9 Replies Latest reply on Nov 1, 2013 4:37 PM by TurgayOrhan

# Ink curves ...

I have some questions about some general terms ...

First of them is ink limit ...

Let's say our ink limit for a particular ink 80% ... does this mean that if the input pixel value is 100% for this ink, then the printer will cover 80% of this pixel area on the paper with this ink?

Secondly, there is an another term ... "ink density",

What is the relationship between "ink limit" and "ink density"?

Thanks a lot ...

• ###### 1. Re: Ink curves ...

Those are press makeready terms, but TIL ( total ink limit ) can be foud in calibrating applications. The total ink limit is typically determined when calibrating a press sheet or any printed sheet. The colorimeter has determined that 80% is the TIL for a given media ( sheet ) on a given press.  It s considered the point where ink will start "puddling".  The ink density is a little different in that in the offset printing scenario, you might gain or lose pressure between blanket and plate in certai areas of the totla print area across the fountain of the press.  So, if you have 80% on the ends, but 70% in the middle or vice versa, then ink density is less or more in that given area.  This can be adjusted for, but could mean the blanket is toast.

• ###### 2. Re: Ink curves ...

'Ink limit' refers to the numbers. In the K-channel the ink limit may be 80 (80% of 100).

The Visual Dot Area (expression according to the Murray Davies theory ) is larger, for

instance 95% because of Dot Gain. The covered area is larger than the numerical value.

Similarly, the 'Total Ink Limit' refers to the sum of numbers of all four channels. For 330% or

340% 'TIL'  the actual coverage may be 400%.

Testing a printed target we may find, that "puddling" or bleeding happens at Total Ink Value

(here it's not yet a limit) of 340%. There is no need for a measurement, because the target

contains the numerical values for each sample or swatch (with increasing values for C,M,Y and K).

[All this based on ink jet adjustments and profiling experiences.]

Density means: negative logarithm (base 10) of percentage (ratio) of reflected light from an

area which is covered by ink. For instance for K (smaller D) and rich black CMYK (larger D):

0.1 (10%) is reflected:      D=1

0.01 (1%) is reflected:      D=2

0.001(0.1%) is reflected:  D=3

Unfortunately, the expression 'Ink density' is nowadays often used instead of 'Ink limit',

especially in Web discussions. A bad habit, in my opinion.

Common instruments, together with programs, measure 'Density', not 'Ink Density=Ink Limit'.

Pages 15-19 here (besides the numerous Web contributions) may shed some light on issues

like Density, Dot Gain etc.:

http://docs-hoffmann.de/a3gencolorhigh.pdf

For a scientific approach I would recommend all books by Henry R.Kang.

Best regards --Gernot Hoffmann

• ###### 3. Re: Ink curves ...

Thanks a lot jdanek ...

• ###### 4. Re: Ink curves ...

Gernot ...

I want to thank you so much for all your helps ...

• ###### 5. Re: Ink curves ...

I would like to give an example from the QuadTone RIP ... for better understanding the effects of different ink densities on the curves of inks.

The following image has ink curves for two inks ... ink limits are set to 100% for both of them ... and ink densities is 100% for black and 50% for cyan. Every other variable are zeroed.

And for four inks with different ink densities:

From these setups, I can see that ink density is for deciding in which tone the peak of the individual ink is placed. Right?

I can also see that ... with reduced densities, they have lower ink limits in practice ... even if their limits are fixed to 100.

But, I don't know ... when we add an ink with a lower density, why other curves are pushed to the right and cover only a portion of tones?

If you have time, I would like to learn ... how would you read these curves with the given limits and densities for each individual ink?

• ###### 6. Re: Ink curves ...

Orhan,

in my opinion these diagrams are showing the black generation by gray inks:

Black (B) and Light Black (LB) in the first case, Black and three Grays in the second.

Cyan, Magenta, Yellow refers to the colors in the diagram and, may be, to ink

cartridges in the printer for standard color printing. QuadTone is obviously –

according to Web information– a RIP for B/W printing.

For the first diagram the process is understandable. From left to right we have

the nominal ink coverage from 0% to 100% (without taking dot gain into account,

or this is hidden somewhere else).

From bottom to top we have the amount of LB and B. LB is exactly half as efficient

as B. Draw a straight line from 0,0 to 100,100. At 40% we have LB=80 which

contributes as 40. At 75% we have LB=50, effectively 25, and B=50, which sums up

to 75.

In the second diagram we have the transitions between LightLightGray, LightGray,

Gray and Black, whatever the names may be.

These curves are to some extent arbitrary.

In a modern inkjet we have Cyan, Light Cyan, Magenta, Light Magenta, Yellow,

Light Gray, Gray, Black (Photo Black or Matte Black).

CyanInk is built by C and LC,  MagentaInk is built by M and LM, with transitions as in the

first diagram. BlackInk is built by LG, G and B, as in the second diagram.

Black is built by ColorInk=(CyanInk+ MagentaInk+ YellowInk) and BlackInk.

ColorInk can be partly replaced by Blackink. This is called Gray Component Replacement

(light, medium, heavy).

The transitions are still somewhat arbitrary, but one goal is to achieve neutral grays for

all (visual) levels. Much more disturbing than moderate color errors are (even small) tints

in the grays.

A B/W print may have an intended cast like 'sepia' or 'cool', which is obviously achieved by

disturbing the balance.

The ink limits themselves are not very important. It's just necessary to avoid too much

ink on the paper. Excellent proofing paper can carry 400% (numerical value, command

value) without bleeding. I'm using less, about 340% because of faster drying. The only

negative effect is a slightly decreased Density (measurable density) for darkest black.

I hope this helps a little – and that I'm not wrong.

Best regards --Gernot Hoffmann

• ###### 7. Re: Ink curves ...

I think I got it ...

If we use a single black ink, then the our ink curve is the following (without any correction) ...

As drawn below, when we add the lighter black ink, it draws a straight line between 0 and the specified tone as the density of this ink (it will be 50 for this case) ... such that the specified ink limit can be reached at the specified density level.

Then, with a kind of transition algorithm ... it draws each ink curve for getting the effect of original dark black ink curve for each tone ... as you said.

From bottom to top we have the amount of LB and B. LB is exactly half as efficient

as B. Draw a straight line from 0,0 to 100,100. At 40% we have LB=80 which

contributes as 40. At 75% we have LB=50, effectively 25, and B=50, which sums up

to 75.

At the end, we have two curves ... and their combined effect is same with the original dark black ink curve ... but now we are using light black ink for lighter tones in place of dark black ink.

And, I think that the reason for why we want to use light black ink for lighter tones is that ... light black ink gives less grainy appearance than the dark black ink for lighter tones. Correct?

Is there any other reason for using light black inks for lighter tones?

• ###### 8. Re: Ink curves ...

Orhan,

I'm happy that I could make myself clear.

Yes, you're right - the use of light inks produces a higher geometrical

dot density (dots per area).

Years ago I had severe problems, color printing by inkjet, with ink stability

and bronzing. At that time the problem could be cured by very strong GCR,

which means, that neutral mixtures of CMY were replaced by K even in

light areas. Unfortunately this resulted in a visible 'pepper effect'.

Meanwhile printers, inks and RIPs have been much improved.

Best regards  --Gernot Hoffmann

• ###### 9. Re: Ink curves ...

All your posts are very informative for me ... thanks a lot.