QTR-Quadtone RIP

I have noticed (and a friend also commented on this) that I often 
will get a higher Dmax in my original InkSeparation.tif print than I 
do in my final profiles.  For example, a given paper may show a Dmax 
at 80% ink in the MK only test strip of 1.68, but when I complete my 
profile, using that ink limit, I usually lose maybe 0.03 to 0.05 
Dmax.  So, my final profile may show a Dmax of 1.63 or 1.64, instead 
of 1.68.

I have tried using a lower ink limit, for example 60%, then used a 
black boost of 80% for the MK ink, and I have also tried using a flat 
80% TIL.  In both cases, my final profile loses a little of its Dmax.

I am guessing that the addition of some lk, lc, or lm inks is 
actually lightening the ink in the blacks, or perhaps contributing to 
slight over inking of the paper.  So, I tried lowering the black 
boost (or overall ink limit) to 70% or 75%, but I still see that loss 
in all my profiles (neutral, warm, sepia, cool, coolSe).  

I am assuming that the settings in the Gray Toner tab might help lay 
down less ink in the deepest shadows.  I hate to give up any Dmax on 
some of the standard cotton papers, and some of them are coming in 
with a Dmax near 1.6 to 1.63 or so on my 4000 (using MISPro inks).

Have others noticed this?  Is there an easy/good way around it?  

Thanks, Lou Dina

I see it pretty consistently.  But, I don't know the answer either.
   
  Tom Baker
  

Louis Dina <lou@...> wrote:
          I have noticed (and a friend also commented on this) that I often 
will get a higher Dmax in my original InkSeparation.tif print than I 
do in my final profiles. For example, a given paper may show a Dmax 
at 80% ink in the MK only test strip of 1.68, but when I complete my 
profile, using that ink limit, I usually lose maybe 0.03 to 0.05 
Dmax. So, my final profile may show a Dmax of 1.63 or 1.64, instead 
of 1.68.

I have tried using a lower ink limit, for example 60%, then used a 
black boost of 80% for the MK ink, and I have also tried using a flat 
80% TIL. In both cases, my final profile loses a little of its Dmax.

I am guessing that the addition of some lk, lc, or lm inks is 
actually lightening the ink in the blacks, or perhaps contributing to 
slight over inking of the paper. So, I tried lowering the black 
boost (or overall ink limit) to 70% or 75%, but I still see that loss 
in all my profiles (neutral, warm, sepia, cool, coolSe). 

I am assuming that the settings in the Gray Toner tab might help lay 
down less ink in the deepest shadows. I hate to give up any Dmax on 
some of the standard cotton papers, and some of them are coming in 
with a Dmax near 1.6 to 1.63 or so on my 4000 (using MISPro inks).

Have others noticed this? Is there an easy/good way around it? 

Thanks, Lou Dina



                         


[Non-text portions of this message have been removed]

Lou,

 

Try printing a 100% patch in the middle of the page instead of at the edge.
Also, avoid any place where a roller has gone across the paper previously.

 

Paul

www.PaulRoark.com <http://www.paulroark.com/>  

 

 

  _____  

From: QuadtoneRIP@yahoogroups.com [mailto:QuadtoneRIP@yahoogroups.com] On
Behalf Of Louis Dina
Sent: Thursday, November 29, 2007 1:51 PM
To: QuadtoneRIP@yahoogroups.com
Subject: [QuadtoneRIP] Dmax Question

 

I have noticed (and a friend also commented on this) that I often 
will get a higher Dmax in my original InkSeparation.tif print than I 
do in my final profiles. For example, a given paper may show a Dmax 
at 80% ink in the MK only test strip of 1.68, but when I complete my 
profile, using that ink limit, I usually lose maybe 0.03 to 0.05 
Dmax. So, my final profile may show a Dmax of 1.63 or 1.64, instead 
of 1.68.

I have tried using a lower ink limit, for example 60%, then used a 
black boost of 80% for the MK ink, and I have also tried using a flat 
80% TIL. In both cases, my final profile loses a little of its Dmax.

