@> 
> wrote:
> >
> > I use an Epson 2100 with Epson UC inks and have been 
experimenting 
> > with QTR to see if I can optimise the print quality. I have 
ordered 
> > some Piezography inks so while I am awaiting their arrival I want 
to 
> > establish a benchmark of what the Epson inks can achieve if 
> > optimised. 
> > I like the warm tones I get by just using the Mk and Lk inks but 
like 
> > the smoothness I get when I add a little LM and LC to cool the 
print 
> > a little. To see if I can improve smoothness in warm prints I 
have 
> > been trying printing at 2880 dpi. I built some linearised curves 
> > using my newly acquired Printfix Pro 2 (I love that little toy) 
for 
> > 2880 printing (ink limit ended up at 37!). I used Epson Archival 
> > Matte paper (called Enhanced Matt I believe in most countries) 
for 
> > the experiments discussed above together with Epson Mk and Lk.
> > 
> > After a lot of time, ink and paper expended I have to report that 
I 
> > cannot see any appreciable difference compared to the prints made 
at 
> > 1440 dpi.
> > 
> > Also, to validate the linearity of my linearisations, I printed a 
> > radial gradient. Again I was a little disappointed to see some 
> > distinct steps that make my radial gradient look a little like a 
> > bullseye target.
> > 
> > My questions of the group are:
> > 
> > 1) Are my findings regarding 1440 vs 2880 printing consistent 
with 
> > the experience of others or should I persist?
> > 
> > 2) Can I reasonably expect a smooth radial gradient from just K 
and 
> > LK on an Epson 2100?
> > 
> > 3) If so how do I "tweak" the linearisation to achieve the smooth 
> > radial gradient? Bear in mind I have already linearised using the 
the 
> > tool in QTR.
> > 
> 
> Hello Alistair, I don't print with a 2100 so I can't comment on the 
1440/2880 issue but I 
> have a comment on the radial gradient and this has come up before 
in other forums.  What 
> is the source of the radial gradient?  I have made them in PS and 
they will show banding 
> on their own.  I have also moved a radial gradient from one monitor 
to another and 
> watched the banding change.  Everyone assumes the gradient is 
perfectly smooth and any 
> banding is the result of the printer but I am not convinced of 
that.  Many think the 
> algorithm used by PS for gradients is the culprit.  Others suggest 
different blurring 
> techniques when making the gradient.
> 
> Somewhere inside QTR you can find a document: Ideal Densities for 
Given dMax Values for 
> a 21 step wedge.  I think this is a more useful tool.  You can also 
add your own curves 
> before linearization to smooth the gradient allowing the linearize 
function to have to do 
> only the fine tuning.  Creating the Curves, I believe, differs 
between Mac and QTR GUI.  I 
> have only a tenuous grasp of the concept myself for Macs creating 
them in text mode, but 
> the info is out there and perhaps others can give you a better 
direction.

> 
> Robert
>

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "robert49brake" 
<robert49brake@...> wrote:
>
> 
> Hello Alistair, I don't print with a 2100 so I can't comment on the 
1440/2880 issue but I 
> have a comment on the radial gradient and this has come up before 
in other forums.  What 
> is the source of the radial gradient? 

Hi Robert, thanks for your response and apologies for my belated 
reply. The source of my redial gradient was the Northlight Test image 
which can be found here: http://www.northlight-
images.co.uk/article_pages/black_and_white_test.html 

I do get a smooth "bullseye-free" gradient using the cooler curve. Of 
course this curve fires 4 inks (Mk, LK, LM, LC) so one would expect 
some additional smoothness when compared to the warm curve. What I am 
really trying to establish is whether anyone has been able to achieve 
a "bullseye-free" radial gradient using just the warm curve to fire 
only Mk and Lk.

 
> 
> Somewhere inside QTR you can find a document: Ideal Densities for 
Given dMax Values for 
> a 21 step wedge.  I think this is a more useful tool.  You can also 
add your own curves 
> before linearization to smooth the gradient allowing the linearize 
function to have to do 
> only the fine tuning.  Creating the Curves, I believe, differs 
between Mac and QTR GUI.  I 
> have only a tenuous grasp of the concept myself for Macs creating 
them in text mode, but 
> the info is out there and perhaps others can give you a better 
direction.
> 

You are right, that is really what I need to find out how to do. I 
have found the "Ideal Densities for Given dMax Values" document and 
my densities are materially different to the ideals given in that 
document. I need to find a way to fiddle the linearisation or the 
curves to get closer to the ideal densities. 

Can anyone offer suggestions to achieve this?

regards

Alistair Owens
owens@...

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Louis Dina" 
<lou@...> wrote:
>
> You might also wish to create your gradient in 16 bit instead of 8 
> bit.  16 bit has so many levels of gray that they tend to look much 
> smoother on print out.  
> 
> FWIW, I find 2880 give me a better Dmax and smoother output.  That is 
> what I use for all my QTR profiles.  This is on a 2200 and a 4000.  
> 
> Lou
> 

Thanks Lou. In that case I will persist with the 2880 curves and refine 
them further.

regards

Alistair Owens