Dan Margulis Applied Color Theory - Total Ink Coverage

From: Andrew Adams, admsfmlyof4@earthlink.net
Date: Sat, Apr 1, 2000, 11:19 AM
RE: Total Ink Coverage

My question deals with how much ink on paper is too much (total ink coverage).

Were I work we have fingerprinted on various stocks at various line screens. The results showed that the best we could print (with a balanced shadow before dot failure begins) were as follows:
133 Line Screen ? 340's
150 Line Screen ? 330's
175 Line Screen ? 320's

With this information in hand I formulated what I felt would be reasonable Photoshop CMYK setup preferences for various stocks and line screens.

Yet the person who has the last say about such things (our scanner operator), adamantly feels that my setups (which keep us within the above stated total ink limits) are a recipe for flat color. He feels (with the full support of the owner) that color on our presses "pop" when he consistently "blows out" the highlights to 0,0,0 and saturates the shadows to total ink coverage of 380 to 400.

I realize that depending on the photo, and if there is no meaningful detail to keep in the shadow area, one could feel free to go beyond the total ink coverage limits (as long as the area was not too large). But in my opinion we print "mud" not "pop" color.

Of course, if the owner wants "mud" then who am I to rock the boat. So far my only reply to our scanner operator was, "If that's how we want to print then could you come up with some CMYK setups so that the rest of us can create separations your way".

Yet privately I feel we are doing a poor job. Am I missing something?

Andy Adams


From: Dan Margulis, 76270.1033@compuserve.com
Date: Sat, Apr 1, 2000, 8:00 PM
RE: Total Ink Coverage

Andrew Adams writes:

>>Yet the person who has the last say about such things (our scanner operator), adamantly feels that my setups (which keep us within the above stated total ink limits) are a recipe for flat color. He feels (with the full support of the owner) that color on our presses "pop" when he consistently "blows out" the highlights to 0,0,0 and saturates the shadows to total ink coverage of 380 to 400.>>

Sure. As long as there isn't any critical highlight or shadow detail, and the pressmen don't object, that'll get more contrast. Get hold of a chrome of something like a black dog and a white cat and see how well this method works. It's also rather hard to keep the picture from getting overall too dark this way, but it can be done.

>>I realize that depending on the photo, and if there is no meaningful detail to keep in the shadow area, one could feel free to go beyond the total ink coverage limits (as long as the area was not too large). But in my opinion we print "mud" not "pop" color.>>

Would it not be possible to run a few tests? You must run lots of jobs with space enough in the margin to run a duplicate image. If you can do better, it should be readily visible.

>>Of course, if the owner wants "mud" then who am I to rock the boat. So far my only reply to our scanner operator was, "If that's how we want to print then could you come up with some CMYK setups so that the rest of us can create separations your way".>>

The chances of him being able to answer that are about as good as his chances of being able to name the 23rd Vice President of the United States. With this style of separation you need, for starters, a very short black plate, which you'll probably have to define in CMYK Setup, probably Photoshop's UCR, which is its lightest setting, would still be too dark. Also, you'll need to set dot gain to a lower number than you would normally.

Dan Margulis


From: dantel@rb-group.com, dantel@rb-group.com
Date: Mon, Apr 3, 2000, 8:44 AM
RE: Re: Total Ink Coverage

Scanning and printing CMYK usually involves some level of compromise. Your company and/or your customers have to decide what you are trying to achieve. Are you looking to match the original or trying to print pleasing color? I am not using "pleasing color" as a negative here, many images can be enhanced to make them pop if that is the term you like. Although the approach your company is using may seem to fly in the face of the scientific method you subscribe to, what does it matter if your clients are consistently happy? Our company prints for major advertising agencies in the Chicago marketplace and we routinely push the limits of total ink density unless the images are destined for publication where they sometimes strictly adhere to SWOP standards and may reject materials which do not comply to the standard.


From: Chris Murphy, lists@colorremedies.com
Date: Mon, Apr 3, 2000, 3:24 PM
RE: Re: Total Ink Coverage

>He feels
>(with the full support of the owner) that color on our presses "pop"
>when he consistently "blows out" the highlights to 0,0,0 and saturates
>the shadows to total ink coverage of 380 to 400.

Unless the RIP or imagesetter is using some kind of transfer curve to limit total ink, I don't see any benefit to more than 340% ink let alone 400. 400% can be a recipe for disaster for lithography (laser printers and many dye subs will need 400% TAC because they have built in total ink limiters).

I'm not aware of any printing specification that calls for 380%+ TAC. To figure TAC, generally you print a TIL/KGEN target (Total Ink Limit, Black Generation) such as one GATF has. Basically you rule out areas where set back (? where the ink hasn't dried and it transfers itself to the back of the next press sheet) has occured, and where the paper has deformed even after drying. The remaining patches you measure with a densitometer and plot out, finding where more CMY does not add appreciable amounts of density and that becomes your ink limit. Only with toner type devices, and those that have built in ink limiting functions have I used higher than 350% TAC.

>I realize that depending on the photo, and if there is no meaningful
>detail to keep in the shadow area, one could feel free to go beyond the
>total ink coverage limits (as long as the area was not too large).
>But in my opinion we print "mud" not "pop" color.

The most saturated colors under "normal" (not necessarily average or common) conditions are, 100% cyan, magenta, yellow; 100% cyan+yellow, 100%, cyan + magenta, 100% magenta + yellow. This would be a 200% ink limit. You aren't going to get more saturated colors by raising the ink limit. You will get darker shadows (up to a point of diminishing returns) and more contrast with a higher ink limit with the possible expense of "fill in" where shadow detail starts disappearing.

>Yet privately I feel we are doing a poor job. Am I missing something?

Not that I can tell. Making separations for 400% TAC is asking for big big trouble. I'd love to see those scans go to newsprint. Boy they'd be P.O'd.

Chris Murphy


From: leeb@ids.net (Lee Blevins)
Date: Tue, Apr 4, 2000, 9:23 PM
RE: Re: Total Ink Coverage

> Unless the RIP or imagesetter is using some kind of transfer curve to
> limit total ink, I don't see any benefit to more than 340% ink let alone
> 400. 400% can be a recipe for disaster for lithography (laser printers
> and many dye subs will need 400% TAC because they have built in total ink
> limiters).

