Digital BW, The Print

Hi all,

As discussed in another thread, I'm planning to move to a k6 inkset. 
I've Paul's Carbon-6 set in mind and will use QTR. However, my 
questions apply to eboni-6 and the Cone inksets and other rips as 
well.

It will be straightforward for me to make a k6 profile using QTR's 
curve creation algotithms. 

However, these algorithms do not use the full power of k6 sets. 
Essentially, in QTR profiles only 2, sometimes 3 jets are firing at a 
certain gray level. As I grasp from older posts, the essence of a 
good k6 profile is to have many jets firing simultanuoulsy. I know 
HOW to do it with QTR (with the user curve feature), but frankly I do 
not WHAT to do to get the "best" possible profile (if there is 
something as "best"...).

My key questions boil down to the following:

1) How many jets should be firing simultanuously? As many as possible 
or is their a sensible/pragmatic optimum?

2) What is for an ink, with a certain density, the optimal location 
of the peak in the "density-load curve"? When I try to reverse 
engineer the QTR algorithms, it looks like that the peak is more or 
less located halfway 100% (black denisty) and its native maximum 
density (e.g. 30% for a typical middle gray), so about at 65%. What 
iactually the optimum? What reasoning is behind it?
  
3) What is the cumulative ink load (adding the loads of all jets at a 
certain gray level) that one can allow? Should I limit this to to the 
ink load level on which one single inks saturates (as determined by 
the procedure tath Tom Moore explains is the QTR manual)? Or can one 
go higher? I've done some simulations in Excel, and it turns out that 
keeping the ink load low is at odds with using as many jets as 
possible? 

Any insight is highly appreciated.

Joost

Joost,

I can't give you a definitive amount on the ink limits, but I can
offer a suggestion. Find the limit of the full black ink by watching
for slight puddling and or printing a pattern with some 1 pixel spaced
white lines surrounded by much larger areas of solid black. When the
edges of those 1 pixel lines start to get fuzzy, you have too much
ink. Now use this ink limit for all the other inks because it will be
much harder to determine the limit for the lighter shades.

As far as mixing goes, you never want to go above the maximum combined
that you set for the full black. With 5 gray inks, you should never
need to go closer than about 75 or 80 percent of the total ink limit
when you overlap the different gray inks. Depending on how much
overlap you use, you may have at least 2 inks firing at all times.
That's about all I can help with since I'm not completely up to speed
with QTR.

Thanks Greg,

I love your trick to determine the maximum ink load. Makes absolute 
sense. Your statement on the combined ink loads confirms what I 
expected. I ran some further simulations in Excel. If you really want 
too keep the ink load to ~80% of the max, it will be a though job to 
get 4-5 jets firing simultanuously. Perhaps lower it to 3. This seems 
to counteract some info I read on Cone's website that the even the 
lightest inks contribute to the very dark gray levels. 

Joost



--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Greg" 
<dfaprinting@...> wrote:
>
> Joost,
> 
> I can't give you a definitive amount on the ink limits, but I can
> offer a suggestion. Find the limit of the full black ink by watching
> for slight puddling and or printing a pattern with some 1 pixel 
spaced
> white lines surrounded by much larger areas of solid black. When the
> edges of those 1 pixel lines start to get fuzzy, you have too much
> ink. Now use this ink limit for all the other inks because it will 
be
> much harder to determine the limit for the lighter shades.
> 
> As far as mixing goes, you never want to go above the maximum 
combined
> that you set for the full black. With 5 gray inks, you should never
> need to go closer than about 75 or 80 percent of the total ink limit
> when you overlap the different gray inks. Depending on how much
> overlap you use, you may have at least 2 inks firing at all times.
> That's about all I can help with since I'm not completely up to 
speed
> with QTR.
>

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Joost Horsten" 
<j.h.j.h@...> wrote:
>
> Thanks Greg,
> 
> I love your trick to determine the maximum ink load. Makes absolute 
> sense. Your statement on the combined ink loads confirms what I 
> expected. I ran some further simulations in Excel. If you really want 
> too keep the ink load to ~80% of the max, it will be a though job to 
> get 4-5 jets firing simultanuously. Perhaps lower it to 3. This seems 
> to counteract some info I read on Cone's website that the even the 
> lightest inks contribute to the very dark gray levels. 
> 
> Joost

The lightest inks can contribute to the darks areas, but why? WHy spray 
all that solvent down when you can more effectively use a darker tone 
to fill in between dots. Truthfully, with the RIPs I have used, I see 
no need to ever have more than 2 inks going at a single time. Maybe 
this is different with QTR, I never really got very far into it.

Now, that said, if you want to use 3 heads firing at once, here is what 
I would suggest. There is a point is each dilution where you get full 
coverage, but not full density. Determine you maximum density, which is 
the point were the paper can take no more ink, or more ink does not 
give you more density. We will refer to this as the full amount, or 
100% (not the real 100% from the heads but 100% for the paper). Now 
start with the lightest shade and work backwards (this is opposite from 
the Epson way). Find the point where you reach complete coverage 
(probably about half way to the cut off), and then you start to mix in 
the next darker shade. You can slowly decrease the the amount of the 
lightest ink, and decrease it until you get to the complete coverage of 
the second shade. Now mix in the third shade following the same 
pattern. All of this needs to be determined by density, not strixctly 
by percentage. If you graph the density vs percentage sprayed on the 
paper it will help you to figure this out. ANd if you graph the 
percentage of ink on paper for the total of the two inks, you will get 
some truly odd looking graphs.

