Fingerprinting Suggestions

[dimitri]

I fail to see any argument anymore in using density or LAB, or XYZ, or of the value of grey balance. If you have spectral data, you can convert to either. Density doesn't relate to appearance. LAB relates to appearance. If press operators want density, spectral data can be converted to density or TVI. Better still, have an automated closed-loop solution.

Metrics derived from spectral data are as linearly related to ink film thickness as density is. Who knows which is better though? I'd go with the appearance argument, and speculate that spectral data would be 'better'. We could control grey balance with CIELAB or CMY density values equally well.

All I'm trying to say is that you got to have an understanding of all these parameters and adapt, learn, and improve. A press can be calibrated to a proof (or the other way around) with density, CIELAB, CIEXYZ, grey balance, or dot gain... so I cannot really see an argument anymore.

 

[Alois Senefelder]

Brunner PCC

Brunner PCC

Hello gentlemen, yet more info on Brunner Systems



Regards, Alois

 

[gordo]

I fail to see any argument anymore in using density or LAB, or XYZ, or of the value of grey balance. If you have spectral data, you can convert to either. Density doesn't relate to appearance. LAB relates to appearance. If press operators want density, spectral data can be converted to density or TVI. Better still, have an automated closed-loop solution.

Metrics derived from spectral data are as linearly related to ink film thickness as density is. Who knows which is better though? I'd go with the appearance argument, and speculate that spectral data would be 'better'. We could control grey balance with CIELAB or CMY density values equally well.

Press operators don't want density values - they want ink film thickness values. Using a densitometer is, I believe, a faster way to get the information when measuring color bars compared with taking spectral readings and then converting them to density values in order to infer IFT. Also, I believe ink density is a non-linear response and when you convert in software it does not simulate the nonlinearities of the varying IFT.

best, gordon p

 

[meddington]

Hey there gordo,

Press operators don't want density values - they want ink film thickness values. Using a densitometer is, I believe, a faster way to get the information when measuring color bars compared with taking spectral readings and then converting them to density values in order to infer IFT.

If we're talking handhelds, the I tend to agree. An Xrite 418 densitometer will give a quicker reading than say a 530 or 939 spectro, and our Pressman always reach for the 418. That said, this doesn't seem to matter too much as we use an ATS (spectro) at the press console. Further, I get just as instantaneous readings as the 418 from my i1 spectro with BabelColor CT&A despite being tethered to a laptop.

Also, I believe ink density is a non-linear response and when you convert in software it does not simulate the nonlinearities of the varying IFT.

best, gordon p

I'm not following you here. Are you suggesting that density values from spectral data are out of sync with values from traditional densitometers? I haven't seen any evidence of this, but I think I'm misunderstanding you. ?

 

[dimitri]

I'm not trying to complicate the argument, rather un-complicate it.

I do not know of any press operators measuring mils on press. Does anyone? They care about IFT in a mechanical sense, and they look/hear at the plates and rollers, or use gauges to measure it, but when it comes to adjusting the press they use density. They also know by experience that 1.80 cyan density is too much; the smarter or more experienced ones know that it is too much because it translates in too much of an ink film.

A pure filter densitometer would indeed be the fastest way to scan a color bar. Most devices, spectrophotometers, spectrodensitometers, colorimeters, abridged spectrophotometers and so on... would fairly fast and easy convert a spectral reading to density. A spectrophotometer would be the slowest, but there are very few real spectrophotometers in pressrooms.

I do not exactly understand the comment about software converting density in a non-linear manner. Please, expand. The linearity (or not) should exist outside of the software, and would be apparent whether we use density or LAB metrics to infer IFT. Do you mean whether filter density would translate linearly to spectral density? In my experience it doesn't (I deal with press scanners and handhleds from different manufacturers)... and it is a hassle when you try to calibrate filter density to spectral density in the pressroom, but, each one by itself, would provide a good enough metric for press operators. Then again, which one is better and in what sense? That's too much of a scientific question for here and I have no answer.

Is density and IFT linearly related? I should have been more accurate in saying that filter density (, spectral density) or LAB metrics are close to each other with regard to their correlation with IFT. They tend to be linear over a certain IFT range and none is perfect. At higher IFTs each of these metrics would saturate.

