Adobe Photoshop Training with Dan Margulis

Hello, colortheory group

i have been on the web for two days looking for information on black generation and gray component replacement. So far i have found descriptions ( including Monaco & GretagMacbeth's website and documentation ) but no real information on why one would use: UCR or Max black, or any of the variants in between. ( as in PM 4.1 GCR 1-4 )

my interest lies in high end inkjet printing and maximum quality.

i believe Andrew Rodney dropped a cryptic comment on this list a month or so ago stating that he liked to use different GCR curves, depending on the image.

What would that criteria be?

What are the pro's and con's, regarding GCR when printing photographs or paintings to inkjet or lightjet or offset?

thank you, for your consideration

John

____________________________________________________________________________

John,

Here are a few things related to UCR and GCR.

The main reason to UCR or GCR when printing images on a printing press is to make them print better.

The original purpose of UCR was to reduce (or remove) the amount of three colors (CMY) of ink on a printing press and put black back in their place. UCR is predominately affects neutral shadow areas of an image. When UCR is applied correctly it makes images print better on a printing press. Basically the image areas that have the CMY reduced can dry faster between each printing unit on a printing press.

The GCR technique affects neutral and color image areas. So neutral shadow areas can have the CMY reduced and black is added back and then color areas that are created with CMY and K have the CMY reduced (in the gray component of the color) and then black is increased.

In theory, if UCR and GCR is applied accurately to an image you should not see any visual difference between images with and without UCR or GCR.

In practice end-user use UCR and GCR to make images print better on a printing press and printers (service providers) use GCR so they have less waste when they begin printing on a press. There for saving time and money.

As a rule of thumb, I believe it is best to use as much ink as you can get away with with printing any thing. When using offset printing this is not always possible due to ink paper and press configurations. For example on a web press that prints pages very fast you need images to dry fast between each printing unit. In those cases UCR and or GCR is used.

However, using GCR with some inkjet and paper combinations turns out to be useful to help reduce problems with metamerisim. What happens when GCR is applied is that color areas with large amounts of color are reduced with black ink in those areas. This makes those color areas visually appear more stable under different lighting conditions.

My .02

Jim Rich

____________________________________________________________________________

John Wickham writes,

The list has several threads on this topic archived at

http: //www.ledet.com/margulis/ACT_postings/SeparationIssues/Separation.htm

Dan Margulis

____________________________________________________________________________

Grey Component Replacement started out as a facility used by drum scanner operators within the pre-press industry.

The theory is that the lightest of a three colour can be removed and replaced by black.

The advantages often claimed include the idea that by making it easier to obtain colour to colour register an image could appear sharper.

This was generally true

In addition it was considered that web-offset printers would save on ink costs because less expensive black ink was replacing more expensive colour ink.

This was most certainly true

However the idea did not gain universal acceptance because many considered that certain originals reproduced badly when using this technique

What exactly the benefits for ink jet are I would be interested to know and how exactly you can vary the GCR curve shape - presumably without altering the other three colours - and still obtain a decent print, escapes me entirely.

Perhaps someone might be mixing up GCR with UCR - which is something else entirely.

Richard Corbett

____________________________________________________________________________

on 12/30/03 4:23 PM, John Wickham wrote:

Then you don1t really need to be too concerned because you really want to send RGB data to these kinds of devices (unless you1re using them to simulate another CMYK device).

Unlike RGB, colors described in CMYK can be defined by many different combinations of CMYK ink.

With this process (which is dealing with separation of RGB to CMYK), you have a great deal of control over what inks get used in the generation of black and other colors. The 3 inks (Cyan, Magenta and yellow) that make up gray, are known as the gray component but this can be also created with black ink as mentioned and when you do substitute the three inks for black, this is called GCR. The basic idea is to use black ink to create neutrals instead of that other possible mix of CMY. The reason is you use less ink and end up using black ink which cost less than the CMY inks.

UCR is preferred by many printers because it only effects the neutral areas, and only gets started later in mid-tones and 3/4 tones. It allows a printer more control on press over color; usually for improperly separated images.

