Something better than density?

General Color & Graphics Press Calibration & Profiling
1

I’ve been playing around with calibrating and profiling our press a couple weeks ago and was wondering if there’s a better way to determine what ink densities to use.

Right now we’re using “standard” densities but could I use a similar procedure as I use on the inkjet when determining my dark ink restrictions? (dark ink restrictions is what Onyx calls it) I would measure in Lch and then look for the combination of hue angle and chroma I like best and then use that as my density. Would something like that work? How would I prepare a test form for that?

That crazy idea got my thinking a little more and I wondered if I could carry it over to process control. Instead of reading densities to determine when I’m up to color, can I read in Lab or Lch instead? I’ve learned that on the inkjet reading densities to determine ink restrictions is basically useless and Lch provides far more information about the colour. Would that same idea hold for a press? Run to Lch values instead of densities?

If I can use something like Lch measurements, would that allow me to try different brands of inks using the same profile? Technically speaking I’d probably need a new profile but if I could get my 100% solids to the same Lch values, my color would be about the same I thinking.

Suggestions, comments, ideas on what I’ve said? Am I crazy or can some of this be done?

Thanks,
Mark

2

Well, I think that your idea is not wrong or bad, in fact, what you are thinking is similar to Heidelbergs proposal. They suggest colorimetric (spectral in reality) readings to control the press instead of pure density. Heidelberg Prinect Image Control is the ultimate tool for doing that. The interesting thing here, is the opposite proposal from Brunner Systems; they say that controlling the press with something different than density is useless; they even say that measuring color difference at press with LAB and delta E units is not effective, do so, they developed their own system of measuring color difference. Im not a press expert and I cant say how is correct, but in my experience using a bit of both systems, I prefer Brunner approach.

I suggest you surf their websites and have your own opinion.

systembrunner.com/

heidelberg.com/www/pages/con … entID=4831

Xavier

3

Hello List
It is helpful to make a difference between the evaluation of testprints for a press and the daily operation.
By measuring Lab / LCH you got very useful informtion about the color of your Ink.
Only with Lab-measuring you can e.g. see, if your cyan is more greenish or blueish according to standards like ISO 12647-2, SWOP. GRAcOL etc.
During testprints, it is useful to measure both Lab and density.
If one density fits best your Lab-expections, this is the density for your daily production.
But the density or the Lab-value of the 100% color is only one part.
Much more important are the complete tonal curves and the TVI.
You should have smooth tonal curves and similar TVI in CMY with black slightly over CMY.
As coming from prepress to colormanagement, I made a lot of press profiles and learned a lot about printing…
Now I say to potential customers, that it makes only sense to make a profile for the press, if the printing house can automaticly measure the controlbar for printing and has controlled curves for ctP wich results to smooth TVI-curves in the complete printing process.

If you are using GMB Profilemaker for profiling presses, you should make spectral measurements with measuretool.
Than you can make directly in the measuretoo an analysis of the tonal curves and TVI according ISO 12647-2 l.

If the curvers are not smooth, you should inform the printer, that the profile will not have optimal quality and he should adjust his process (especialy the CtP-curves)

4

Current press systems use ‘closed loop inking’. This maintains a certain density value with user defined upper and lower control limits. I believe System Brunner has the best approach. It opens and closes ink keys based on gray balance not solid ink density. What is more important to color printing than gray balance? Color comes from the color of the dot (density) and the size of the dot (TVI). Lab is so fine a measurement - try to find neutral gray as in L=50 a*=0 b*0 - you’ll hunt all day!
The ‘secret’ to color printing is to print to gray balance and use GCR to control it.
Dan

5

If we do spectral measurement, it is possible to calculate Lab-values, density and TVI from one measurement.
Than there are three main steps for the use of measurement-data:

  • process optimization from CtP to print (tuning the process to reach optimal densities, TVI and greybalance according ISO / GRACoL / SWOP

  • process monitoring / adjusting (is my process stable, or has something to be adjusted to print according defined tolerances ?)

  • closed loop printing (adjusting the ink keys for the actual job by measurement data)

For some steps density is fine and for other steps Lab is fine.
I prefer solutions with spectral measurement, which shows Lab. density and TVI, so I can choose, what I need for my actual step.