There is research on the subject. Bassemir, Robert W, and Jean S. Lavelle published a 1993 TAGA paper on it “Colorimetric Parameters of Lithographic Prints at Various Film Thicknesses”, pps. 327 – 347 (attached doc is just my own graphing out of the original paper and would be slightly inaccurate). There is more work than just this paper.

What I'm trying to say, is that I hope that someone comes up with the perfect math formula for printing, but until then, we can make it work with a variety of ways.

Best,
-D

 

[gordo]

Are you suggesting that density values from spectral data are out of sync with values from traditional densitometers? I haven't seen any evidence of this, but I think I'm misunderstanding you. ?

I think Dimitri explained what I meant better than I could explain it. Unfortunately I'm at the end of my knowledge/experience on this aspect of the topic.

best, gordon p

 

[danremaley]

On a side bar, Brunner's gray balance is NOT based on color.
>>>>I know, he and Heid. got in a great tiff about it
He does not use colorimetry (page 5 from your PDF: "Colorimetry is absolutely unsuitable for controlling the printing process and detecting process faults in illustration printing.").
>>>>I know, never said he used colorimetry, but he does use gray balance along with TVI and SID. Have you ever been on-site running this systen - I have, at Quadtek - AWESOME. It has a 1 thru 5 star rating system, taken from that thing you dislike, the color bar!

Instead, he uses densitometry and assumes a connection to color. A connection that does not always occur.
>>>>ASSUMES? I DON'T THINK SO! Felix Brunner knows more about color than any human - alive or dead. -Dan

best, gordon p[/QUOTE]

 

[Deleted member 16349]

Not only is the press inconsistent but it seems general prepress knowledge is too.

I understand that there is a lot of experience here in dealing with prepress issues within the existing methods used to get jobs done but I have heard so many faulty views on the physics of the problem that I find it very depressing.

IMO there is a need to rethink the whole prepress approach to reproducing colour on press. It now seems to be a mess and this discussion from very experienced people is proof of that. The lack of clarity of views confirms the existence of inherent faults in the process.

I blame the graphic institutions for perpetuating fuzzy thinking.

 

[danremaley]

some color theory

some color theory

GATF designed a great little product called a proof comparator, it was designed to measure a proof's neutrality in the 25-50-75% areas with ONLY Y-M-C (black would only make it darker). With 0 and 100% + the 25-50-75% you can develop a gray balance curve. So the theory is if these areas are neutral then the entire scale is neutral. The same thing applies at press, if your neutral in the 25-50-75% the entire scale would be neutral.
Note that the standard deviation (best crew and machine) is +/- 2% dot change. This is critical because if your "casted" even as little as 2% at the start, (towards Magenta for example) at standard density, it won't be long before your 4% casted (2% start +2% std. deviation) and the color is horrible! Now some pictures won't be harmed because their values are saturated and away from the midtones. Ever look at a 40" press sheet and notice that most of the color was OK but a few photo's are "off" - the 'off' ones have values near the midtone, no matter the color. I'll attach a couple items to show this gray relationship and a note on an upcoming training event.

Dan
412.889.7643

 

[danremaley]

Not only is the press inconsistent but it seems general prepress knowledge is too.

I understand that there is a lot of experience here in dealing with prepress issues within the existing methods used to get jobs done but I have heard so many faulty views on the physics of the problem that I find it very depressing.

IMO there is a need to rethink the whole prepress approach to reproducing colour on press. It now seems to be a mess and this discussion from very experienced people is proof of that. The lack of clarity of views confirms the existence of inherent faults in the process.

I blame the graphic institutions for perpetuating fuzzy thinking.

There nothing to rethink, nothings changed- dots are put on plates, either film or CTP but the press IS where change is needed, so the standard deviation can be improved.
Are you saying that Felix Brunner's documented research, is "fuzzy"?
I believe he has done more than all of us put together!
How about the old PIRA Research Group from England - and the book Color Control for Lithography by Kelvin Tritton and Color Reproduction in the Printing Industry by Anthony Mortimer.
I'll let you read them, when you pry them from my cold, dead, hands. .