I am guessing that the addition of some lk, lc, or lm inks is 
actually lightening the ink in the blacks, or perhaps contributing to 
slight over inking of the paper. So, I tried lowering the black 
boost (or overall ink limit) to 70% or 75%, but I still see that loss 
in all my profiles (neutral, warm, sepia, cool, coolSe). 

I am assuming that the settings in the Gray Toner tab might help lay 
down less ink in the deepest shadows. I hate to give up any Dmax on 
some of the standard cotton papers, and some of them are coming in 
with a Dmax near 1.6 to 1.63 or so on my 4000 (using MISPro inks).

Have others noticed this? Is there an easy/good way around it? 

Thanks, Lou Dina

 



[Non-text portions of this message have been removed]

Louis Dina wrote:
> 
> I am guessing that the addition of some lk, lc, or lm inks is 
> actually lightening the ink in the blacks, or perhaps contributing to 
> slight over inking of the paper.  So, I tried lowering the black 
> boost (or overall ink limit) to 70% or 75%, but I still see that loss 
> in all my profiles (neutral, warm, sepia, cool, coolSe).  
> 
To avoid the lighter ink firing at 100% you need to set the Overlap 
figure to zero (on the Gray Curve tab). Your black boost figure should 
then replicate the limit on the calibration print. I don't usually find 
this is necessary but then I'm using grayscale inks which probably have 
a darker LK.

Howard

--- In QuadtoneRIP@yahoogroups.com, Howard Shaw <glassman@...> wrote:
> To avoid the lighter ink firing at 100% you need to set the Overlap 
> figure to zero (on the Gray Curve tab). Your black boost figure 
should 
> then replicate the limit on the calibration print. I don't usually 
find 
> this is necessary but then I'm using grayscale inks which probably 
have 
> a darker LK.
> 
> Howard
>

Thanks, Howard.  

All my profiles have the overlap set to 0.  I agree, the black boost 
SHOULD replicate what I got on the calibration target, but it 
doesn't.  The final profile always has a Dmax that is lighter by 0.03 
to 0.05 density.  I kind of wish I could set the overlap to a 
negative number to move the light ink curves further to the left, but 
QTR doesn't seem to like that.  It always reverts back to zero.  I 
also tried setting the Shadow setting to a lower number in the Gray 
Curve tab, since this moves all the inks to the left and darkens 
everything in the shadow end, but that didn't seem to do it.

So, I remain suspicious that light inks may be stealing just a little 
Dmax, and the only way I can currently see getting around this is to 
create custom curves using the plot list, but I haven't tried this 
yet.  It may be easy, but it seems like a lot of work.  

Lou

When I was using IJC quite a bit I got suspicious that the smoothing
algorithm was putting some light inks in the 100% patch even though they
were supposedly at 0 at that point.  I could "see" from the IJC graphs that
the final line (to the extent it was accurate) did not always go through the
point.  To counteract this I moved the point at which the light inks hit the
horizontal axis a little to the left of the 100% right edge of the IJC
graphical interface.  It seemed to work.

 

Paul

www.PaulRoark.com <http://www.paulroark.com/>  

 

 

  _____  

 On Behalf Of Louis Dina



All my profiles have the overlap set to 0. .

So, I remain suspicious that light inks may be stealing just a little 
Dmax, . 



[Non-text portions of this message have been removed]

Louis Dina wrote:
 >
> 
> All my profiles have the overlap set to 0.  I agree, the black boost 
> SHOULD replicate what I got on the calibration target, but it 
> doesn't.  The final profile always has a Dmax that is lighter by 0.03 
> ...

Perversely then, I might suggest that you try a positive figure, say 
10-20, for the overlap.

Also, for testing purposes you could try a black only curve - that would 
confirm or eliminate the LK as being the problem.

Howard

Howard,

Good point.  I will try both.  I am guessing that increasing the 
overlap will hurt rather than help, but you never know until you try 
it.  The black only curve should make it clear whether the light inks 
are the culprit or not.  If I get the same Dmax in a BO profile as I 
do when I print the inkseparation target, that would seem to confirm 
it.  If not, then something else is at play here.