How I arrived at the values was from the linarizing step in the RIP 
which gives you the percentage of ink vs the measured density. Yes it 
is a little obscure, but it sure did work for me. And I do realize that 
describing it in words is a little difficult to translate between what 
I say and what is really happening.

And to make all this harder, the linearity of the full black is your 
guide post. You need to mix the inks so that they end up doing the same 
that the linearized full black is doing, otherwise you end up with a 
non-linear output. Often the RIP software will handle this for you 
automatically, but sometimes not.

Hi Greg,

Thanks once again for your elaborate answer. 

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Greg" 
<dfaprinting@...> wrote:
> 
> The lightest inks can contribute to the darks areas, but why? WHy 
spray 
> all that solvent down when you can more effectively use a darker 
tone 
> to fill in between dots. Truthfully, with the RIPs I have used, I 
see 
> no need to ever have more than 2 inks going at a single time. 

Recall that I don't claim any hands-on expertise here. I'm just 
reflecting what I picked up here in the past. What I thought that was 
common sense is to have a number of jets firing simultanuously two 
reasons: 1) less dottier prints, 2) less microbanding.

If what you say is true, it's one argument less for a carbon-6 setup 
(note that I'm still in doubt whether to use carbon-6 or 4k+)
  
> Maybe this is different with QTR, I never really got very far into 
it.

I don't think this has anything to do with QTR. As a matter of fact 
the type of 2-jet curves you describe can be created with QTR very 
easily.


> Now, that said, if you want to use 3 heads firing at once, here is 
what 
> I would suggest. There is a point is each dilution where you get 
full 
> coverage, but not full density. Determine you maximum density, 
which is 
> the point were the paper can take no more ink, or more ink does not 
> give you more density. We will refer to this as the full amount, or 
> 100% (not the real 100% from the heads but 100% for the paper). Now 
> start with the lightest shade and work backwards (this is opposite 
from 
> the Epson way). Find the point where you reach complete coverage 
> (probably about half way to the cut off), and then you start to mix 
in 
> the next darker shade. You can slowly decrease the the amount of 
the 
> lightest ink, and decrease it until you get to the complete 
coverage of 
> the second shade. Now mix in the third shade following the same 
> pattern. All of this needs to be determined by density, not 
strixctly 
> by percentage. If you graph the density vs percentage sprayed on 
the 
> paper it will help you to figure this out. ANd if you graph the 
> percentage of ink on paper for the total of the two inks, you will 
get 
> some truly odd looking graphs.
> 
> How I arrived at the values was from the linarizing step in the RIP 
> which gives you the percentage of ink vs the measured density. Yes 
it 
> is a little obscure, but it sure did work for me. And I do realize 
that 
> describing it in words is a little difficult to translate between 
what 
> I say and what is really happening.

I think I understand what you try to explain. I made a model in excel 
to simulate what you describe above.

> And to make all this harder, the linearity of the full black is 
your 
> guide post. You need to mix the inks so that they end up doing the 
same 
> that the linearized full black is doing, otherwise you end up with 
a 
> non-linear output. 

This is actually rather logical, but why didn't I think of this 
myself?? I have been struggling to understand what the target ink 
load should be. But indeed a black-only curve can be used as 
guideline.

I think I'm done simulating. I have to make sure I get this Epson 
9000 and start making some real curves....

Joost

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Joost Horsten" 
<j.h.j.h@...> wrote:
> 
> I think I'm done simulating. I have to make sure I get this Epson 
> 9000 and start making some real curves....
> 
> Joost
>

Theory and practice are often much different.

And if you are using a multipass print spped, there really shouldn't 
ever be RIP induced microbanding. There could be paper feed induced 
microbanding, but it should never come from the RIP when working with 
only a single print head.

Hi Joost,

>... I'm planning to move to a k6 inkset. 
> I've Paul's Carbon-6 set in mind and will use QTR.

I actually usually use the Epson driver with an ICC made with QTR's 
(Roy's) Create ICC-RBG.  One of the handouts at the recent workshop 
outlined different levels of controlling Eboni-6 with the Epson 
driver, and I've posted that piece at 
http://www.paulroark.com/BW-Info/Eb6-C6-Workflows.pdf 

Basically, the ICC approach gives me direct, "color managed" printing 
from Photoshop in my Windows environment.

> However, my questions apply to eboni-6 
> and the Cone inksets and other rips as well.

Note that Jon Cone uses a different ink order than I do, and the 
Epson driver approach will not work with his ink order.

...

> 3) What is the cumulative ink load (adding the loads of all 
> jets at a certain gray level) that one can allow? 

Frankly, the fact that the Epson driver takes care of this total, 
cumulative ink load is one reason it's often easier to profile with 
PS curves and the Epson driver.  With widely overlapping curves the 
total ink load can get too high in QTR.

On a related subject, I have in the past said that I get the best 
dmax with only Eboni firing -- no lighter inks.   This is very 
different than Cone's style.  I have to modify my position now.  I 
have found on some printers using some "C" (darkest gray -- 33.3% 
Eboni in the C-6 inkset) will achieve a slightly higher dmax than 
straight Eboni.  A rip is needed to do this.  Presumeably the darkest 
gray ink is filling in between the 100% Eboni dots.

Paul
www.PaulRoark.com

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "pr_roark" 
<pr_roark@...> wrote:
>
  Presumeably the darkest 
> gray ink is filling in between the 100% Eboni dots.
> 
> Paul
> www.PaulRoark.com
>

This must depend on the RIP being used. I have never found white space 
at 100%, in fact I normally get "full coverage" at about 75% and the 
Dmax continues to get darker down to 100%. Yet I also get 
differentiation with a 1 pixel line going through the 100% areas, so it 
isn't like the ink is spreading past the boundaries.