Dan

 

[meddington]

I know this is heresy, however, the 3/C gray patches in a color bar do not share the same screen tint builds as the same tone values in the live 4/C image area and therefore cannot reflect the color shift caused by a change in solid ink density. As a result, the 3/C gray patch in the color bar is fairly meaningless for that purpose.

By the same account, an individual 50% cyan patch doesn't represent the live image area either (unless you print a 50% cyan screen in the image area), but this doesn't negate its usefulness as a process control indicator. In fact, a change in solid ink density might go unnoticed on this patch both visually and slogged off as insignificant by way of TVI measurement, though the live imagery and 3/c gray may indeed show a cast. I've seen a lot a pressruns, and casted printing is casted printing...always represented in kind by the 3/c gray patch. Certainly valuable even if you never measure it....but I think you expected me to disagree.

On a side bar, Brunner's gray balance is NOT based on color.
He does not use colorimetry (page 5 from your PDF: "Colorimetry is absolutely unsuitable for controlling the printing process and detecting process faults in illustration printing."). Instead, he uses densitometry and assumes a connection to color. A connection that does not always occur.

best, gordon p

Densitometry has its place as a process control metric. Colorimetry has its place as a defining perceptual metric. I think there can be some latitude to be had by crossing over to some degree, but attempting to switch the the roles of the two can lead to trouble.

Regarding the Brunner PDF, I'm not all that comfortable with the sentence: "Instrument Flight* also precisely measures the gray balance and makes any corrections needed with the patented
regulation algorithm. It doesn’t need colorimetry to do this", and then later..."There is
no mathematical association between colorimetric measurement values and the
process variables ink film thickness and halftone area coverage used for color
control". Now that last sentence I agree with, which is why I'm uncomfortable with the first. Perceived gray balance is dependent on a lot of factors, not the least of which is the substrate's white point. An arbitrary target of gray balance that isn't substrate relative can lead to an perceived cast even when the target is achieved. This can happen with either densitometry or colorimetry, but I'd rather start with a metric that more closely aligned to human perception. Further, the paragraph under "Colorimetry is not picture pertinent" implies that densitometry would be, which is also false. Moreover, though delta Eab is less than suitable for gauging perceptual differences, other formulae are, such as delta E CMC, 94, & 2000. But I didn't mean to make this into a critique.

 

[gordo]

>>>>I know, never said he used colorimetry, but he does use gray balance along with TVI and SID. Have you ever been on-site running this systen - I have, at Quadtek - AWESOME. It has a 1 thru 5 star rating system, taken from that thing you dislike, the color bar!
Instead, he uses densitometry and assumes a connection to color. A connection that does not always occur.
>>>>ASSUMES? I DON'T THINK SO! Felix Brunner knows more about color than any human - alive or dead. -Dan

First off I am a big fan of System Brunner - especially the idea of the star rating system.

Yes, I have been on-site at several locations with presses running System Brunner. That being said. I've also seen System Brunner fail when the Yellow printer gets contaminated and color in presswork shifts but System Brunner reports all is well because the relative density is maintained.

To my knowledge there has never been a study of the relationship of gray balance in the color bar vs gray balance in images (done with today's typical separation methods) in presswork. If something has been published, please provide a reference. All the literature I've seen on the subject takes experience from scanning and proofing and applies it to the press - assuming that what works in one area should work in the other. Until someone can provide some real world press data to support the claims made by Brunner or other gray balance in the color bar advocates, I have to rely on my experience doing daily production inside a sheetfed shop and from what I know about the behavior of screen tint builds. And they tell me that there is no useful relationship between gray balance in a color bar and gray balance in the live image area of a press sheet.

best, gordon p

 

[Deleted member 16349]

There nothing to rethink, nothings changed- dots are put on plates, either film or CTP but the press IS where change is needed, so the standard deviation can be improved.

Dan

Let me see now.
You don't think rethinking is required on the prepress side.
Press manufactureres don't think rethinking is required on the press side.
That implies everything is perfectly OK.

The people in their particular fields can't imagine anything better.