The way you evaluate this is to print some test patches.

A simple way is to print squares of density in a series like a gray scale.

You evaluate the results with visual was well as densitometric measurements.

Line the squares up with the lowest dmax on the left and the highest dmax on the right. If you can see the division, take the square to the right as you density.

When you can't see the difference between the square to the right and the square you're looking at, that's your dmax.

I've ran this test on a Komori 6C on a #1 coated sheet and after evaluating it I determined that one could see a difference between 380 and 400.

This approach to manipulating tonal range and a system of creating tests for calibrating is covered very well in Minor White's "The Zone System Manual."

Although it is a book primarily intended for photographers it is a wonderful method for maximizing and learning to predict the behavior of a system of printing. The book is about printing photographically but the tonal characteristics and methods of measuring are very relavent to any form of graphic reproduction.

After reading it one could get the idea of why getting the blackest black you can get [or at least knowing where it is] in an image is so important.

But in response to your statement...

Yes, in a "good" printing situation one should be able to tell the difference between 340 and a 400 dmax. If you can't, you have issues with ink trap. Pressman doing dumb things like running the black down last. I've seen jobs that were bid low that prompted the shop to take all of the old ink off the shelf and mix it to use as black.

In "bad" printing like color lasers and ink jet you probably can't tell the difference.

Life is like a print job. You never know what you're going to get.


From: leeb@ids.net (Lee Blevins)
Date: Wed, Apr 5, 2000, 7:01 AM
RE: Re: Total Ink Coverage

> The goal is not to select 400% ink just because there is a visible
> difference between it and 390% or 380% or whatever is on your target.
> The point is to plot all the amounts and graph it and find out when there
> isn't appreciable increase in density. There usuaully still will be (some
> processes you get a dip in density) an increase in density, but not the
> amount of increase going from 250% to 280% for example. At some point the
> increase starts to plateau and that is where the ink limit should be.

It's hard for me to imagine what the goal would be if it's not to get a richer shadow density. "More contrast" is probably the most common markup on any sep submitted for proof. Since you're dealing with transparencies that have much longer tonal ranges than cmyk proofs it seems to me that getting the longest range out of your material would be to your benifit.

A "textbook" sep would have a shadow of 95C, 88M, 88Y, and 75K. That would create 346% total ink density. While that may have been acceptable in the 70's, today that's too low for good quality images.

You are correct on one point, that there is no reason to use a higher dmax if there is no increase in density. But that *IS* the point. If there is an increase in density and you don't capitalize on that, your images are not as good as they *CAN* be.

This is one of the weaknesses of CMS systems. Some CMS systems don't let you make values higher than 360. Control is turned over to pre-determined calculations. That limits the user from making separations that are specific to their printing process and getting the most out of their range.


From: Chris Murphy
Date: Wed, Apr 5, 2000, 12:53 AM
RE: Re: Total Ink Coverage

>Yes, in a "good" printing situation one should be able to tell the
>difference between 340 and a 400 dmax.

The goal is not to select 400% ink just because there is a visible difference between it and 390% or 380% or whatever is on your target. The point is to plot all the amounts and graph it and find out when there isn't appreciable increase in density. There usuaully still will be (some processes you get a dip in density) an increase in density, but not the amount of increase going from 250% to 280% for example. At some point the increase starts to plateau and that is where the ink limit should be.

>In "bad" printing like color lasers and ink jet you probably can't tell
>the difference.

I disagree. Just because adding a lot more CMY component does NOT increase density does not mean it's a bad process; one MIGHT conclude black ink alone is already very dense. One might also conclude the process simply can't print that much ink (like color lasers, they would break if they printed more than 250% ink, and most I've worked with have built-in limits to as low as 200% ink).

Chris Murphy


From: Steve Upton
Date: Wed, Apr 5, 2000, 12:41 PM
RE: Re: Total Ink Coverage

At 7:01 AM -0400 4/5/00, Lee Blevins wrote:

>This is one of the weaknesses of CMS systems. Some CMS systems don't let
>you make values higher than 360. Control is turned over to
>pre-determined calculations. That limits the user from making
>separations that are specific to their printing process and getting the
>most out of their range.
>

Which CMS are you referring to?

I haven't seen this limitation with ColorSync or ICC profiles yet.

Regards,

Steve Upton


From: Dan Tesch
Date: Wed, Apr 5, 2000, 3:02 PM
RE: Re: Total Ink Coverage

I've been following this thread for a couple of days now and I've gotta tell you - we do it all the time. Certainly there are situations where you don't want to have a very full shadow but by and large we print images all the time that have large areas of 400% density. Obviously you don't hold detail there but it doesn't hurt the rest of the image to push the contrast that high and we have no problems with it on press - "setoff" or plugging or sheet distortion and other things that have been described. We are running a MAN Roland 700 7 color and 6 color but I have seen it done on Heidelbergs also and suspect that any modern press with good operators/paper can pull it off. Believe me when we first started doing this ten years ago, the scanner operators that were doing it were freaking me out so I began asking in the pressroom about problems they have running certain stuff and I just got shoulder shrugging "no problems". The one argument that I heard which seems to possibly have some merit is the point of diminishing return where after a certain point it don't get any blacker - but as someone else pointed out a frequent comment from buyers/creatives is more contrast or more guts/punch. By pushing up the black point O/A and not worrying about max. den. you get more contrast and buyers like it - technically correct or not.


From: Dan Margulis
Date: Wed, Apr 5, 2000, 2:30 PM
RE: Re: Total Ink Coverage

Lee writes:

>>A "textbook" sep would have a shadow of 95C, 88M, 88Y, and 75K. That would create 346% total ink density. While that may have been acceptable in the 70's, today that's too low for good quality images.>>

You mean too *high*, right? Certainly no web printers and no digital printers, and very few sheetfed shops, will accept ink coverage that high.

