Yahoo Groups archive

Digital BW, The Print

Index last updated: 2026-04-28 22:56 UTC

Thread

Ink Limits in IJC

Ink Limits in IJC

2003-04-01 by Nick Wheeler

Antonis:

The behavior of the Ink Jet Control application does not seem to conform to
the traditional prepress standards for ink limiting and I am trying to
figure out what is happening here. By that I mean that traditional cmyk
curves relate to actual ink percentages as used to print and the ink limit
is built in to the curves.

In the IJC app it seems ink limiting is handled behind the scenes somehow,
where one dials in the appropriate step value and that value determines the
ink limit.

Once this new limit is applied it is not reflected by a redrawing of the
crossover curves. Therefore I'm led to conclude that the crossover curves do
not represent real ink percentages, but conform to some other standard which
I don't see documented anywhere.

In addition the "ink color" patch values seem hard to determine, one is left
having to visually match them to actual ink colors somehow?

It seems that one is left trying to generate a new "ink formula" flying
blind (the use of the word profile here I believe to be inappropriate). One
is drawing curves that relate to indefinate values and the only visual
feedback is via color patches which are likely to be extremely inaccurate.

Am I missing something here? How would one accurately assign values to the
ink patches to get a reasonable preview of banding problems etc. And what
values do the crossover curves actually represent, a percentage of a
percentage?

The linearization feature seems superb, very well thought out. It would be
nice if this could be used "stand alone" for linearizing a variety of
printing behaviors, not just OPM. ImagePrint, or other quadtone workflows
come to mind. All that would be needed is rgb target values documented
somehere, or just have a small target provided that could be printed via
other applications. 200.00 for this feature alone would be money well spent.

I think it is also important to add that the "print dialog" is hands down
the best I've ever seen. It might be nice to pull the single vs multi pass
option out of preferences but other than that it's great.

Best wishes,

Nick Wheeler

Re: Ink Limits in IJC

2003-04-01 by Antonis Ricos

Nick,

thanks for the thoughtful review of IJC and its unique "take" on controlling ink. 
I agree with you that if you see it from a traditional prepress perspective, it 
won't conform to what we expect from color profile-making programs. This one 
is built purely on the logic of how best to lay down a grayscale and partition 
the gray inks. 

Some responses to your comments:


By that I mean that traditional cmyk
> curves relate to actual ink percentages as used to print and the ink limit
> is built in to the curves.

As you know those percentages are derived from the ink coverage of a given 
dot area of a traditional screen (when half that dot is covered with ink we call 
it 50%). In the case of IJC, the percentage numbers merely refer to a place on 
a grayscale from 0 to 100.


> 
> In the IJC app it seems ink limiting is handled behind the scenes somehow,
> where one dials in the appropriate step value and that value determines the
> ink limit.

IJC lets you see the effects of letting the printer run with the tap completely 
open, i.e. deliver all the ink the hardware  can handle, even if it runs off the 
paper !  It allows you to print a grayscale for each ink under that state, so you 
can see how much ink your paper can take, then assign a limit of your 
choice. That's a crucial feature in IJC that allows for true determination of the 
maximum density of a paper/ink combo. In that sense, it does not do 
something "behind the scenes", but rather hands the control to you so you 
can dial a value from 1 to 26 based on the results of the test print above. 

Again, it doesn't really match the traditional prepress notions of "total ink" or 
UCR. For that matter, it doesn't even address the effects of the total ink that is 
layed down at maximum black when more than one ink may be running at the 
same time. The only way to find out what happens if you overload the paper 
by - for example - running full "cyan" _and_  full black at the 100% patch, is 
trial and error and good densitometry. In traditional prepress terms we would 
have called that 200% total ink (100%cyan+100%black) - a limit well below 
even newspaper applications. In the case of inkjets, however, things are 
very different. 200% with dyes may indeed produce a deeper density than 
black alone, but 200% with pigments would be a disaster and dmax will suffer. 

IJC completely bypasses any notion of "percent"  laydown. It gives you a 
way to simply determine how much ink each head will lay down at a given 
point in the grayscale. It does this by showing a 0-100% scale under the ink 
curves. As you place and move a point up or down the curve, more or less 
ink is delivered for those parts of the grayscale that are affected. Simple as 
that. This has nothing to do with the percent of dot coverage as used in 
prepress.

Unfortunately, IJC ommits to give us a value on the vertical axis of that curve, 
such as a scale of 0-255, so we know numerically how much we move the 
points by. Worse yet, it also ommits a scale of 0-26 running in parallel to the 
0-100% scale at the bottom. Such a scale would correspond to the 26-step 
patches that IJC prints for measuring. Since the patches only have numbers 
from 0-26, it would be nice to have these numbers under the ink curves . 
Instead, we now have to print a scale and using a pencil (!!!)  write down what 
percent on the grayscale each patch corresponds to, so we can find it on the 
percentage scale under the ink curves. Of course, you only need do this 
once and then stick it on the side of your monitor (!!), but it makes for an 
inelegant solution.