Regards, Lou

--- In QuadtoneRIP@yahoogroups.com, Howard Shaw <glassman@...> wrote:
>
> Louis Dina wrote:
>  >
> > 
> > All my profiles have the overlap set to 0.  I agree, the black 
boost 
> > SHOULD replicate what I got on the calibration target, but it 
> > doesn't.  The final profile always has a Dmax that is lighter by 
0.03 
> > ...
> 
> Perversely then, I might suggest that you try a positive figure, 
say 
> 10-20, for the overlap.
> 
> Also, for testing purposes you could try a black only curve - that 
would 

> confirm or eliminate the LK as being the problem.
> 
> Howard
>

Thanks, Paul, for this and your previous post.  Your experience with 
IJC seems to suggest light inks (or even additional dark inks, such 
as C or M) can and sometimes do reduce the Dmax, especially if the MK 
ink is already at optimum ink density and there are no small white 
voids in the printed pattern.  Like you, I have adopted 2880 dpi, 
unidirectional, ordered printing for all my QTR profiles, which seems 
to give me better results.  

I am toying with the idea of generating a series of neutral, warm, 
sepia and cool curves for one specific paper based on using the plot 
list as you have done in some of your profiles.  I will be using the 
MISPro color inkset.  I am hoping that most papers respond the same 
to the <b>"ink relationship"</b>, so that I can simply adjust the 
overall ink limits and relinearize when I migrate from one paper to 
another.  So, theoretcially, a neutral QTR profile on my initial 
paper will still give me neutral results on new papers using the same 
plot list (ignoring the color of the paper base, of course), and I 
would just have to change the overall ink limit so it is appropriate 
for the new paper (whether matte, luster, or whatever).  This is may 
be wishful thinking.  If it works, though, it would sure make profile 
creation easier, and possibly even more repeatable from one paper to 
the next (ie, all sepia profiles would tend to have the same 
coloration).  And yes, I still believe in Santa Claus.

Any thoughts or experience you have on this would be appreciated.  If 
it works, I can quickly generate great profiles, plus make certain 
that I don't have ANY light inks in the deepest shadows, thus giving 
me the better Dmax I seek.

Regards, Lou


--- In QuadtoneRIP@yahoogroups.com, "Paul Roark" <paul.roark@...> 
wrote:
>
> When I was using IJC quite a bit I got suspicious that the smoothing
> algorithm was putting some light inks in the 100% patch even though 
they
> were supposedly at 0 at that point.  I could "see" from the IJC 
graphs that
> the final line (to the extent it was accurate) did not always go 
through the
> point.  To counteract this I moved the point at which the light 
inks hit the
> horizontal axis a little to the left of the 100% right edge of the 
IJC
> graphical interface.  It seemed to work.
> 
>  
> 
> Paul
> 
> www.PaulRoark.com <http://www.paulroark.com/>  
> 
>  
> 
>  
> 
>   _____  
> 
>  On Behalf Of Louis Dina
> 
> 
> 
> All my profiles have the overlap set to 0. .
> 
> So, I remain suspicious that light inks may be stealing just a 
little 

> Dmax, . 
> 
> 
> 
> [Non-text portions of this message have been removed]
>

Howard.....a little feedback.  I ran both tests you suggested.  

I found that using BO in my profile did give me the same Dmax as I 
originally got in the MK step wedge when printing the original 
inkseparation file.  In this test, I printed on Red River Aurora 
Natural paper, and hit a Dmax of 1.65 on my 4000 at 85% ink.  I got 
this on the inkseparation target and a BO profile using only the MK 
with an 85% ink limit.  (This is using the MISPro color inkset - UC 
equivalents).

I also tried the counterintuitive move in increasing the overlap of 
the lighter inks (it was originally set to 0).  I tried an overlap of 
10 and 20.  They both slightly reduced the Dmax to about 1.63, but 
that was less loss than I anticipated.  The neutral profile I built 
had LK, LC and LM inks for neutrality and smoothness in the light 
tones.

This leads me to conclude that the addition of light inks into the 
shadows is, in fact, slightly reducing Dmax in most of my profiles.  
I don't see a way of completely eliminating this using the normal 
tools in QTR's Gray Ink tab.  You can minimize it by moving the 
Shadow field in the Gray Curve Tab to a lower number (closer to 
zero), which limits the use of light inks in the shadows.  To 
completely eliminate light inks in the last step of the step wedge, 
I'd probably have to create a plot list or import a curve from PS 
(which may be my next experiment).  