The numbers cited are, though inferior, typical of what scanner vendors were recommending in the 1980s. The reasons they couldn't come up with anything better were:

1) The numbers are about right for seps made with a process camera and the industry had not yet figured out that hard-edged laser dots print significantly differently than soft-edged camera dots;

2) Clients were naive then about the quality that could be expected in shadows;

3) In an ideal world one would certainly want higher black and lower CMY in the shadows but setting up a means of doing that was beyond the computing power of 1980s scanners, which typically had either 32K or 64K of RAM. One's basic choices were an extremely light black, lighter than would be generated with Photoshop's UCR setting, or an extremely heavy one, almost to the extent of Photoshop's Maximum setting. Given these two grim choices, the first one is usually the lesser of the evils.

>>You are correct on one point, that there is no reason to use a higher dmax if there is no increase in density. But that *IS* the point. If there is an increase in density and you don't capitalize on that, your images are not as good as they *CAN* be.>>

Right, *if* there is no critical detail to hold in the shadows and *if* the printer will accept the job that way.

>>This is one of the weaknesses of CMS systems. Some CMS systems don't let you make values higher than 360. Control is turned over to pre-determined calculations. That limits the user from making separations that are specific to their printing process and getting the most out of their range.>>

The CMS itself will support almost anything you plug into it, which makes it in principle the most flexible method. The problem you are referring to is our current inability to edit ICC profiles in a timely and cost-effective way.

Anyhow, we don't *want* ink that high on separation, because for the typical image it does more harm than good. Cases do exist where we might want it and nothing stops us from adding weight *after* the separation, which is not a CMS issue.

But I do agree, that with current implementations of color management, it's more than slightly ironic that they have less ability to maneuver than was the case in a 1980s scanner with its 64K of RAM.

Dan Margulis


From: Chris Murphy

>A "textbook" sep would have a shadow of 95C, 88M, 88Y, and 75K. That
>would create 346% total ink density. While that may have been acceptable
>in the 70's, today that's too low for good quality images.

Keep in mind today's black ink absorbs more light, is better gray balanced, and retains dot shape better than in the 60's or 70's. It used to be black ink was the lesser expensive ink, and that's not always the case today. Only economies of scale keep it in the ballpark of the process inks.

>You are correct on one point, that there is no reason to use a higher
>dmax if there is no increase in density. But that *IS* the point. If
>there is an increase in density and you don't capitalize on that, your
>images are not as good as they *CAN* be.

This is an ink on paper process. If you add more ink to a black box target, yes you get more density, yes you get more dynamic range. If you do it in an image, the more ink the greater the fill in, the greater the loss of detail. There is a cost associated with getting the additional density. Also I would argue if you're getting significant increases in density as you add CMY to K through 400% ink, that you look into getting a higher quality black ink.

>This is one of the weaknesses of CMS systems. Some CMS systems don't let
>you make values higher than 360. Control is turned over to
>pre-determined calculations. That limits the user from making
>separations that are specific to their printing process and getting the
>most out of their range.

Of the top packages, I haven't seen any TAC limitation. I have seen limited black generation support where it's either automatic (same with UCA) or described only in terms of none, light, medium, heavy and maximum (no UCR or GCR, no black ink limit, no black ink onset, etc.)

Chris Murphy


From: Jonathan Clymer

on 4/4/00 21:20, Lee Blevins at leeb@ids.net wrote:

> This approach to manipulating tonal range and a system of creating tests
> for calibrating is covered very well in Minor White's "The Zone System
> Manual."
...snip...
> > After reading it one could get the idea of why getting the blackest
> black you can get [or at least knowing where it is] in an image is so
> important.

I have only a passing familiarity with White's book, but I am very well acquainted with other books covering the Zone System including Adam's original work, and I have used these methods in my own photography and darkroom printing.

The topic of the thread was total ink coverage on press, and that has been answered by others more knowledgeable than myself. But in several of the replies it was noted that while more ink indeed yeilds more density, it does so with a loss of detail.

This is exactly what happens when you print a black-and-white silver print. You can for sure get deeper blacks by pouring more light onto the paper. The problem is that eventually you are putting a lot of light on the paper to get a small density change. That's another way of saying that you are losing contrast in the shadow areas. When the contrast goes down, the detail disappears. Anyone who does serious B/w printing learns this pretty quickly. Some images will benefit from the darker blacks even though the shadow detail disappears, but as a general rule, most photographs look better with a compromise between holding shadow detail and getting an acceptable black, which means a dMax with is less than the maximum possible.

Jonathan Clymer


From: Jsweengatf@aol.com
Date: Thu, Apr 6, 2000, 10:21 AM
RE: Re: Total Ink Coverage

In a message dated 4/5/00 9:16:32 PM Eastern Daylight Time,
jeclymer@bellatlantic.net writes:

> But in several of the
> replies it was noted that while more ink indeed yeilds more density, it does
> so with a loss of detail.

Yup.

But let's keep in sight why we have spec's like SWOP and SNAP in the first place.

The idea is that anyone can prepare film / files for reproduction: for SWOP for #5 Groundwood, web offset reprtoduction (magazines) and SNAP for Non-heaset Advertising Printing (Newspapers, inserts etc.)

It is NOT the optimum, but rather a common denominator to avoid inline conflicts, and so ALL the pages can "match" the supplied proof on press.

$ .02

John Sweeney
Graphics Microsystems, Inc
(and SNAP, GRACoL, SWOP Advocate)


From: Chris Murphy
Date: Thu, Apr 6, 2000, 1:51 PM
RE: Total Ink Coverage

> Certainly there are situations where you don't
>want to have a very full shadow but by and large we print images all the time
>that have large areas of 400% density.

It's risky doing this. Maybe you aren't having a problem because you've been lucky. This is the first time, on this list, that I've heard of people busting total ink limits by more than a little. A little would be by 5%. Pushing it would be 10%. Running 400% ink inherently makes the process not conform to any lithographic printing specification, and there also isn't a single proof based on 400% ink either (just because the Matchprint will let you doesn't mean it's designed for it).

I've asked a highly respected trainer in process control about this and he won't even work with anyone printing this much ink on press. If you are gaining so much more contrast in black that you need to print 400% ink, you need to talk to your ink man about getting a higher quality black ink, or just accept the contrast you get. The increase in contrast from 340% to 400% should not be all that significant. On a web press, 400% ink would be nuts.