Maybe that's the problem.

 

[gordo]

[SNIP] Note that the standard deviation (best crew and machine) is +/- 2% dot change. This is critical because if your "casted" even as little as 2% at the start, (towards Magenta for example) at standard density, it won't be long before your 4% casted (2% start +2% std. deviation) and the color is horrible! Now some pictures won't be harmed because their values are saturated and away from the midtones.[SNIP]

and

[snip]I've seen a lot a pressruns, and casted printing is casted printing...always represented in kind by the 3/c gray patch. Certainly valuable even if you never measure it....but I think you expected me to disagree.

Here is an example of what I've tried to describe. (Pardon the ugly face.)

Below is a 4/C image on the left and 4/C grayscale on the right.* These were converted from RGB using PShop defaults.
Below the "live images" are the 3/C gray patches.

I've split the images and grayscale patches in half (but the split is not oriented the same way**). I then added 3% Magenta to one half of the grayscale patches and also to one half of the live image areas.

It's pretty clear that the 3%M shift shows up in the 3/C gray patch. However, at least on my monitor, there is no apparent or equivalent color shift in the live image area. I doubt very much if you can see which section of the live image has had Magenta added and therefore should be suffering from a color cast.

Conclusion: On press, the 3/C gray patch and live image color do not have a useful or meaningful connection.

As I said, I have a lot of respect for Brunner's work***. However, when we actually put gray balance to the test on press at Creo when we were trying to quantify the color stability of Staccato FM screening compared with AM/XM screening when solid ink densities moved, the 3/C gray balance patches did not perform how advertised/promoted.

As a result, I don't really put much credence on what some expert says. Until I see the results of some graphics organization real and proper testing of this "axiom" to me it's just an industry myth.

best, gordon p

* Converted back to RGB in order to be embedded in this post.

** I did not split the live image areas using the same orientation (vertical) as the 3/C gray patches because I felt that if a color shift actually happened it would show up and reveal the orientation of the split - and hence prove me wrong.

*** I have quite a library on System Brunner, however, in all the material I have he does not show or describe the CMYK separations he used to demonstrate his ideas. Perhaps the default separation methods used today are very different than those used when he was doing his tests. I don't know.

 

[meddington]

I've split the images and grayscale patches in half (but the split is not oriented the same way**). I then added 3% Magenta to one half of the grayscale patches and also to one half of the live image areas.

It's pretty clear that the 3%M shift shows up in the 3/C gray patch. However, at least on my monitor, there is no apparent or equivalent color shift in the live image area. I doubt very much if you can see which section of the live image has had Magenta added and therefore should be suffering from a color cast.

Conclusion: On press, the 3/C gray patch and live image color do not have a useful or meaningful connection.

Hi Gordo,
Nice visuals, and there’s no disputing that the “cast” imparted by the 3% increase in magenta is difficult to detect on your samples. I’d take this further and state that the higher level of GCR used in the separation, the less effected the imagery is by deviations such as the one you outlined above. In fact, with strong GCR, one might need to have a midtone magenta shift of 10% or so to get the same visual impact of a 3% shift in a 3/C gray patch, but I certainly wouldn’t advise that magenta TVI be allowed to drift that much on press! Does this indicate that Magenta TVI has the same disconnect (or more) with the live image? IMO, no.

Moreover, the example you provided is dealing strictly with a change in midtone TVI...something that we can all agree is not a metric that a press operator has an effective direct control over once on press with a job. During make ready, the press operator will be dealing with adjustment of ink film thickness by way of solid ink density, adjusting ink keys based in the requirements of the imagery. One element your samples cannot show is the visual deviation associated with different ink films when making ready a job. Different ink films that result in the increased magenta TVI would likely show more profound differences in the imagery, perhaps a stronger difference at the shadow end and tapering off in the highlights (making your ears redder), which again, a synthetic example such as this can't show.

Lastly, the split image (without a reference proof) makes detecting the visual difference harder than it would if we were to simply apply a Photoshop curve increasing 3% magenta over the entire image, then toggle that curve on and off...much like for example, comparing a press sample to a reference proof. As you know, we humans have poor color memory.