I also think Paul Roark's comment may be part of the explanation 
too.  Since I tweak a profile by repeatedly sending the same sheet of 
paper through the printer, (printing the 21 step random test target), 
perhaps wheel marks are having some impact and skewing readings.  
Also, I use the microwave for 12-15 seconds after each test strip is 
printed, and perhaps the repeated use of the microwave has some 
impact on paper moisture, ink receptivity, or something.  Not sure 
about those, but I do feel reasonably confident after these tests 
that light inks are most most of the reason for a loss of 0.03 to 
0.05 Dmax in the deep shadows.  

Lou


--- In QuadtoneRIP@yahoogroups.com, "Louis Dina" <lou@...> wrote:
>
> Howard,
> 
> Good point.  I will try both.  I am guessing that increasing the 
> overlap will hurt rather than help, but you never know until you 
try 
> it.  The black only curve should make it clear whether the light 
inks 
> are the culprit or not.  If I get the same Dmax in a BO profile as 
I 
> do when I print the inkseparation target, that would seem to 
confirm 
> it.  If not, then something else is at play here.
> 
> Regards, Lou
> 
> --- In QuadtoneRIP@yahoogroups.com, Howard Shaw <glassman@> wrote:
> >
> > Louis Dina wrote:
> >  >
> > > 
> > > All my profiles have the overlap set to 0.  I agree, the black 
> boost 
> > > SHOULD replicate what I got on the calibration target, but it 
> > > doesn't.  The final profile always has a Dmax that is lighter 
by 
> 0.03 
> > > ...
> > 
> > Perversely then, I might suggest that you try a positive figure, 
> say 
> > 10-20, for the overlap.
> > 
> > Also, for testing purposes you could try a black only curve - 
that 

> would 
> > confirm or eliminate the LK as being the problem.
> > 
> > Howard
> >
>

Louis Dina wrote:
> 
> I found that using BO in my profile did give me the same Dmax as I 
> originally got in the MK step wedge when printing the original 
> inkseparation file.  In this test, I printed on Red River Aurora 
> Natural paper, and hit a Dmax of 1.65 on my 4000 at 85% ink.  I got 
> this on the inkseparation target and a BO profile using only the MK 
> with an 85% ink limit.  (This is using the MISPro color inkset - UC 
> equivalents).
> 
> I also tried the counterintuitive move in increasing the overlap of 
> the lighter inks (it was originally set to 0).  I tried an overlap of 
> 10 and 20.  They both slightly reduced the Dmax to about 1.63, but 
> that was less loss than I anticipated.  The neutral profile I built 
> had LK, LC and LM inks for neutrality and smoothness in the light 
> tones.
> 
> This leads me to conclude that the addition of light inks into the 
> shadows is, in fact, slightly reducing Dmax in most of my profiles.  
> I don't see a way of completely eliminating this using the normal 
> tools in QTR's Gray Ink tab.  You can minimize it by moving the 
> Shadow field in the Gray Curve Tab to a lower number (closer to 
> zero), which limits the use of light inks in the shadows.  To 
> completely eliminate light inks in the last step of the step wedge, 
> I'd probably have to create a plot list or import a curve from PS 
> (which may be my next experiment).  
> 

Louis

I just did a quick test to see exactly what the overlap figure does to 
the numbers. (You probably know that you can open the .quad files that 
are produced by the curve creation process.) With an overlap of zero the 
output for the LK curve at point 255 (ie 100% black) is 0 although at 
point 254 it does start to come in. As overlap is increased the amount 
of LK present at point 255 also increases as one would expect. In theory 
therefore 100% black with an overlap of zero should be giving you only 
the black ink. Changing the highlight & shadow settings did not alter this.

I don't therefore understand the Dmax results you are getting. As you 
suggested earlier using the point list curve input might be worth 
trying. You could take the visual representation of the curves to 
approximate your points but then just alter the LK slightly so it 
doesn't come in until say 90%.

regards
Howard

--- In QuadtoneRIP@yahoogroups.com, Howard Shaw <glassman@...> wrote:

> I don't therefore understand the Dmax results you are getting. As you 
> suggested earlier using the point list curve input might be worth 
> trying. You could take the visual representation of the curves to 
> approximate your points but then just alter the LK slightly so it 
> doesn't come in until say 90%.