Maybe you've been making your regular sacrifices to the press gods and that's why you haven't had a problem running 400% ink. The specifications are based on thousands and thousands of press runs, not just by the request of someone's pet squirrel.

> Obviously you don't hold detail there
>but it doesn't hurt the rest of the image to push the contrast that high and we
>have no problems with it on press - "setoff" or plugging or sheet distortion
>and other things that have been described.

What is your ink film density? What percentage of the printed area is higher than 350%?

> Believe me when we
>first started doing this ten years ago, the scanner operators that were doing
>it were freaking me out so I began asking in the pressroom about problems they
>have running certain stuff and I just got shoulder shrugging "no problems".

You've been printing 400% ink for 10 years without any problems?

>By pushing up the black
>point O/A and not worrying about max. den. you get more contrast and buyers
>like it - technically correct or not.

So why not print 100% black ink before you go up to 400%? Use more black,
get a better black ink, you won't need to go higher than 340%.

Chris Murphy


From: leeb@ids.net (Lee Blevins)
Date: Thu, Apr 6, 2000, 7:49 PM
RE: Re: Total Ink Coverage

> Which CMS are you referring to?
>
> I haven't seen this limitation with ColorSync or ICC profiles yet.

The last version of Kodak Colorflow would not let us put in a DMAX over 360.


From: Chris Murphy
Date: Thu, Apr 6, 2000, 8:01 PM
RE: Re: Total Ink Coverage

>The last version of Kodak Colorflow would not let us put in a DMAX over 360.

That's one package. This isn't a CMS, ColorSync, or ICC imposed limit. I
would say 360% is not an unreasonable upper limit on TAC and Kodak is likely realizing this; yet at the same time limiting the ability of their software to profile other devices that need to specify 400% ink.

Chris Murphy


From: Chris Murphy
Date: Thu, Apr 6, 2000, 8:00 PM
RE: Re: Total Ink Coverage

>We have no "absolutes" on this. That's why I don't think statements like
>"you'll never need more than 340% DMAX" aren't really accurate.

I have YET to find a European printer than can relate to printing substantially more ink than their applicable specification and they aren't having problems getting more than adequate density and contrast. I'm still looking of course.

Testing to find out the ideal TAC is the best way to go about it. If the test indicates it ought to be substantially higher than an applicable printing specification, I would be wondering why I need so much ink to get the needed density. I wouldn't just accept that I needed 400% ink. I'd want to know why and then I'd work on finding a way to get the density I needed at a lower TAC.

Chris Murphy


From: leeb@ids.net (Lee Blevins)
Date: Thu, Apr 6, 2000, 7:50 PM
RE: Re: Total Ink Coverage

> >>A "textbook" sep would have a shadow of 95C, 88M, 88Y, and 75K. That
> would create 346% total ink density. While that may have been acceptable
> in the 70's, today that's too low for good quality images.>>
>
> You mean too *high*, right? Certainly no web printers and no digital
> printers, and very few sheetfed shops, will accept ink coverage that high.

No, I do mean too low. UCR, like Solid Ink Density is not something you get out of book and arbitrarily apply to your printing condition unless you're making publication ads and must adhere to SWOP's 280 or 300 percent limit.

An inhouse separator should test to see what dmax limits really are if they are concerned with making high quality separations. I am amazed at how many printers show me their samples with grayed out shadows while boasting of their high quality work.

The printers/separators that take the time to make these tests are the top percentages of quality printers that when you look at their work you know you are seeing extremely high quality printing.

Unfortunately this represents a small percentage of printers at the top of the heap.

> The numbers cited are, though inferior, typical of what scanner vendors
> were recommending in the 1980s. The reasons they couldn't come up with
> anything better were:

I got the feeling that DS (in the 80's) was suggesting those numbers for much the same reasons I see them tossed around here. No good reason. They were just numbers somebody got from somebody else and were passed on as aim points for beginners.

The point of my rant is that one can't really say that any particular DMAX fits all situations. If you're interested in doing quality work you should test and get your own numbers.

I received a call from a manufacturer with an interesting question. He had his book printed over the years by many vendors. When he saw the printing after we did the seps he called and wanted to know what we did that made them look so much brighter. The answer is simple. You can't make the paper white any whiter than "no dots" but you can make it look whiter if you make the black "more blacker."

We don't arbitrarily load shadows on seps, we need to know how they're going to be printed. I ask the customer what paper they've chosen and who's going to print the job. If they tell me they're printing on a number 1 sheet on a new 8 color press we do one thing. It is' going on a number 3 sheet on an old web we do another.

We have no "absolutes" on this. That's why I don't think statements like "you'll never need more than 340% DMAX" aren't really accurate.


From: leeb@ids.net (Lee Blevins)
Date: Thu, Apr 6, 2000, 7:50 PM
RE: Re: Total Ink Coverage

> This is exactly what happens when you print a black-and-white silver print.
> You can for sure get deeper blacks by pouring more light onto the paper. The
> problem is that eventually you are putting a lot of light on the paper to
> get a small density change. That's another way of saying that you are losing
> contrast in the shadow areas. When the contrast goes down, the detail
> disappears. Anyone who does serious B/w printing learns this pretty quickly.
> Some images will benefit from the darker blacks even though the shadow
> detail disappears, but as a general rule, most photographs look better with
> a compromise between holding shadow detail and getting an acceptable black,
> which means a dMax with is less than the maximum possible.

The whole point of the Zone System is curve shape adjustment so you can get max black without sacrificing midtone or shadow. (Acutally, the whole point is about "previsualization" so you know where you're placing tone.)

I remember my freshman year of photography when the instructor told us to throw away the data sheet for the B/W print paper that suggested a developing time of 1.5 minutes and required us to leave the paper in the soup for a full 3 minutes.

It only took one demnonstration to see the difference in black density without sacrificing any shadow detail.