GCR takes UCR to another level and and will replace gray, even in areas that have some color. GCR usually starts just before Midtones. If the printer is very familiar with both GCR and UCR, and the Separations well made to reflect the printing process, GCR is usually preferred.

If you have a series of images that have a lot of neutrals (let's say a catalog of metal parts), you'd want to use a higher GCR (use more black ink to create the neutrals) which would insure a better chance that they will indeed print neutral. Images that are low-key with lots of shadow detail will need a profile with none or light black generation in order for detail to be retained. For mid-key images, light to medium is going to be better.

Andrew Rodney

____________________________________________________________________________

on 12/30/03 7:39 PM, Richard wrote:

I'd have to disagree with this. As an ex-scanner operator, I see GCR as more an extension of UCR than something entirely different.

I think of UCR as a relatively narrow "bandwidth" technique that primarily affects neutrals and near-neutrals (browns for example) where amounts of CMY are replaced with K. GCR merely extends this same technique into colors where the "graying" component or contaminating color (Cyan in Red for example) gets reduced or replaced with K. GCR is simply a broader bandwidth variation of UCR for the most part.

If there's a key difference, it's that UCR is primarily used to keep the overall maximum density "up" while at the same time keeping the total ink limit within the bounds of the target printing process. Bottom line is that Black ink alone does a better job of making black compared to increasing amounts of CMY.

Which, in reference to inkjet printers, is all pretty much moot. If you're using a CMYK RIP for your inkjet printer, I'd suggest using as much GCR (and start it early) as you can get away with while at the same time being careful to avoid the "graininess" that excess amounts of black ink can produce. You'll find with certain printers and ink types (dye vs. pigment) that the amount of GCR and how early you can start it can vary greatly.

My .02 along with Jim's .02 = .04

Cheers,

Terry

--

__________________________________

WyseConsul

Color Management Consulting

v 704.843.0858

__________________________________

___________________________________________________________________________

GCR is the removal of Y-M-C inks and replaced with black. Have you ever tried to print a 4/color Black & White? Without GCR it can't be done. On press the standard deviation of TVI (Tone Value Increase) is (+)or (-) 2% of all colors. A neutral color with a 2% shift - towards Mag. for example - will produce a Mag. "casted" gray. On the high end scanners GCR was named as a percentage of replacement. Example: 50% GCR means - 50% of the "graying component" is replaced with Black. 100% GCR means 100% of the graying component is replaced with Black.

The reason GCR isn't used mor e, is because printers lack the tools to measure and control color on press. The Heavy Black, GCR scan is exactly opposite what they are usually printing - a "ghosted Black". Since the ghosted black has little to do with the image it can be printed heavier to accent the black type. If you print a GCR image heavier in the black, the TVI will muddy the color. This means you need to measure gray balance and TVI at press - and what do printer's measure? Solid ink density! We print dots not solids!

The secret to quality 4/color printing is GCR and gray balance measurement at press. At our website <www.gain.net> under process control I have a PDF showing the differences. Did you know that Xerox uses GCR on everything that goes through its copiers?

Newspapers must use GCR because of it's TVI standard deviation is so high +/- 4%. With GCR even deviations this high won't "change" the color at press.

Try this, convert the same image, RGB to CMYK. Do one with SWOP UCR and one with SWOP Max. GCR. Open both images and "push" the Mag. 4% in the midtones on each image - check out the results!

The UCR image is terrible - the GCR image hardly changes. Technically, you could do the same for Y-C. The impact would be if you changed the Black 4%. So is it easier to watch one color BLACK (GCR) or 3-colors Y-M-C(UCR).

Dan Remaley

Process Control Mgr.

Graphic Arts Technical Foundation

412.741.6860x450

___________________________________________________________________________

Dan Remaley writes,

No. These were all proprietary algorithms in which the numbers had meaning only to the manufacturer. No sane method of separation would remove the same percentage of graying component from 10c6m6y as it would from 60c50m50y. Similarly, the only meaning of "100% GCR" is that it's considerably more than "50% GCR". If 100% of the graying component was actually being removed, there wouldn't be any CMY ink at all in the deepest shadows, which doesn't work.

Very true.