Howard, 

It is true that the light inks are at 0 at 100% when overlap is set to 
0, but they start coming in right after that.  It seems to be just 
enough to have some impact on Dmax.  

I created a whole series of profiles yesterday on Red River Aurora 
Natural paper (neutral, warm, sepia, cool, selenium).  These are for my 
4000, using MISPro color inks, and I used the plot list to create all 
the curves.  I did precisely as you suggested above, with one 
adjustment.  I started with a previously made, linearized profile 
(using the standard QTR tools).  I duplicated the points in a plot 
list, but I turned off all the light inks at 97% instead of at 100%.  
This gave me back some of the Dmax I had lost previously.  My 
adjustment was as follows.  My ink limit for this paper is 85%, so I 
multiplied all the points in my curves by 100/85 (1.18), so my plot 
list would be based on 100% ink limits.  Then by telling QTR that my 
default ink limit is 85%, the final plot list curves are automatically 
adjusted downward to match the numbers in the original profile.  This 
method will allow me to use any ink limit up to 100%, and the curves 
will all adjust up or down proportionally based on the default ink 
limit the user inputs into QTR.

This was a LOT of work initially, but I am hoping (perhaps foolishly) 
that it will allow me to migrate these curves to other papers and 
simply select the appropriate ink limit and relinearize.  If it works, 
it will be very fast and easy to create curves for new papers.  I am 
also hoping that my toned curves will be much more consistent from one 
paper to another (ignoring the color of the paper base, of course).  I 
haven't created any profiles for other papers yet, but hope to do this 
in the next 2 or 3 days when time is available.  

It would be a whole lot easier if QTR allowed the user to insert a 
negative overlap, or better yet, gave us a chance specify the point 
where light inks dropped to zero (ie, 90%, 95%, 100%, etc).

I'll let you know how things work out when I migrate these profiles to 
other papers.  I plan to try moving them to different matte papers, but 
I also want to see what happens if I migrate a matte profile to glossy 
or luster paper.  And finally, I will try moving a profile from my 4000 
to my 2200 using the same inkset.  Should be interesting.  

Lou

Louis Dina wrote:

> It is true that the light inks are at 0 at 100% when overlap is set to 
> 0, but they start coming in right after that.  It seems to be just 
> enough to have some impact on Dmax.  

> Lou

Lou,

Where do you limit black ink itself, right at the point 
where no density increase happens or do you go a bit higher 
where the extra ink shouldn't add more density in theory but 
isn't taking density away either ? I have done the last in 
the past with the idea that some reduction by the 
linearisation curve just below 100% will solve a flat part 
there but a stricter ink limit may loose density in the two 
steps of linearising and profiling and/or fluctuations of 
printer and paper coating. In short a slight bleed to be 
preferred above any white appearing.

On the darkest grey ink running close to the 100% and 
reducing density I think it depends much on how dark that 
grey ink is and how consistent the printer's dot pattern. I 
had my doubts about the HP Z3100 quad use for  some matte 
papers:  MK,PK,Medium Grey and Light Grey where the PK can 
not bridge much of the tonal range so why use it.  But 
before that I used the same configuration in my 9000 MIS 
quad to fill in (to a degree) irregularities of the black 
dot pattern. The same concept that is used in multihead BO 
printing now en vogue. I have no idea how the HP Z3100 B&W 
mode adds the PK in the partitioning but I think it is more 
to deepen the black than for the shadow range. The HP B9180 
doesn't have that quad partitioning but has superb 4800 
lines stepping resolution hardware and 1000 nozzles per head 
so also for the black. Based on that is a superb screening 
resolution and patterns with 16 bit screening masks. No need 
for PK to close the ranks. When Roy made QTR the printers he 
had to work with then could use a bit of overflow from the 
dark grey ink instead of relying on the black only. It may 
not have been intentional but I guess he did work with trial 
and error. No problem, no solution needed. Now with better 
resolution and consistency this may be no longer the case.