From: leeb@ids.net (Lee Blevins)
Date: Thu, Apr 6, 2000, 7:49 PM
RE: Re: Total Ink Coverage

> buyers/creatives is more contrast or more guts/punch. By pushing up the black
> point O/A and not worrying about max. den. you get more contrast and buyers
> like it - technically correct or not.

Well said.

I'm glad to see another printer who isn't just pumping out seps by the numbers but is actually looking at the work and trying to make good art.From: leeb@ids.net (Lee Blevins)


From: leeb@ids.net (Lee Blevins)
Date: Fri, Apr 14, 2000, 6:22 AM
RE: Re: Total Ink Coverage

> Imaging CTP allows you to print higher density with lower dot gain. It
> doesn't have anything to do with ink tack and total ink limits.

I would think it odd to attribute that to CTP alone. It seems that the resulting curve shape on the plate would be the reason for printing higher densities with lower dot gain.

In fact, since one must linearize a plate exposing device just as one must do with a film exposing device the possibility exists that a reverse curve could produce the same result on both [film and CTP] systems


From: Chris Murphy
Date: Thu, Apr 13, 2000, 10:15 PM
RE: Re: Total Ink Coverage

>Putting boundaries on DMax and ink densities in today's workflows, with CTP
>and superior press's is for the everyday shop and the uncreative people of
>our industry.

Imaging CTP allows you to print higher density with lower dot gain. It doesn't have anything to do with ink tack and total ink limits.

What I find uncreative as well as irresponsible is printing different from an appropriate standard and telling the customer to piss off when it comes to someone else making separations for this non-standard condition. I know more people who deviate from standards as a way to get customers stuck with them for scans and seps, not just as a way to print better. What they need to do is profile the press and provide a profile to the customer so they can make appropriate separations for this non-standard printing condition. At that point I don't have a problem with deviating from a standard so long as it isn't a distributed print need (like magazine printing).

Chris Murphy


From: Chris Murphy
Date: Fri, Apr 14, 2000, 1:38 PM
Re: Total Ink Coverage

>I'm an advocate of the ends justify the means. If a better image
>results, do it.

Better image compared to what? Compared to the same press at 340% but at 400% they look better? If I brought you my separations at 340% total ink and you complained the total ink should be 400% or I won't get sufficient contrast, but I print with Joe Blow and his press prints just fine at 340% I'm going to ask you to do the job over and get it right next time. Other people don't have problems getting sufficient density and contrast running 340% ink, so as a print provide, I would be wondering why 400% is needed. I wouldn't just accept it because the ends justify the means.

>Process control stats are guidelines until they are tested and tuned.

Printing specifications are just that, specifications. GRACoL is a guideline. SWOP isn't. I would argue those who wrote GRACoL wouldn't have a problem with going 10% over the recommended total ink limit (so if your guideline calls for 300% ink, printing 330% isn't going to get you into trouble); but printing 400% is asking for big trouble.

Chris Murphy


From: Chris Murphy
Date: Fri, Apr 14, 2000, 1:37 PM
RE: Re: Total Ink Coverage

>I would think it odd to attribute that to CTP alone. It seems that the
>resulting curve shape on the plate would be the reason for printing
>higher densities with lower dot gain.

The reason is you get a sharper dot that doesn't gain as much once on press, so you can go with higher ink film density without risking too high dot gain and getting muddy colors.

>In fact, since one must linearize a plate exposing device just as one
>must do with a film exposing device the possibility exists that a
>reverse curve could produce the same result on both [film and CTP]
>systems.

Delinearizing the imagesetter to compensate for dot gain isn't a good way to go about it.

Chris Murphy


From: Chris Murphy
Date: Fri, Apr 14, 2000, 7:00 PM
RE: Re: Total Ink Coverage

>> Delinearizing the imagesetter to compensate for dot gain isn't a good way
>> to go about it.

>Really?
>
>What would you suggest?

Either reducing ink film density, or using a lower lpi; adjusting blanket pressure, ink viscosity and/or tack; there are other things you can do I'm not thinking of, but delinearizing the imagesetter doesn't compensate for dot gain. If dot gain is too high, it's too high (too high as in colors are getting muddy) and you need to go about controlling it, not blaming it on something that isn't causing it. Plus when you dork the imagesetter, not a single contract proof is going to come out right.

Chris Murphy


From: leeb@ids.net (Lee Blevins)
Date: Sat, Apr 15, 2000, 5:35 AM
RE: Re: Total Ink Coverage

> Either reducing ink film density, or using a lower lpi; adjusting blanket
> pressure, ink viscosity and/or tack; there are other things you can do
> I'm not thinking of, but delinearizing the imagesetter doesn't compensate
> for dot gain. If dot gain is too high, it's too high (too high as in
> colors are getting muddy) and you need to go about controlling it, not
> blaming it on something that isn't causing it. Plus when you dork the
> imagesetter, not a single contract proof is going to come out right.

I deserve that answer since I asked the question rhetorically.

Reducing ink density to compensate for dot gain is about the worst thing
you can do. That will flatten the print and produce some real ugly
printing.

I don't know why you think that altering linearization is a bad thing.
It's the best approach to dealing with high gain situations.

We do this on a regular basis for printers. We plot their curve shape
and create compensated film. We also create curves for our Iris to that
allows us to match their print in both shape and density so we can see
the effect.

I guess we didn't read the book that says we're not supposed to do this
but since the customer is very happy with the results and they pay us
I'll skip that chapter.


From: Cmyk54@aol.com
Date: Sat, Apr 15, 2000, 10:50 PM
RE: Re: Total Ink Coverage

HMMMMM

Chris
..profile my press conditions and send them to my clients? We can't even get most clients to send us the right fonts or files?? . I guess we do not print at an appropriate standard, but I know the fashion and automobile clients pay top dollar to produce their work one step above the rest. Our shop produces top quality work every day of the week and I will share my secrets with anyone.... its called hard work , skilled craftsman and up to date equipment..

Alan Kochis


From: Chris Murphy
Date: Sun, Apr 16, 2000, 8:33 PM
RE: Re: Total Ink Coverage

Lee writes:
>Reducing ink density to compensate for dot gain is about the worst thing
>you can do. That will flatten the print and produce some real ugly
>printing.