That's for sure. But we live in the real world. If printers would get their process control act together, there would be a lot more of a case for more use of GCR. However, we may not know who the printer is, and/or we may not know how good his process control is, in which case we have to consider the pluses and minuses.

1) If the magenta is unexpectedly +4% in the midtone and you're using a skeleton black, there will be a color shift. If the image is a fleshtone, it may become unacceptably beet red. If not, it's probably OK.

2) If the magenta is unexpectedly +4% in the midtone and you're using a heavy black, there will be less of a color shift although it will most certainly still be there. If the image is a fleshtone it will be better than in #1 but it still may not be acceptable.

3) If the *black* is unexpectedly +4% in the midtone and you're using a skeleton black, if there's critical shadow detail it will plug up. If not, the picture is likely to look *better* than the proof--more contrast.

4) If the black is unexpectedly +4% in the midtone and you're using a heavy black, the average picture will turn to mud and be completely unacceptable. The only ones that won't be are the ones where you're looking for a duller and less saturated look as when the main object is neutral and you don't care what color the background is.

#4 is therefore a far more serious problem than #1. Plus, it is far, far more likely that the black will run too heavily than the magenta. Printers are always jacking around the black ink density to try to control the look of the type.

If you don't know who's going to print the job, therefore, the decision is a no-brainer. You have to take a calculated risk either way. Either you're going to open up the possibility of #1 or #4. So, look at the image. If it's one of those rare images where #1 is a bigger potential problem than #4 is, go with a heavier GCR. Otherwise, stick with the prudent method. Skeleton blacks are standard for a reason.

None of this applies, of course, to a desktop or any other printer that isn't a press, or to the situation where you know the printing company well and know that they're good.

Dan Margulis

___________________________________________________________________________

Couple of comments:

I agree with Dan Remaley in principle but the only scanner I ever ran that used "percentage of replacement" as a method of setting GCR was a Crosfield Maganscan. The Screen and Hell scanners I ran just had a knob you twisted but you had to go to a mounted gray scale(!) to figure out how much GCR you were actually getting. Then again, that was 10 years ago so maybe the newer scanners (Hell 3000 series?) let you call it out by percent. (Age creeping up on me again!). And let's not even get started on PCR, ICR, ACR .... :-)

I think Remaley's right on what the percentage of GCR actually meant but then there's also the starting point of GCR/K. So in Dan Margulis's example of 10c6m6y and 60c50m50y, it wouldn't likely take the same percent out because the GCR is probably not getting that far into the highlight. I personally never like to see the black start below about 25% or so (fleshtone territory) but I think if the black started at 0% (insane for a press), it should indeed remove 50% of the graying component. But that kind of K start point would never happen in the real world of offset printing unless it's something specialized like full-range 4c B&W halftones (huh?).

It always seems this UCR/GCR discussion turns into an either/or thing when MODERATION should be the word of the day. Give me a separation that uses about 30-40% (Medium?) GCR with a K start point of 30-35%, 300% total ink and I'll make good separations all day long. And the pressroom will never know I'm using GCR to help them out.

One last pet peeve: why is it that we always have to ASSUME there is no process control in the pressroom when this should be an EXPECTATION instead? It's always kind of irked me that "we in prepress" have had process control for DECADES and yet we're still afraid to expect the same from the pressroom because, God forbid, we should offend the pressmen. I always hear "well, look at all the variables in the pressroom, blah blah..." which is EXACTLY WHY the pressroom should be the ones screaming the loudest for good process control (and not just measuring density!).

[What's that Nurse Rachet? You say it's time for my medication? OK.]

I feel better now,

Terry

___________________________________________________________________________

I have observed that, regardless of how you configure your spearation with regard to GCR and UCR, once the "separation" is processed through the driver for the inkjet, it is reseparated to configure the inks for that inkjet. The inkjet printer generally applies its own UCR to optimize total ink coverage and ink drying for that ink jet model. This may not be the case with each workflow or inkjet printer, but I suspect that most configurations work this way.

What have others observed?