 
> Once this new limit is applied it is not reflected by a redrawing of the
> crossover curves. 

Choosing an ink limit causes the effects of the whole ink curve to increase or 
decrease  - as if the whole thing went up or down the grid. It would't cause a 
redraw because of the logic of "limits" as it applies to inkjets vs prepress (as I 
discussed above).  Changing ink limits after you have arrived at a good 
curve, gives you a chance to reduce or intensify an ink while maintaining it's 
"shape". This is almost an unintended "feature", very useful in many 
situations. IOW, besides the limiting function as I discussed earlier, these ink 
limits are an added control over and above the curves. An example would be 
the use of a toner. You decide how you want it to behave at each point in the 
grayscale, then adjust overall intensity using the limits. Very cool in pratice!


> 
> In addition the "ink color" patch values seem hard to determine, one is left
> having to visually match them to actual ink colors somehow?

Indeed. Apparently the "wysiwyg" intentions of the  preview that the scales 
give you, could not be materialized. As a result, these are useful as an 
overall "icon" of what curve corresponds to what ink. You pick a color from 
the Apple color picker to assign to each ink by simple eyeballing. It helps you 
only to know that this ink starts and ends here and to see it in relation to other 
inks. It has no effect on the profile you save.

 The composite grayscale preview is also "inaccurate" for showing banding 
and the like, but useful for showing gross error or if you have turned off the 
wrong ink, etc. I was never concerned with using these previews to base 
precision moves on the curves, but they offer useful feedback overall.



> 
> The linearization feature seems superb, very well thought out. It would be
> nice if this could be used "stand alone" for linearizing a variety of
> printing behaviors, not just OPM. 

Not a bad idea - but for that to work, the Aims would have to be editable. As it 
is IJC internally generates a set of 26 density values that represent an ideal 
transition from the max black to paper white (as read in a densitometer from 
the printed target). Not every program, user or image need agree with that 
ideal. But if the aims were user-defined, the internal engine in IJC could 
indeed "linearize" to those  numbers instead of the current built-in "ideals". I 
think this would be a good feature for future IJC versions, anyway.  


> I think it is also important to add that the "print dialog" is hands down
> the best I've ever seen. It might be nice to pull the single vs multi pass
> option out of preferences but other than that it's great.

I agree.... but would like to see simple capabilities added, such as 90 degree 
rotations of the image, user-defined page size, typable image dimentions and 
printable area preview. But we are only in v.1xx. I am sure there is more to 
come down the line.

Thanks for sharing your views and feedback on IJC. 


Antonis

Re: Ink Limits in IJC

2003-04-01 by Nick Wheeler

Antonis:

> From: "Antonis Ricos" <antonisphoto@...> Subject: Re: Ink Limits in IJC
> 

> open, i.e. deliver all the ink the hardware  can handle, even if it runs off
> the paper !  It allows you to print a grayscale for each ink under that state,
> so you can see how much ink your paper can take, then assign a limit of your
> choice. That's a crucial feature in IJC that allows for true determination of
> the maximum density of a paper/ink combo. In that sense, it does not do
> something "behind the scenes", but rather hands the control to you so you can
> dial a value from 1 to 26 based on the results of the test print above.
> 
That is actually the problem I am having at the moment, I'm setting the ink
limits too high (choosing the wrong step obviously), and the ink is running
off the paper :=) I'm trying to figure out how to dial this in better. There
is the problem of the interaction between the step setting one chooses and
then how one draws the curve - I guess that is what I was driving at.

It would seem that if the proper ink limit was selected then the proper
crossover curves would all be close to the same height as at some point one
would want to use 100%, at least in theory. Yet a lot of the curves in the
provided examples don't even come close to this, so I assume I'm not
thinking about this properly.

> In traditional prepress terms we would have called that 200% total ink
> (100%cyan+100%black) - a limit well below even newspaper applications. In the
> case of inkjets, however, things are very different. 200% with dyes may indeed
> produce a deeper density than black alone, but 200% with pigments would be a
> disaster and dmax will suffer.
> 
It would be nice to be able to quickly arrive at what a black black might be
for a given ink set, I haven't figured that one out at all. I am using the 4
color in six position MIS Full Spectrum Neutral inks, care to hazard a guess
at a good starting point for a black black with this type of ink set?
(Similar to the now outdated Cone system)

> IJC completely bypasses any notion of "percent"  laydown. It gives you a way
> to simply determine how much ink each head will lay down at a given point in
> the grayscale. It does this by showing a 0-100% scale under the ink curves. As
> you place and move a point up or down the curve, more or less ink is delivered
> for those parts of the grayscale that are affected. Simple as that. This has
> nothing to do with the percent of dot coverage as used in prepress.
> 
Well dot percentages don't make much sense in ink jet printing anyway as I
can now see clearly from your explanation.