I use e.g. the new software basiccolor certify for process optimization ans process monitoring. For 500,- EUR I can use a EyeOne for measuring one strip and I get imedeatly an analysis for Lab-values for primaries and secondaries, and TVI curves with comparision to ISO 12647-2 targets.

6

At 3:09 AM -0800 3/9/05, janpeter wrote:

If we do spectral measurement, it is possible to calculate Lab-values, density and TVI from one measurement.

one set of measurements yes (TVI requiring more than one patch)

Than there are three main steps for the use of measurement-data:

  • process optimization from CtP to print (tuning the process to reach optimal densities, TVI and greybalance according ISO / GRACoL / SWOP

  • process monitoring / adjusting (is my process stable, or has something to be adjusted to print according defined tolerances ?)

  • closed loop printing (adjusting the ink keys for the actual job by measurement data)

For some steps density is fine and for other steps Lab is fine.
I prefer solutions with spectral measurement, which shows Lab. density and TVI, so I can choose, what I need for my actual step.

yes, I do too. Many instruments (all X-Rites now I think) measure only spectro, then do internal calculations. So all X-Rite densitometers are now spectro-densitometers.

Regards,

Steve

o Steve Upton CHROMiX www.chromix.com
o (hueman) 866.CHROMiX

–

Post generated from email list

7
  • process optimization from CtP to print (tuning the process to reach optimal densities, TVI and greybalance according ISO / GRACoL / SWOP
  • process monitoring / adjusting (is my process stable, or has something to be adjusted to print according defined tolerances ?)
  • closed loop printing (adjusting the ink keys for the actual job by measurement data)

  1. What exactly is the difference between ISO 1247-2 and GRACol/SWOP data files?

  2. Where can you get these reference files and/or how do you determine the ideal TVI curve for your presses.

  3. Should the dot gain I read from my inkjet ideally read the same as the press. Right now, I have my inkjets calibrated toward a GRACol reference, but the yellow dot gain from the inkjet reads 5% below the cyan and mag, yet the gray balance looks good on the proofs.

  4. Should I worry about using the Murray-Davies verses Yule-Nielson formula which according to an XRite manual tries to adjust dot gain for the type of paper being printed on.

Dave B

8

Hello Dave

  1. What exactly is the difference between ISO 1247-2 and GRACol/SWOP data files?

The ISO provides an international standard. Some national organisations
provides material (reference data, reference-prints, control strips
etc.) according the ISO 12647-2.

You can buy the ISO 12647-2 via iso.org or national resellers.

Reading the ISO 12647-2, you will find differences between a positice
platemakin process (mainly used in EUROPE) and negative platemaking
process (mainly used in the USA)

negative platemaking-process results to an higher TVI compared to
positive platemaking.

  1. Where can you get these reference files and/or how do you determine the ideal TVI curve for your presses.

In Germany, the federal associations of printers produced a lot of
material helping printeres and prepress standardizing their processes
according ISO 12647-2 / positive platemaking process.

At altonatestsuite.com you get:

  • Reference-prints for 5 different papertypes according ISO 12647-2
  • The data, which was printed
  • Characterization-data and profiles
  • The TVI curves in 5% steps for every papertype
  • A manual in german and english language

A free reference-file for coated-paper and free profiles and
characterization-data for all papertypes according ISO-12647-2 positive
platemaking is available at eci.org

For the USA GRAcOL delivers materials for working according ISO 12647-2
(negative platemaking process).

SWOP / TR 001 represents papertype 3 according ISO 12647-2

CGATS / TR 004 represents papertype 1/2 according ISO 12647-2

I dont know, if somebody is deliviring a free profile for TR 004. (at
profile central ???)

But if, you have a profiling software, you can download the
characterization-data at cgats.org

There you can also buy reference-data and prints.

  1. Should the dot gain I read from my inkjet ideally read the same as the press. Right now, I have my inkjets calibrated toward a GRACol reference, but the yellow dot gain from the inkjet reads 5% below the cyan and mag, yet the gray balance looks good on the proofs.