-- 
Met vriendelijke groeten, Ernst


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

Hi Ernst.  

--- In QuadtoneRIP@yahoogroups.com, Ernst Dinkla <E.Dinkla@...> wrote:

> Where do you limit black ink itself, right at the point 
> where no density increase happens or do you go a bit higher 
> where the extra ink shouldn't add more density in theory but 
> isn't taking density away either ? I have done the last in 
> the past with the idea that some reduction by the 
> linearisation curve just below 100% will solve a flat part 
> there but a stricter ink limit may loose density in the two 
> steps of linearising and profiling and/or fluctuations of 
> printer and paper coating. In short a slight bleed to be 
> preferred above any white appearing.

When I was using the typcal QTR controls, I found I got better 
profiles using a slightly lower default ink limit.  I would typically 
select an ink limit at the "shoulder" of the black black ink (for 
example, if Dmax was 1.68 at 75% ink, I might select 55 or 60% where 
the Dmax was around 1.6 or so and the curve was just 'beginning' to 
steepen).  This resulted in a lower total ink and better separation 
in the shadows.  I often found that setting the ink limit exactly at 
the Dmax caused the shadow steps to bunch up.  To get back the better 
Dmax, I set the black boost to 75%, corresponding to the best Dmax in 
my calibration print.  I have never gone to a higher ink load for the 
black, since my intial ramp in the calibration print always showed 
that the Dmax either stayed the same or dropped slightly.  

> On the darkest grey ink running close to the 100% and 
> reducing density I think it depends much on how dark that 
> grey ink is and how consistent the printer's dot pattern. I 
> had my doubts about the HP Z3100 quad use for  some matte 
> papers:  MK,PK,Medium Grey and Light Grey where the PK can 
> not bridge much of the tonal range so why use it.  But 
> before that I used the same configuration in my 9000 MIS 
> quad to fill in (to a degree) irregularities of the black 
> dot pattern. The same concept that is used in multihead BO 
> printing now en vogue. I have no idea how the HP Z3100 B&W 
> mode adds the PK in the partitioning but I think it is more 
> to deepen the black than for the shadow range. The HP B9180 
> doesn't have that quad partitioning but has superb 4800 
> lines stepping resolution hardware and 1000 nozzles per head 
> so also for the black. Based on that is a superb screening 
> resolution and patterns with 16 bit screening masks. No need 
> for PK to close the ranks. When Roy made QTR the printers he 
> had to work with then could use a bit of overflow from the 
> dark grey ink instead of relying on the black only. It may 
> not have been intentional but I guess he did work with trial 
> and error. No problem, no solution needed. Now with better 
> resolution and consistency this may be no longer the case.

Very interesting comments on the HP printers and the older printers.  
I'm using an Epson 4000 and a 2200.  What I have been doing recently 
is creating my own curves using the Plot Lists in QTR.  This allows 
me to totally control the shape of the curves, and also where each 
ink starts and ends.  I am using MISPro color inks (UltraChrome 
equivalents), so in addition to the Eboni black, I typically use LK, 
LC and LM (I only use Y with my sepia profiles).  These inks are 
fairly light, and I believe they do reduce Dmax slightly due to 
overlap, at least in my testing and using my methods.  Using plot 
lists, I have stopped all light inks past about 95%, so only the 
Eboni is used for solid black.  I am using 2880 dpi, unidirectional 
printing and the ordered dither pattern, so I get good coverage with 
Eboni only, and I don't see any white at all when viewing with a 
loupe.  

Another advantage of using plot lists is that my curves are almost 
perfectly linear before performing final linearization.  That should 
provide better profiles overall, though I am not sure if it is 
noticeable or not.  Building the initial set of curves with plot 
lists is time consuming.....but....over the next weeek, I am planning 
to repurpose my initial profiles and see how they migrate to 
different papers, and even different printers.  I am hoping I can 
simply determine the correct ink limit for the new paper, print a 
test strip, and relinearize.  If that is successful, it will be very 
fast and easy to build new profiles, and hopefully, they will be more 
consistent from one paper to another.  It should also allow me to get 
the best Dmax that paper is capable of, without any loss due to light 
ink overlap.  I'll report back.

Regards, 

Lou