> Unfortunately, IJC ommits to give us a value on the vertical axis of that
> curve, such as a scale of 0-255, so we know numerically how much we move the
> points by. Worse yet, it also ommits a scale of 0-26 running in parallel to
> the 0-100% scale at the bottom. Such a scale would correspond to the 26-step
> patches that IJC prints for measuring. Since the patches only have numbers
> from 0-26, it would be nice to have these numbers under the ink curves .
> Instead, we now have to print a scale and using a pencil (!!!)  write down
> what percent on the grayscale each patch corresponds to, so we can find it on
> the percentage scale under the ink curves. Of course, you only need do this
> once and then stick it on the side of your monitor (!!), but it makes for an
> inelegant solution.
> 
A pencil! Wow, I'll have to scrounge around and look for one of those, what
a good idea. Doh :=) I've already got stickies all over my monitor - it's
gonna get lost!

I think I get your point though - set the ink limits, then print that head
firing just by itself by checking it's box on and print as a tweaked file
right? I don't see how you correlate a patch number to a point on the scale
however.

> Choosing an ink limit causes the effects of the whole ink curve to increase or
> decrease  - as if the whole thing went up or down the grid. It would't cause a
> redraw because of the logic of "limits" as it applies to inkjets vs prepress
> (as I discussed above).  Changing ink limits after you have arrived at a good
> curve, gives you a chance to reduce or intensify an ink while maintaining it's
> "shape". This is almost an unintended "feature", very useful in many
> situations. IOW, besides the limiting function as I discussed earlier, these
> ink limits are an added control over and above the curves. An example would be
> the use of a toner. You decide how you want it to behave at each point in the
> grayscale, then adjust overall intensity using the limits. Very cool in
> pratice!
> 
Ok I think I've got that - if you want a cooler or warmer print color you
could just alter the ink limits a step or two for a given ink color. That is
cool, er excuse the pun, that IS a nice feature.

> Indeed. Apparently the "wysiwyg" intentions of the  preview that the scales
> give you, could not be materialized. As a result, these are useful as an
> overall "icon" of what curve corresponds to what ink. You pick a color from
> the Apple color picker to assign to each ink by simple eyeballing. It helps
> you only to know that this ink starts and ends here and to see it in relation
> to other inks. It has no effect on the profile you save.
> 
I understand there is no interaction, one could save a lot of time with an
accurate screen preview. I posted a method of doing this a while back that
involved reverse engineering a ProfileMaker Profile when trying to draw
curves with the traditional Epson driver. Dan Culbertson also posted a
similar method. I wonder if there is a way to do that here, hmmm...

> I agree.... but would like to see simple capabilities added, such as 90 degree
> rotations of the image, user-defined page size, typable image dimentions and
> printable area preview. But we are only in v.1xx. I am sure there is more to
> come down the line.
> 
No doubt these will come in time, nevertheless there is something about the
simplicity of the product as it now stands that I find tremendously
appealing. Too many of these printing apps/RIPs have all these "features"
that don't really work. File rotations seem especially problematic, I'm sick
of watching RIPs choke on file rotations. I can do that in Photoshop, no big
deal.

I will peck away at this profile/recipe whatever for the MIS FSN inks on the
7000. If I ever achieve any success is there a place to post them for others
to use?

Best wishes,

Nick Wheeler

Re: Ink Limits in IJC

2003-04-01 by Antonis Ricos

Nick,

my comments  follow  yours:


> That is actually the problem I am having at the moment, I'm setting the ink
> limits too high (choosing the wrong step obviously), and the ink is running
> off the paper :=) 

Setting ink limits for individual inks should be a simple procedure: you print 
the "unlimited" scales, you pick the darkest patch (going from left to right) and 
use the number for that patch as the limit for that ink. Check the manual for 
more details.

So, for example, as you look at a scale, especially of the black ink, it will get 
darker and darker until at some point it either won't get any darker, or it will in 
fact get lighter. Say, step 15 may read 1.64, step 16 may read 1.65,  step 20 
may be 1.65,  but step 26 may be 1.59. The correct limit would be 16.


> 
> It would seem that if the proper ink limit was selected then the proper
> crossover curves would all be close to the same height as at some point one
> would want to use 100%, at least in theory. 