The top 20 printers in the country start out with an applicable printing specification first, and then optimize the press from there. NONE of them use the imagesetter as a means to control dot gain. NONE. They modify their separation method to compensate for dot gain, not the imagesetter. The imagesetter is to remain linearized and calibrated. You don't screw up the imagesetter in order to fix some other problem. You find that some other problem and fix it. Leave the imagesetter linear.

Factors that affect dot gain are ink density, screen ruling, packing, ink tack, water-ink balance & temperature, and other items. That's where you should be investigating if you have dot gain that is too high (muddy colors, sacrificed print contrast).

>I guess we didn't read the book that says we're not supposed to do this
>but since the customer is very happy with the results and they pay us
>I'll skip that chapter.

Who ever said the customer was always correct and can always identify high quality printing? Compared to what? MicroSoft and others have proven if you're nice to the customers and use lots of good marketing you can sell anything (until they figure it out, then you've got a wee problem.)

Chris Murphy


From: leeb@ids.net (Lee Blevins)
Date: Mon, Apr 17, 2000, 6:52 AM
RE: Re: Total Ink Coverage

> The top 20 printers in the country start out with an applicable printing
> specification first, and then optimize the press from there. NONE of them
> use the imagesetter as a means to control dot gain. NONE. They modify
> their separation method to compensate for dot gain, not the imagesetter.
> The imagesetter is to remain linearized and calibrated. You don't screw
> up the imagesetter in order to fix some other problem. You find that some
> other problem and fix it. Leave the imagesetter linear.

I'm not sure who you're calling the "top 20." Perhaps you're including TV Guide and Watchtower Publications. Could you list the top twenty you are referring to? I've believe I've worked for at least two of the top twenty commercial printers in the country and I'd like to see if I'm on the list? I'm thinking commercial color lithographer with sales in excess of 60 million makes the top twenty in that category. Lumping all printing and publishing and then taking the top 20 of that would produce a group that probably has little in common with how they manage their overall color processes.

The problem with using the separation only for controlling dot gain is it only applies to contone images separated with a means to control dot gain. That would imply that the "top 20" scan in RGB and convert in Photoshop.

If you want to control the whole page you must do it at either film or platesetter. Just controlling the contones is a half measure. Imagine a designer choosing a tint value by reading an area of an image and finding out that only the image was controlled for dot gain.

The truth be told, most commercial printers don't attempt to control dot gain at all beyound keeping systems linear. The only ones who can really control it are the publication printers who print on the same stock all the time. Commercial printers may print on that stock one time and never see it again. Many times they're just able to get the paper to the press in time to run and never have the ability to test it beforehand.

We don't re-linearize the imagesetter, our system lets up apply curves to every piece of film independently. The printers this applies to have carefully evaluated their curve shape on the substrate they print on. We have them print 21 step scales and measure their dot gain along that scale and build a reverse curve to compensate for it. I'm not really clear on why you think that's a bad idea. It seems so practicle.

> Factors that affect dot gain are ink density, screen ruling, packing, ink
> tack, water-ink balance & temperature, and other items. That's where you
> should be investigating if you have dot gain that is too high (muddy
> colors, sacrificed print contrast).

My point was that trying to reduce solid ink density in a high gain situation would produce bad printing. The midtone would increase and the solid ink density would reduce. READ - VERY LOW CONTRAST

The biggest factor is really paper. Most printers have the items you mentioned under tight control. The substrate becomes the biggest variable they face in controlling dot gain.

If you know you're headed to high gain (offset and uncoated papers) the best think you can do is reduce the screen ruling.

* While we're on the subject, I'm amazed at how many people think that UCR is a method of controlling dot gain.

> Who ever said the customer was always correct and can always identify
> high quality printing? Compared to what? MicroSoft and others have proven
> if you're nice to the customers and use lots of good marketing you can
> sell anything (until they figure it out, then you've got a wee problem.)
>

Basic business 1A says "the customer is always right." They're the people who give us the money that funds what we do. Without them and their checks, we can all go home and this whole discussion means nothing.

Resisting a customers request for color change on the basis that you are "in technical spec" is a bad bad idea. I would sell my stock in your company quickly.

My fatherinlaw was a pressman at the Boston Globe for years. He complained about management telling them to run the black density higher than he felt was "technicaly correct." Point was, it made the ad copy look better and guess who pays the bills in a newspaper?

In my humble opinion, the real skill of a color technician is in being able to understand the customers request and execute it and make them happy. This requires a big bag of tricks. Years of satisfying agency buyers has made my bag of tricks very large.

Color printing where there is no revision cycle is like shooting fish in a barrel. That's easy. Whatever YOU decide is good color is what prints.


From: Chris Murphy, lists@colorremedies.com
Date: Mon, Apr 17, 2000, 1:25 PM
Re: [ColorTheory] Total Ink Coverage

>I'm not sure who you're calling the "top 20." Perhaps you're including
>TV Guide and Watchtower Publications. Could you list the top twenty you
>are referring to? I've believe I've worked for at least two of the top
>twenty commercial printers in the country and I'd like to see if I'm on
>the list? I'm thinking commercial color lithographer with sales in
>excess of 60 million makes the top twenty in that category.

>Lumping all
>printing and publishing and then taking the top 20 of that would produce
>a group that probably has little in common with how they manage their
>overall color processes.

Precisely, yet they don't use the imagesetter as a means to control dot gain.

>The problem with using the separation only for controlling dot gain is
>it only applies to contone images separated with a means to control dot
>gain. That would imply that the "top 20" scan in RGB and convert in
>Photoshop.

What? All drum scanning software and scanner operators have a means to compensate for dot gain. You tell them what the dot gain is, and they compensate for it in separations. It's been done that way for years. Imagesetters remain linear and calibrated. Dot gain, ink hue, paper white is compensated for at the separations stage. Plenty of other software makes separations while compensating for dot gain than just Photoshop.

>If you want to control the whole page you must do it at either film or
>platesetter. Just controlling the contones is a half measure. Imagine a
>designer choosing a tint value by reading an area of an image and
>finding out that only the image was controlled for dot gain.