Steven Barton

____________________________________________________________________________

On 12/31/03 Steven Barton wrote:

Now to me that's a brand new observation. It took a long while for me to realize that my Epson printers were RGB, still longer to figure out which options to choose for printing (Epson is no help; their manuals are bare-bones). Now that I'm correcting in CMYK then converting back to RGB for printing, I wonder how the problem you described is affecting my output. There's no visible problem, but yours was an intriguing discovery and I'm anxious to see what kind of response you receive. For that matter I now wonder if it really matters which CMYK profile is active when I convert to CMYK in the first place since the files will not be sent to a CMYK printer. Probably doesn't, but it was fun figuring out how CMYK profiles work and how to write custom profiles.

Howard Smith

____________________________________________________________________________

Terry Wyse writes:

Printers do whatever they do because enough customers accept it to keep them in business. If you don't like the odds, you could talk to the customer about finding a different printer. If enough customers insist on it, they'll do it.

This is a classic idealist vs pragmatist question (as is the thread about photographers who want everyone to get their exposures right). Everyone agrees that it would be nice if everyone paid attention to what they were doing. You can "expect" anything you want, the question is what you want to happen when Murphy's Law kicks in. Do you want to point fingers and say "I told you so", or do you want to stack the deck in your favor? Either position can be defended (and often is, at great length, but we've heard it before).

And if the customer insists on having it printed by the lowest bidder, there's not much you can do about it (that's another classic argument that's not going to go away).

___________________________________________________________________________

From: Terry Wyse

Would'st care to indicate exactly what variables - apart from ink density, C to C register and dot gain - you would care to see controlled in the press room.

Richard Corbett

___________________________________________________________________________

on 1/1/04 8:16 AM, Richard wrote:

Gray balance, dot gain and density. In that order.

Terry

___________________________________________________________________________

on 12/31/03 8:17 PM, howller2 wrote:

While this may be true of OS-level printer drivers (profile as RGB and then the driver performs a "mystery" conversion to CcMmYKk or whatever), this is not true of proofing RIPs (ColorBurst, Best, et al). With a good RIP (and profiling package) you may have control over all aspects of the conversion including light/dark ink conversion, individual ink channel limits, total ink limit, UCR/GCR, linearization, etc.

Cheers and Happy New Year!

Terry

--

__________________________________

WyseConsul

Color Management Consulting

v 704.843.0858

__________________________________

___________________________________________________________________________

On Wednesday, December 31, 2003, at 11:26 AM, Dan Margulis wrote:

What would the reasons for a printer to call out for a lighter GCR in SWOP when using a Web Press? As opposed to Medium.

Is it to speed the drying as I believe I understood it? Does this mostly benefit the pressman or improve most images in most cases with regards to color shifting?

And what primarily will one see if changing from Medium to Light GCR? Or better still, what precautions in what types of images would one want to look out for.

HNY,

Doug Walker, FP

___________________________________________________________________________

Subject: Re: GCR

Doug,

One reasons I have found since the beginning of making color separations with Photoshop is to use the GCR option of light instead of medium is because you have less problems with black in the midtones (gaining on the black plate) when printed on a press.

As for changing the setting in GCR setting in Photoshop and seeing different results, say on your computers monitor, is that there will be very little visual difference. The color separation options in Photoshop were created that way.

Hmmm.... There are a few situations when you might or might not want to use GCR. One that is obvious is screen shots. I have found that using Heavy or Max GCR reduces the 3 colors of CMY and allow black to be the predominate ink. This kind of situation makes neutral screen shots (or images with a lot of neutrals print easier and in some cases better.

The down side of too much GCR is that your black image areas can suffer by not having enough ink to get a rich shadow. Though that can be overcome by using UCA (undercolor addition) in the neutral shadow areas.

Jim Rich

___________________________________________________________________________

I read some excellent replies to this question, but they all related to offset press conditions. John also asked about inkjet and Lightjet output.

Of course the answer is to give the Lightjet (also Lamda, Chromira or any other photographic output) RGB files - no GCR, no UCR, no K, ever. It could only hurt.

Also, certain inkjet printers and rips work best from RGB files. The better inkjet rips that support CMYK also provide control for these issues within the rip, so it shouldn't be a question on the file prep end. Besides, many of the offset press problems these tools address, like registration and TVI, don't even exist with inkjet printing.