If you want to control your proof, you should do this with an spectro
photometer reading Lab-values and comparing the results with a
reference- based on profile (e.g. SWOP or TR004)

Dotgain and density is not working for evaluating digital proofs,
because the technology is completly different to offset-printing. (E.G.
pure offset-colors are always represented by mixed colors on a proper
calibrated and profiled proofing systems)

  1. Should I worry about using the Murray-Davies verses Yule-Nielson formula which according to an XRite manual tries to adjust dot gain for the type of paper being printed on.

As measuring density or dotgain makes no sense on a digital proof, both
formulas are not important for evaluating a digital proof.

:slight_smile: Jan-Peter

Dave B

Sent from colorforums.com
To start new topics, do not reply to this message. Cut and paste the forum address into a new email.
Return to www.colorforums.com for subscription administration
(c)2004-5 CHROMiX, Inc.

–

homann colormanagement ------ fon/fax +49 30 611 075 18
Jan-Peter Homann ------------- mobile +49 171 54 70 358
Kastanienallee 71 ------- colormanagement.de
10435 Berlin ---------

Post generated from email list

9

I totally agree with Jan-Peter regarding positive and negative film process, but I just want to point that now that the CTP is very spread, things are changing.

While using a CTP for making the plates, the difference between positive and negative platemaking is gone. With the CTP there’s one step less in the platemaking chain and there’s no dot-gain/dot-lose in the process. That’s the reason why now America has less dot-gain and Europe has more dot-gain, in fact now the TVI tends to be very similar in Europe and in America (if using the same conditions).

10

Hello Xavier, hello list

TVI and CtP is a very critical point.
Changing from filmbased workflow (CtF) to CtP should maintain, that the
resulting TVI is the SAME !!!
Otherwise you will get problems with reprints, if the former print was
done with CtF.
You will get also problems, if the customers expects higher TVI and
delivers data and proofs made for higher TVI.

Analysing printing-problems during change from CtF to CtP shows mainly
problems with wrong calibrated CTP-Systems.

Good CtP-Curves needs definitly a testprint. Only measuring direct on
the plate is not enough.

Measuring direct on the plate is good and necessary to see, if your
process is stable, but you cant predict, what TVI you will get in the
machine, or even, if you get the same TVI compared to CtF-production.

:slight_smile: Jan-Peter

xavier_DE0 schrieb:

[quote:f145e8d60c=“janpeter”]

Reading the ISO 12647-2, you will find differences between a positice
platemakin process (mainly used in EUROPE) and negative platemaking
process (mainly used in the USA)

negative platemaking-process results to an higher TVI compared to
positive platemaking.

[/quote:f145e8d60c]

I totally agree with janpeter regarding positive and negative film process, but I just want to point that now that the CTP is very spread, things are changing.

While using a CTP for making the plates, the difference between positive and negative platemaking is gone. With the CTP there’s one step less in the platemaking chain and there’s no dot-gain/dot-lose in the process. That’s the reason why now America has less dot-gain and Europe has more dot-gain, in fact now the TVI tends to be very similar in Europe and in America (if using the same conditions).

Sent from colorforums.com
To start new topics, do not reply to this message. Cut and paste the forum address into a new email.
Return to www.colorforums.com for subscription administration
(c)2004-5 CHROMiX, Inc.

–

homann colormanagement ------ fon/fax +49 30 611 075 18
Jan-Peter Homann ------------- mobile +49 171 54 70 358
Kastanienallee 71 ------- colormanagement.de
10435 Berlin ---------

Post generated from email list

11

I’m having a debate with my prepress manager about how TVI is measured.

I thought a 20% gain called for by GRACol/SWOP meant the densitometer would read 70% at the 50% patch. He says a 20% dot gain means you take 20% of 50% for a reading of 60% at the 50% patch.

I suspect this gets into using the Murray-Davies vs the Yule-Neilson formulas. Can anyone clear this up?

Thanks,
Dave

12

correct.

nope.