You haven't printed for a magazine before have you. You expect designers to compensate for your press dot gain, if they don't, it's their problem, not yours. You aren't going to dork the imagesetter for each and every ad. The press has a specific dot gain and if the designer is going to specify tints, they need to compensate for dot gain when they make the tint. The ad right below and the images to the left of her ad and tint are going to be made with dot gain compensation in advance the way it ought to be.

What you're suggesting is that you try to change the world and tell them to no longer compensate for dot gain in their images, something people are doing with varying degrees of accuracy (but they are doing it) since the beginning of lithography; all in the name of having a single place to compensate for it because of a few designers who never learned, for whatever reason, they need to build their tints with dot gain in mind.

Well too bad for them. The whole world is not going to change. They will continue to build CMYK images with ink hues and dot gain in mind like they've been doing whether it happens on a drum scanner, a flat bed scanner, or in Photoshop, or in LinoColor, or in ColorShop, or in any number of other apps.

>The truth be told, most commercial printers don't attempt to control dot
>gain at all beyound keeping systems linear.

Well most commercial printers don't check for gray balance either, so what you're trying to say here has less weight than implied.

Keeping sysems linear isn't even a method for controlling dot gain. Controlling dot gain (keep it the same day to day, hour to hour) is to control SID, packing, tack, ink-water balance, temperature, etc. You keep it the same. You fingerprint the press, as has been done for decades, to find out what the dot gain actually is - you optimize press behavior and record the behavior so you can reproduce it day in and day out. Once you know what the dot gain is, you tell the scanner operator and HE compensates for it.

> The only ones who can really
>control it are the publication printers who print on the same stock all
>the time.

Publication printers who are controlling it at the imagesetter are asking for a whole lot of trouble and I'd like some examples if you know a publications printer who's doing exactly that.

>The printers this applies to have
>carefully evaluated their curve shape on the substrate they print on. We
>have them print 21 step scales and measure their dot gain along that
>scale and build a reverse curve to compensate for it. I'm not really
>clear on why you think that's a bad idea. It seems so practicle.

First off, this is NOT delinearizing the imagesetter. The imagesetter is still linear. You are simply applying different transfer curves to the data either before or after it's RIP'd, then it goes on to a LINEAR imagesetter. What you've done is a CMYK to CMYK conversion to compensate for dot gain using a curves based profile. Color management, repurposing, whatever you want to call it. Delinearizing the imagesetter is not what's going on here.

Second, it's a really sloppy workflow for you to have to figure out what pieces of each job has proper compensation for dot gain and which pieces don't. You used the example of an image and a tint. If you have an image and a tint on the same page and the image has dot gain compensated for and the tint doesn't, you method of using a transfer curve doesn't help. You will need to go into the document and adjust the tint first, then make films that are consistent (now if those still need to go through transfer curves, fine, but they need to start out on a level playing field).

>My point was that trying to reduce solid ink density in a high gain
>situation would produce bad printing. The midtone would increase and the
>solid ink density would reduce. READ - VERY LOW CONTRAST

That's only the case if the SID is where it ought to be. If SID is too high and that's what's causing the high dot gain, print contrast will also be reduced. Reducing ink density in this case will reduce dot gain, and increase print contrast.

Print contrast is computed at a 73% tint. (Ds - Dt) / Dt (Density of shadow minus density of tint divided by density of tint). For publications it would be 35-45% for K, 30-40% for C and M, and 25-35 for Y. For a publication you shouldn't be doing better than that (publications should be printed per SWOP because that's what everyone is expecting, and might have ads in other publications and are expecting uniformity). If you are following GRACoL, the numbers vary depending on substrate, but an opimitized press should be pretty easy to make perform better than GRACoL.

>The biggest factor is really paper. Most printers have the items you
>mentioned under tight control. The substrate becomes the biggest
>variable they face in controlling dot gain.

Yes, I mentioned this and a good quality control system will have a means to ensure you expect as consistent a product as is possible from your vendor for both ink and paper. That way you aren't having to recalibrate the press every month, or however often you go through a shipment of supply.

>If you know you're headed to high gain (offset and uncoated papers) the
>best think you can do is reduce the screen ruling.

Yes and I mentioned that as well.

>* While we're on the subject, I'm amazed at how many people think that
> UCR is a method of controlling dot gain.

Well if the dot gain of K is substantially less than the other three channels, I suppose you could use it as a way to control the effect of dot gain in neutral shadows; or GCR for that matter. I'd want to know why K dot gain was so much lower than the other three of course.

>Basic business 1A says "the customer is always right."

That's false. That started back in the day of Jimmy's Diner, not in the day of the printing business. The idea the customer is always right being brought into the printing business has caused at least as much friction as any other change. You're not going to let a customer dictate ink densities, print contrast, ink tack, or what you use for packing are you? I don't think so.

In the color reproduction business, we must rely primarily on people fitting into a system and where the system isn't complete, faking it by using skill and experience. Otherwise, we have a free for all (and we're close to a free for all in this industry as it is).

> They're the
>people who give us the money that funds what we do. Without them and
>their checks, we can all go home and this whole discussion means nothing.

Yes and that is happening right now as we speak. Those with the real high quality printing, and the others who adapt and change with the times will be the ones to survive.

>Resisting a customers request for color change on the basis that you are
>"in technical spec" is a bad bad idea. I would sell my stock in your
>company quickly.

You do whatever is necessary to make the customer happy within limits. If they want 6000 more for free, they can forget it. If they want to change a dress from pink to puce at press check after they signed the proof, be happy to change it and make new plates while the press waits, but you're charging them for it. You're making it sound like the customer is the master and printers are slaves. There are limits. You fit into a system and if you don't, you can pay for the printer to be the system for you.

Where I will agree with "the customer is always right" is when the customer has a better idea of the system than the printer does. I run into that as well, where the printer checks for SID once a day, never checks dot gain, never checks print contrast, never checks gray balance. Hey if they don't care, I can find someone who can.

>My fatherinlaw was a pressman at the Boston Globe for years. He
>complained about management telling them to run the black density higher
>than he felt was "technicaly correct." Point was, it made the ad copy
>look better and guess who pays the bills in a newspaper?