Happy New Year Everyone,

john castronovo

tech photo & imaging

___________________________________________________________________________

This is not a GATF position, it is mine from 30 years of 'color matching' experience - working at a high end color trade house. Here is my reasoning for a higher percentage of GCR. I'm speaking of the publication advertising industry where 'color' is extremely important (hair color for example). When color is critical and has to 'match' a proof that has been color corrected several times, I don't want the pressman to 'decide' what color is important to 'him'. With a GCR scan he has little chance of changing the color at press. Yes, he can run the black too high and "muddy" the picture - but at least he, and everyone, will notice - and take corrective action (lower the black!). I will provide him with a sep. that has the correct amount of black gain removed to standard SWOP or GRACoL print conditions. Without GCR the color is changing constantly throughout the press run - think anyone notices that? The pressman decides which color 'he' likes the best - it may not be yours!

A recent GATF/NAPL study revealed the number one complaint of print buyers - "inconsistent color" was the answer! GCR prevents this. I think we would all agree that gray balance is important in photography, scanning, and proofing - so how many printers measure gray balance at press? Is it in their color bar? On their proofs? So how do we control color on press?

Dan Remaley

Process Control Mgr.

Graphic Arts Technical Foundation

412.741.6860x450

Print: The Original Information Technology at <www.gain.net>

___________________________________________________________________________

Terry writes,

Yes, but since it doesn't start at 0%, which would certainly be insane, the percentage of gray component actually removed is never the same for any two given darknesses, which is all I was trying to say. If you start generating black at the 25% point, then a 24% point has 0% GC reduction, a 26% point has 1% GC reduction, a 35% point has 10% reduction and so on. The darker the color gets, the higher percentage of gray component is removed. So, saying "50% GCR" doesn't have a meaning unless one also says at what darkness point the black generation begins, and at what darkness point it reaches the level of eliminating 50% of the gray component.

Nobody has assumed anything of the kind. Dan Remaley's and my examples talked about 4% variation in specific inks. This variation is within many industry standards, yet it's enough to ruin some pictures, and more than enough to justify a conversation about what GCR to use with quality-oriented printers, let alone bad ones.

*My* pet peeve is that people are so anxious to blame the printer for their own ignorance of color realities that they refuse to take obvious safety measures in their workflows to protect themselves.

I have a CMYK image with lots of subtle colors the purity of which I am anxious to retain. It's being printed on #5 web sheet by a printer I don't know well. Should I use extra GCR?

To answer, you have to ask what happens when things go right, and when things go wrong, and what the gains and losses are.

I. IF THINGS GO RIGHT.

If I use extra GCR, I expect absolutely, precisely and exactly zero benefit. If the printer's process control is good, it won't matter whether I use it or not; the result should be the same. The fact that the printer saves money on ink if I do is a matter of no concern to me as the price I pay is the same. But I have no reason whatever to think that printed quality will be better--the whole theory behind GCR is that it creates the same colors in a different way.

II. IF THINGS GO WRONG.

The black ink runs significantly heavier than expected, either because I didn't estimate dot gain correctly or because the printer had bad process control or both. The chances of this happening are quite high. Many separation methods, including Photoshop's built-in engine AND its supplied profiles, underestimate black dot gain. Plus, there are a number of technical factors that can cause black ink to run heavier than normal, and a lot of times that pressmen run it heavier for aesthetic reasons pertaining to the layout. And, this is an unknown printer. So, this happens a *lot* of the time. When it does, there isn't any ifs or buts or whereases about it. The job is a loss, if I've used a heavier than normal GCR. If I've used a skeleton black, there probably won't be any damage--it may even look better than the proof.

With heavier GCR in a subtle-color image, there's no possibility of gain and a very clear and very high possibility of ruin. What lunatic would run such a risk for no purpose? Only one, I suggest, who's more interested in blaming the printer than in taking the easy steps to make sure that there's no blame to be spread because the client is satisfied with the job.

I've seen heavy-GCR workflows function well. But when they do, it's always been in the context of a tightly controlled operation, usually doing a single category of images that come from the same source. In that case, the pressman can actually become accustomed to how a certain type of heavy black behaves and some of the benefits that Dan Remaley discusses become available.