Depending on your densitometer, you can have it report either just dotgain or actual dot value at 50%. The former would read 20%, the latter 70% for a 20% TVI at 50%.

mike

13

An other link for downloading the most recent GRACoL caracterization data with and without UV filter is:

ipa.org/conferences/tech2005 … loads.php3

Xavier

14

One should note that this is not official “Gracol” data. This data is averaged from 35 sheets from one pressrun at one particular printer in the ongoing testing for Gracol’s new approach to define specifications, so it may not even adhere to the Gracol 6.0 specs, and from what I can see from the Vender Guidelines, some of the SIDS are higher than Gracol 6.0. At this point, the purpose of the data specific for the IPA Proofing RoundUp.

15

In terms of “making sense;” measuring density, no. Dot gain, however, yes. If you’re measuring dot gain with a spectrodensitometer, then the readings are valid and useful. Gain is gain irrespective of density when deriving gain from visual data. (Like the X-rite and Gretag units)

Of course the profile will describe the discrepancy in the relationship of D to Lab, and that’s the missing piece in this scenario.

Digital proofing certainly has thrown a wrench in the proverbial works! I’m amazed at the nonsense espoused by manufacturers who insist that (for example) running Magenta at higher-than-specified density will yield neutral grays. Anyone ever seen a GRACoL sheet? Every one I’ve seen is casted magenta in the gray ramp… I can’t figure out why they think this is alright!

16

Not sure I follow…dot gain is usually based on either the Murray/Davis or Yule-Nielson equations, both of which determine dot area as a function of density of a solid and density of a tint. Can you elaborate on “deriving gain from visual data”…are you refering to spectral data, XYZ/Lab or something else?

17

Spectral data… like the X-Rite 500 series does.

I was simply commenting that the dot area and therefore gain data calculated from spectrophotometric readings is valid, so long as paper and solid references are taken into account. Gain is gain, in this sense, regardless of the method or material of tints.

Perhaps, though, I misunderstand Jan. His comment that Density and Dot Gain readings “don’t make sense” in digital proofing is what I’m commenting about.

The calculated dot area (in stocastic or contone as in some digital proofs) is equivalent to actual halftone dot area (in halftone printing) because these measurements are equivalent to our perception of that simulation. In terms of RIP inputs, these kinds of measurements are acceptible for gain specifications. The calculated dot area in halftone dot digital proofs is absolutely valid; measuring this process is identical to measuring print.

Obviously density is less relevant; Lab analysis of the primaries would of course capture (and then some) the printing ink equivalents… What’s important is that the spectrodensitometer is taking this into account by referencing a solid and paper.

Forgive my long-windedness and feel free to correct me. I’m certainly not a color scientist, and my experience is only from practice. The theory is still muddy waters for me! I learn more from this type of a discussion than I ever could from a manual!

18

Agreed…gain is gain…and neither a densitometer or spectrophotometer is actually measuring dot size. both the 500 series and the 939 (which I have) measure spectral data…obviously, but by default, use the Murray/Davis equation to determine dot area. I would assume they’re converting spectral to density using a formula that I won’t (can’t :wink: ) comment on.

They make sense and the values are accurate, but they often don’t help when attempting to match color as they don’t take hue into account. You can’t really set up an inkjet printer by defining density/dotgain alone and expect a match to your target. Even if you could, keeping such a system in check by density/dot gain calibration would not compensate for hue shift, such as a colorimetric calibration would. This is less an issue with say a Kodak Approval, which is very stable being calibrated with mere density/dot gain values.

19

He he… don’t get me started on “N factors” :unamused:

Right on. I didn’t realize there were hue shift issues in ink jet. It makes sense, however, considering the radical effects of the paper in a system that relies so heavily on absorbtion… we have the luxury of a brand new Fuji proofer that has the ability to adjust solid ink densities; if it weren’t for the fact that proofs are leaving that machine for other vendors, I would just dial back the density to get where our inks are…

The colorimetric calibration acheived with icc profiles is the missing piece in our proof at this time, our inks have (thankfuly) very low hue error, but their grayness is not as good as some. Visualy it just looks like we started printing at 90% or 80%… easy to account for in an ICC workflow. None of this is to mention the discrepancy in paper vs primer white… that’s another thread altogether.

Once that piece is in the puzzle, I can start working on the GCR implementation. Neat fun!

Hope I didn’t hijack too bad. :wink:

20

_

When I look at proofs in office lighting (not 5K) they seem to cast magenta.