So, I agree that a CMY gray patch will certainly be more sensitive to fluctuations that general imagery, but I disagree that there is no meaningful connection.

 

[gordo]

@ meddington

meddington --> Nice visuals, and there’s no disputing that the “cast” imparted by the 3% increase in magenta is difficult to detect on your samples. I’d take this further and state that the higher level of GCR used in the separation, the less effected the imagery is by deviations such as the one you outlined above.

gordo--> Yes. Put another way, the greater the disconnect the 3/C gray in the color bar has with the live image area.

meddington --> In fact, with strong GCR, one might need to have a midtone magenta shift of 10% or so to get the same visual impact of a 3% shift in a 3/C gray patch, but I certainly wouldn’t advise that magenta TVI be allowed to drift that much on press!

gordo--> Neither would I, but that's not the point.

meddington -->Does this indicate that Magenta TVI has the same disconnect (or more) with the live image? IMO, no.

gordo --> Not sure I understand you there. Magenta TVI in the live image area will respond the same as the same tone area in the 3/C gray patch. My point is that it is the 3/C gray target in the color bar that is disconnected from the live image area. As I think you agreed with in your first paragraph above.
If the live image area contains areas of 50C, 40M and 40Y then those areas will respond to SID shifts the same as the 3/C gray patch in the color bar. However, I don't believe that 50C, 40M and 40Y occur very often, if at all, in RGB to CMYK separations. And it's not just the fact that black is added, but also because the remaining CMY end up being placed at a higher tone value than the reference 50C, 40M and 40Y in the 3/C gray patch in the color bar. Being at a different, higher, point in the tone curve means they gain less when SIDs are increased. Put another way, the 3/C in the live image area is more stable than the 3/C in the color bar.

meddington --> Moreover, the example you provided is dealing strictly with a change in midtone TVI...something that we can all agree is not a metric that a press operator has an effective direct control over once on press with a job. During make ready, the press operator will be dealing with adjustment of ink film thickness by way of solid ink density, adjusting ink keys based in the requirements of the imagery. One element your samples cannot show is the visual deviation associated with different ink films when making ready a job. Different ink films that result in the increased magenta TVI would likely show more profound differences in the imagery, perhaps a stronger difference at the shadow end and tapering off in the highlights (making your ears redder), which again, a synthetic example such as this can't show.

gordo --> This starts to get a bit fuzzy. But you are correct, I cannot simulate ink film thickness, and I did use TVI to simulate a 3% Magenta shift. But I'm not sure it matters at all since I applied the same shift to both picture elements. If they share a common response - which what is argued as the reason for the 3/C gray patch in the color bar - then I think that the shift I applied would have effected both equally. But it didn't (which is my point).

You bring up two points that I have seen research on but that has not been published (that I'm aware of anyway). First is that TVI is measured as a relationship between a solid and a tone area. It is quite possible for a solid to be increased and the measured dot area to remain the same or even go down.
Your second point is ink film thickness. I would argue that the press operator is not adjusting ink keys based in the requirements of the imagery, but on a need to try and have the ink film the same thickness throughout the tone range. I.e. The ink film on the halftone dots (lots of small solids) be the same as on the larger solid areas.
So, another reason for the lack of color shift in the live image area compared to the 3/C gray patch is not just related to TVI but to the fact that the smaller dots in the live image area cannot take on as much ink (and therefor color saturation - not TVI) compared with the larger dots in the 3/C gray patch.

It would be great to see some research on this.

meddington --> Lastly, the split image (without a reference proof) makes detecting the visual difference harder than it would if we were to simply apply a Photoshop curve increasing 3% magenta over the entire image, then toggle that curve on and off...much like for example, comparing a press sample to a reference proof. As you know, we humans have poor color memory.

gordo --> I used the split image for two reasons. First because it shows that the strong hue shift in the 3/C gray patch is not mirrored in the live image area - otherwise you'd see the edge of the magenta shifted area. Secondly, it mimics what press operators do when they check critical color areas. They overlay the cut press sheet over the proof. If they cannot see a color shift or transition from proof to press sheet then they have a match.
I've only seen press operators toggle between press sheet and proof when they're looking for artifacts - not for color.