A commercial printer who accepts files from anybody off the street isn't in this position. The pressman doesn't have an easy way to know whether a given image has lots of GCR or a little or even whether there are several different styles on the same page. They have no choice but to use a one-size-fits-all approach. Even if you tell them that every image on the page uses Heavy GCR, it won't help quality unless they have a lot of experience working with such images.

Dan Margulis

____________________________________________________________________________

Dan Remaley wrote,

You assume the pressman will pay any attention to the black. I think that historically pressmen have tended to focus on the three colors and give less notice to the black. In this case lowering the three colors would make matters even worse.

Steven Barton

____________________________________________________________________________

Dan wrote:

One benefit of GCR, especially a heavier application of GCR, will be an increase in sharpness and fidelity due to the general practice of laying the black ink down first. The three colors often print with less fidelity because of chemical ink trapping problems.

Steven Barton

Imaging Sciences & Partners

___________________________________________________________________________

On 12/31/03, Steven Barton wrote:

This is another one of those things that seems to be of more theoretical than practical interest. It's the same thing with with color management where so much emphasis is put on white points and targets and other things of which I am happily ignorant. At least the reasoning behind all this is understandable now that I know that color managment is important for specific applications such as volume production where there isn't time for giving much attention to detailed corrections in individual images.

But when it comes to inkjet printing does it really matter what the printer driver does to the data if the output comes out looking like the image on the monitor (allowing for obvious differences in viewing transmitted vs. reflected light)? Admittedly it would be interesting to know what is actually going on inside the computer when you click on the Print button, but it's still the result that pays the bills. Aside from humorous comments from the computer illiterate ("Oh! It's done on a computer...my nephew can do that. He just puts in the little card and the printers spits it out."), no one really knows or cares what went into producing the final image.

Howard Smith

____________________________________________________________________________

Date: Fri, 2 Jan 2004 21:48:10 -0500

From: Terry Wyse

Subject: Re: Re: GCR

If you're printing through the OS inkjet driver, it's RGB anyway so discussing GCR in this context is meaningless. If you're printing through a RIP to your inkjet, then it's an important thing to consider. I've found that finding the right amount of GCR and K start % is a BIG deal if you want to get the best results possible with a given RIP and inkjet ink/paper combination. Metamerism (as Jim Rich mentioned), banding and overall "smoothness" all tend to be influenced by the separation settings you use when building your inkjet profile. The reasons you would use more or less UCR/GCR with an inkjet are not all related to the discussion about UCR/GCR on press.

Terry

____________________________________________________________________________

Thank you for the information. I had no idea it could be so complicated, but the ame time it makes me glad I didn't invest in a RIP. It's enough fun doing things the hard way. What continues to amaze me is how well my prints turn out without me knowing why. Of course no one is allowed to look at them up close. I may not be smart, but I'm not stupid, either...

Dang! And just when I thought I understood this stuff. On the other hand it does give me a much greater appreciation for the printing crew that used to do all my work..

Howard Smith

___________________________________________________________________________

It's the same thing with with color management where so much emphasis is put on white points and targets and other things of which I am happily ignorant.

Get you to a commercial studio mon ami' and you would not remain both happy and ignorant for too long.

If you actually worked in one and had to move from work station to workstation using all those "not profiled" monitors, life would take on a less than rosy glow.

Richard Corbett

___________________________________________________________________________

Dan Remaley writes,

He may not be able to do that.

First, particularly in advertising, there may be other elements on the page that can't afford to have the black come down. If the choice is to have small type that's broken and illegible or too much black in the picture, you can guess what will happen. If there's a large area of solid black in the layout, a muddy picture is better than a ratty-looking solid.

Second, the sep itself may prevent it. If you're going to be using a heavier GCR, the file preparer has to have a nearly exact knowledge of what your press's black dot gain is. Preparers typically do *not* have this knowledge, and Photoshop's settings don't help them get it. If I'm using a skeleton black and I underestimate black dot gain by 4% it's no big deal. If it's heavy GCR and I make the same mistake I can kiss the image goodbye. If the sep comes in with a black that's too heavy for the press conditions, the pressman *could* compensate by reducing black ink coverage, but that's highly likely to wreck any type or other black element in the immediate vicinity.