The one ink that needs 100% would be the black and you can see how that 
curve reaches to the top right corner of the grid in a typical profile. Other inks, 
may or may not need to get to the top of the grid depending on how they 
actually print.
For a new user, it may be best to start with an existing profile and tweak. This 
way the basic shapes are already dialed in.




Yet a lot of the curves in the
> provided examples don't even come close to this, so I assume I'm not
> thinking about this properly.

I think you are doing fine. A given ink may not hit the top of the grid (i.e. 
maximum output) because of other reasons, such as the density of that ink 
may be too much for that particular part of the grayscale - or there is more than 
one ink running simultaneously.



> It would be nice to be able to quickly arrive at what a black black might be
> for a given ink set, I haven't figured that one out at all. 


That's not so easy!....    What blend of black and "cyan" or other ink positions 
gives max black varies from inkset to inkset and paper to paper. It's back to 
trial and error, I'm afraid. For FS-type blacks, for example, the less the cyan 
the better.  the same goes for the Original Piezotone black. 

If you flood the paper with cyan, you loose density. In the case of the piezo 
plug in and piezo pro software, according to Jon Cone, that's exactly what 
they did: put a lot of cyan and very little black. Their reasoning was that 
because their Black faded so catastrophically on its own,  they put a lot of 
cyan under it to hide the fade. So, you had two things fighting each other: a 
black ink that had dyes to make it blacker, but a software that was putting less 
of it down because that ink couldn't hold its original density!  Fortunately, there 
is now Museum Black which will reach above 1.66 on Hahnemuhle, similar to 
Ultrachrome Black, when profiled properly. 


I am using the 4
> color in six position MIS Full Spectrum Neutral inks, care to hazard a guess
> at a good starting point for a black black with this type of ink set?

I think the included profiles should give you enough of a starting point for 
these kinds of issues.

 

> I think I get your point though - set the ink limits, then print that head
> firing just by itself by checking it's box on and print as a tweaked file
> right? 

Not exactly - but check if you can find the procedure in the manual. Your 
description above sounds confusing, and I'd rather not repeat what's already 
in the manual...




I don't see how you correlate a patch number to a point on the scale
> however.


Here is a guide:
 
step 2 = 4%
step 4 = 12%
step 6 = 20%
step 8 = 28%
step 10 = 36%
step 12 = 44%
step 14 = 52%
step 16 = 60%
step 18= 68%
step 20 = 76%
step 22 = 84%
step 24 = 92%
step 26 = 100%



> I understand there is no interaction, one could save a lot of time with an
> accurate screen preview.

May be you can suggest it to Bowhaus (software@...).



 File rotations seem especially problematic, I'm sick
> of watching RIPs choke on file rotations. I can do that in Photoshop, no big
> deal.

Very true...

> 
> I will peck away at this profile/recipe whatever for the MIS FSN inks on the
> 7000. If I ever achieve any success is there a place to post them for others
> to use?

Best thing to do is send it to me to upload to the Files section of this group.
And, btw, you can start with my 7000 profile that's included in the download, it 
shouldn't be that far off since the black is the same as yours.



Regards,

Antonis

Re: [Digital BW] Re: Ink Limits in IJC

2003-04-02 by Robert Morrison

On 4/1/03 10:00 AM, "Antonis Ricos" <antonisphoto@...> wrote:

> Choosing an ink limit causes the effects of the whole ink curve to increase or
> decrease  - as if the whole thing went up or down the grid. It would't cause a
> redraw because of the logic of "limits" as it applies to inkjets vs prepress
> (as I 
> discussed above).  Changing ink limits after you have arrived at a good
> curve, gives you a chance to reduce or intensify an ink while maintaining it's
> "shape". This is almost an unintended "feature", very useful in many
> situations. IOW, besides the limiting function as I discussed earlier, these
> ink 
> limits are an added control over and above the curves. An example would be
> the use of a toner. You decide how you want it to behave at each point in the
> grayscale, then adjust overall intensity using the limits. Very cool in
> pratice!

Hey now...its not an unintended "feature"...it was my idea and a damn good
one at that :-)  I felt it was important to have an easy way to have the
same shape and amplitude of curve do the same thing across various inks.
Before ink limits it was much more difficult to know how to set the
amplitudes of the various ink curves relative to each other...at least in my
opinion.  The feature also allows for those subtle tweaks that are
frequently necessary when you move from one paper to the next.  Paul Roark
has frequently reported problems with mottling with Museo using the Epson
driver.  IJC makes childs play of this...because it was a simple ink limit
problem in the midtones. The ease which you can tweak profiles for different
papers and really tune them in perfectly is one of the most powerful
features of the program and makes it much easier to use than RGB workflows
with the Epson driver.

Robert

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.