meddington --> So, I agree that a CMY gray patch will certainly be more sensitive to fluctuations that general imagery, but I disagree that there is no meaningful connection.

gordo --> I don't see how you can agree and disagree at the same time. But I'd be happy to accept a beer and a more complete explanation if we ever meet!

best, gordo

 

[TerryWyse]

Press matches proof, not the other way 'round

Press matches proof, not the other way 'round

Gentlemen and fellow "Lithographers"

I maybe just a "Fossil from the Stoneage" but I Repeat -----

THE PRESS "Still the largest variable, the press system's Ink, Water,Paper,Blankets, Pressures, Temperature and Humidity all impact the print quality"

That's exactly right....and why the press should never be used to establish the standard printing (proofing) condition for a company. The "standard" should be be based on the most stable and repeatable process in the plant and that's generally going to mean the proofing system. The job of the proofing system is to emulate a standard specification that is achievable by the pressroom. The key here is choosing the correct specification for the type of printing you do (don't choose GRACoL if you're printing on newsprint!).

Bottom line, the press is much more "flexible" than a proofing system....the press can easily vary it's "standard printing condition" to suit the job at hand, and it can do it mostly on-the-fly.

A proofing system on the other hand is NOT very flexible but it CAN be deadly accurate and, more importantly, consistant. The press can be changed in "real time" by simply tickling the ink keys whereas a proofing system might take a 1/2 day to a day (or more) to be altered to suit a different printing/proofing condition. In my opinion, using the ever-variable press to establish an in-house proofing standard is a recipe for distaster...or at least frustration.

Regards,
Terry

 

[meddington]

Hi Gordo,

I think we may be diverting a bit far off topic to the original post, but debating with you is always fun and educational, so here goes...

gordo --> I used the split image for two reasons. First because it shows that the strong hue shift in the 3/C gray patch is not mirrored in the live image area - otherwise you'd see the edge of the magenta shifted area. Secondly, it mimics what press operators do when they check critical color areas. They overlay the cut press sheet over the proof. If they cannot see a color shift or transition from proof to press sheet then they have a match.
I've only seen press operators toggle between press sheet and proof when they're looking for artifacts - not for color.

Yes, that does happen, but it also occurs that the press operator has a single page proof that he lays up on a multi page press form. In this case there is no cutting of press sheets, and the image is compared 1 to 1, and IMO, overall deviations are more readily identified. I don’t wish to argue this point further as it seems rather petty given your example, but I think if you had provided “unsplit” imagery where a 1 to 1 comparison could be made, the deviation would have been more apparent.

Also, a solid field of color will always show a more pronounced visual change than a dynamic image when subjected to varying SID or TVI. If your face were a flat tint, I think we’d readily see the 3% magenta deviation (no, that wasn’t a threat ). For example, refer back to your face in Photoshop and take a sample reading off your cheek, then paint a large swatch with those CMYK values. Applying the same 3% magenta increase, you’ll certainly see this change quite readily, and nearly as profound as the 3/c gray. These values have a direct connection with the imagery (they’re taken right out of it), yet the visual difference is still more profound on the flat tint than on the image of your dynamic, yet dapper face.


All process control metrics that we are measuring press side will have a connection to the imagery, whether it be solid ink density, TVI, or gray balance. Your point is well taken that typical imagery is “protected” from such deviation, though some imagery will more readily show the connection than others. I consider a 3/c gray patch and early indicator...the canary in the coal mine. The connection to the imagery might not be profound with all separations, but its there, and has value for indicating the balance of the individual printing units. Whether or not a pressman needs to react to a deviating 3/c gray might be another point to discuss, as if there is still a close correlation between proof and press sheet, it wouldn’t be necessary.

 

[Deleted member 16349]

You bring up two points that I have seen research on but that has not been published (that I'm aware of anyway). First is that TVI is measured as a relationship between a solid and a tone area. It is quite possible for a solid to be increased and the measured dot area to remain the same or even go down.
Your second point is ink film thickness. I would argue that the press operator is not adjusting ink keys based in the requirements of the imagery, but on a need to try and have the ink film the same thickness throughout the tone range. I.e. The ink film on the halftone dots (lots of small solids) be the same as on the larger solid areas.