The color better not be changing constantly throughout the press run if there's management around that gives a hoot about quality. If pressmen start making artistic decisions about how much ink to use that are necessitated because they don't bother to check midtone and solid values of their inks, or the allow them to become contaminated because they've made a pigsty of their press, they belong at a different company.

Yes, the inking performance will vary slightly throughout the run. But if it's varying so much that there are gross color changes to the point that the pressman has to start making judgment calls, this is an out-of-control situation that no amount of GCR, profiling, or sermons about gray balance is going to correct.

Dan Margulis

___________________________________________________________________________

Steven Barton writes,

I don't understand why the positioning of the black on press would have any effect on this phenomenon. In any event, if black is being printed first, that's a strong argument *against* heavier GCR.

The thing that irritates me about the posts on this topic from Terry and Dan, although they make some good points, is that the overlook the obvious one. Terry, when asked about what he'd like to control in the pressroom, replied "Gray balance, dot gain and density. In that order."

To that, I say, "Consistency, Reliability, Predictability." In any order. If you don't achieve this, all talk about gray balance, GCR, profiling, and dot gain is irrelevant. The only way to get quality results is to hunt down and ruthlessly exterminate all sources of variability. It is far better to be able to hit 28% dot gain every day than it is to be anywhere between 18 and 24% depending on the phase of the moon and other factors.

Very, very high up on the list of things that cause variability are inks that aren't the correct color any more because they were allowed to become contaminated by other inks. If we're slathering four hits of different inks onto a piece of paper at very high speed, it's not that easy to keep traces of Ink #1 out of Ink #2 and so on.

There are two schools of thought on how to minimize this problem. The modern one is to try to keep the sheet as dry as possible for as long as possible by printing the inks in the order of how heavy the coverage is in typical images, lightest coverage first. That order is clearly KCMY.

The more traditional one is to print the lightest *inks* (not coverages) first on the theory that if contamination occurs it won't be as noticeable. It's hard to mess up black ink, so this theory puts it last. The rotation in this method is therefore YCMK.

If a much heavier black is being used, it defeats a lot of the purpose of printing KCMY or any other rotation that puts black anywhere but last, because it becomes more likely that black ink would find its way into the other ink fountains, which is a Very Bad Thing indeed. If I were committed to a heavy-GCR workflow I would think real hard about printing in the order CMYK.

Dan Margulis

P.S. I did a quick search of magazine rate cards, which almost always specify their ink rotation in case we feel like doing press proofs. It looked like over half were KCMY, which is what SWOP recommends. There were probably half as many YCMKs, and there were a couple of CMYKs and even a MYCK.

___________________________________________________________________________

"If" the pressman measure the 3/colors - then they would have a gray patch - closed loop inking (on most presses) have no gray patch. They may visually access the color, but measure - no way. In fact only 'System Brunner' and GMI measure gray and control the ink keys based on gray.

Dan Remaley

Process Control Mgr.

Graphic Arts Technical Foundation

412.741.6860x450

____________________________________________________________________________

Dan Margulis wrote:

Where printing takes place on multi-unit presses the first ink lays down with the best fidelity because it is transferring to dry paper. Subsequent inks are attempting to adhere to a surface coated with a layer of wet ink and tend not to transfer as well. This is the principle of (chemical) ink trapping. It follows that when the black ink is printed first, it tends to have better sharpness and fidelity than the subsequent inks.

A familiar example of poor ink trapping is where black ink attempts to print over a metallic ink, and often prints a about a quarter strength or less.

I believe that the reasoning behind the laydown order of KCMY has more to do with ink trapping, ink transparency and gloss (in heatset printing), than ink contamination.

In fact, I would argue that heavy GCR (and UCR as well) tends to inhibit contamination from the Black. Where the Black is printing heavy, the color plates will have more non-image area and therefore permit more water on their plates. The presence of water on the plate inhibits transfer of ink from the sheet to subsequent inks.

If we consider what I have attempted to explain about ink trapping, I would suggest that heavy GCR would be least successful with Black going down last.

Steven Barton

Imaging Sciences & Partners