I have seen and still have some plots somewhere that show a clear relationship between ink film thickness (density) and dot gain. Of course this was from testing with fresh ink that had no water emulsified in it. I think it was done with a Little Joe machine.

On press things are a bit different. It is wrong to say that density is related to ink film thickness because the printed film has ink and water present. One could say that density is closely related to the ink pigment content in the printed film. It is closely related because the smoothness of the ink lay down affects density. A smoother film will result in a higher density than a rough ink film that has the same amount of ink per area.

So it is very important to think in terms of printing a film that has both ink and water. This is important to understand because it helps one understand some factors that affect dot gain and wet trap. So when one is talking about printing ink films, always remember that there is water in that film too.

I would say that on a conventional press, the operator has no direct control of the dots and for sure is not trying to print the same ink film on the dots as is on the solid. They have no control over this. They control dots indirectly by adjusting solid density ink films.

Dots do not have the same ink film thickness as solids. Dots are not small solids. Models that assume that are wrong. Think about this. You put ink on a 50% dot on the plate. Assuming that the same amount of ink per area has gone to that 50% dot area as would have gone to a solid area, then after mechanical dot gain, the ink film thickness must be less than the ink film thickness on a solid due to the squeezing of ink out from the original dot circle.

This thinning out of the ink can cause a hue change. Also the thickness of the ink on the dot have very little affect on dot gain (TVI) but spreading of the ink into the adjacent non image area around the dot has a very great affect on TVI.

If an image was generated from some data from an original test form that was run to specific solid densities, the press operator is just trying to duplicate those conditions by running to the same solid densities. The hope is that the image in line with those colour bars will reproduce predictably. Some times they don't.

Since the printed ink film has both ink and water in it, we get to the problem of density control and control of dot gain. On conventional presses, the operator has enough problems just trying to maintain density let alone any issues about water content in the printed film. So after he gets to the target density, he may at times be running slightly more water and at other times slightly less. He does not want to get into the problem of chasing the balance. And with shorter run, he has not time to spend anyhow.

Here is where the potential benefits of positive ink feed comes in. Since with positive ink feed, the ink portion of the ink film is much more consistent. It becomes very much easier to make adjustments in the water content because adjusting the water will not affect the ink content. Increasing the water content might be useful in making adjustments to dot gain, on the run since it is rational to think that a thicker and less tacky ink film will spread out more. Water in the ink reduces tack. Laboratory tests for ink tack do not represent conditions on press. It also might help control wet trap. By adjusting the water content of the first down and second down inks slightly, wet trap might be adjustable while you run.

These would be very interesting issues to study but one needs positive ink feed to move forward. Oh shoot, I forgot, the press manufacturers don't think that rethinking the press design is required.

 

[Greg Imhoff]

Balancing Color Bars and Images on press? OK

Balancing Color Bars and Images on press? OK

We measure and control both the color bar and full image on virtually any press. This technology is proven in the demanding Securities sector so we are now qualifying commercial beta sites and welcome inquires.

Process automation is one key to consistent production quality. So if printing a good color bar but with a poor image, we see correct or flag. If printing a good sellable image but with a poor color bar likewise we correct or flag. We automatically close the loop to the operator console and to the ink keys.

Snip Edit in Gordo >> “ … I've also seen System Brunner fail when the Yellow printer gets contaminated and color in presswork shifts but System Brunner reports all is well because the relative density is maintained. … To my knowledge there has never been a study of the relationship of gray balance in the color bar vs. gray balance in images (done with today's typical separation methods) in presswork. If something has been published, please provide a reference. …” <<

From first sheet forward we intelligently measure and control all pixels, Densities, and LCHab targets, across and around each impression.

For more information see the attached 1 page PDF. Any questions please do not hesitate to contact me.

Sincerely,
Greg Imhoff
Director EPG Color Solutions
(800) 394-7130 office
(708) 557 – 2021 cell
Essex Products Group | Integrated Color Control Systems
[email protected]