As long as it's not screwing up someone else's job, I don't have a problem with it. The technically correct black density is going to be really important when you start printing color images though.

I would be even more pissed if I were a customer following SNAP or ISO 12647-3 for my separations and someone said "well most of our customers don't do it right so we pump up the black." As long as I get a free ad. If you won't give me a profile or separation table for your newsprint press, expect me to assume you print per a specification such as SNAP or ISO 12647-3.

Chris Murphy


From: digital1@inmind.com, digital1@inmind.com
Date: Tue, Apr 18, 2000, 12:59 PM
Re: [ColorTheory] Total Ink Coverage

BTW, what do the letters TAC stand for?

Les Schofer


From: Chris Murphy, lists@colorremedies.com
Date: Tue, Apr 18, 2000, 10:19 PM
RE: Re: [ColorTheory] Total Ink Coverage

>BTW, what do the letters TAC stand for?

Total Area Coverage. I'm not sure of the derivation, maybe it's an ISO thing (more international). It's essentially the same thing as total ink limit.

TVI is Tone Value Increase, also known as dot gain. I haven't yet heard of TVD which would be Tone Value Decrease, also known as dot sharpening; but I don't do a lot of work with positive plate making. The effect does occur (dot sharpening) I just don't know if TVD is an actual term.

Chris Murphy


From: Dan Margulis, 76270.1033@compuserve.com
Date: Tue, Apr 18, 2000, 3:42 PM
RE: Re: [ColorTheory] Total Ink Coverage

I hate to join this battle, but,

Chris writes:

>>What? All drum scanning software and scanner operators have a means to compensate for dot gain. You tell them what the dot gain is, and they compensate for it in separations. It's been done that way for years.

It has not. At least half the time the scanner operator had no clue about the destination of the sep and therefore assumed that it was "SWOP", whatever that means.

>>You haven't printed for a magazine before have you. You expect designers to compensate for your press dot gain, if they don't, it's their problem, not yours.

Lee may not have done magazines but I am fairly experienced in the field. Process control is a wonderful thing, but making a 30-year old press behave like a much newer model from a different manufacturer is a lot more difficult than one might think. The client often doesn't know which specific press is going to be used, and assumes "SWOP", whatever that means. AFAIK every magazine printer uses some kind of dork to accommodate its different printing conditions in a manner transparent to the client.

In fairness, the dork is more commonly in contacting or platemaking than it is in imagesetting, because historically magazine printers haven't done much imagesetting.

>>What you're suggesting is that you try to change the world and tell them to no longer compensate for dot gain in their images...>>

On the contrary, what Lee suggests is the status quo: that the client prepare the files to a loose standard and allow the printer to do the rest.

>>First off, this is NOT delinearizing the imagesetter. The imagesetter is still linear. You are simply applying different transfer curves to the data either before or after it's RIP'd, then it goes on to a LINEAR imagesetter. What you've done is a CMYK to CMYK conversion to compensate for dot gain using a curves based profile. Color management, repurposing, whatever you want to call it. Delinearizing the imagesetter is not what's going on here.

I agree. It is a form of color management, and probably what Lee is talking about is more accurately called an in-RIP curve than a imagesetter dork. That said, however, directly addressing the output device is a time-honored way to do this. In the early 90s Southwest Software had a package specifically for this, where the printer would send a patch over to the imagesetter for varied output conditions. Iris printers have traditionally been calibrated for different papers by sending over a direct call to the output device, not any kind of curve. Gravure printers dork their Helioklischographs with a direct call, not any type of preprocessing curve. And on and on.

>>Second, it's a really sloppy workflow for you to have to figure out what pieces of each job has proper compensation for dot gain and which pieces don't. .

He doesn't have to. He assumes that they have compensated for "SWOP", whatever that is, and to the extent that his presses vary from that "standard", he dorks somewhere. An advocate of ICC color management would do approximately the same thing, except he would call it a precision color transform, so that people wouldn't know that it was still a dork.

Dan Margulis


From: Chris Murphy, lists@colorremedies.com
Date: Tue, Apr 18, 2000, 10:22 PM
RE: Re: [ColorTheory] Total Ink Coverage

>It has not. At least half the time the scanner operator had no clue about
>the destination of the sep and therefore assumed that it was "SWOP",
>whatever that means.

At least half of the time TODAY. It's more an more common for people not to know where they are printing when they are getting scans done than it used to be.

The implication was that scanner operators don't compensate for dot gain. If they know where it's being printed, they are compensating for specific dot gain. If they don't kow where it's being printed, they are compensating for SWOP dot gain. And yes, SWOP is a very specific thing, it's just that the U.S. application of SWOP isn't a very specific thing.

Chris Murphy


From: Lee Blevins, leeb@ids.net
Date: Tue, Apr 18, 2000, 6:59 PM
RE: Re: [ColorTheory] Total Ink Coverage

> I hate to join this battle, but,

Thanks Dan.

I think I finally get it. It came to me while driving to work. Chris's use of the term "dot gain" is probably what he's perceiving from the separation setup in Photoshop.

Since we don't scan in rgb and convert to cmyk in photoshop we don't even think of that setting.

I am referring to "dot gain" that happens in printing ink on paper.

We consider "no gain" to be the normal 15 or so percent that happens in commercial printing.

Our use of the imagesetter to compensate for "excessive gain" such as printing on offset stock that is higher than the "norm."

Before we had digital imagesetting we used other methods such as altering the plate exposure or using deep etch plates that sharpened a few percent to help out on the web printing. (remember PDI plates?) In some cases we contacted to positive and then contacted it back overexposed. There were all sorts of tricks.

The point was we did it to the whole page. Not just the images.

I think possibly Chris's objection has to do with some confusion about what I'm calling gain and what he's calling gain.

I remember getting into a similar confusing discussion with a color scientist at Kodak. He swore all scans were made with GCR and I was telling him that wasn't true.

I was referring to the actual scanner knob that controls GCR and I think we was referring to selective color correction that would reduce the graying component and replace it with black. In his mind all scanner operators would adjust their color computer to do some substitution of unwanted color for black. I was thinking of actually turning the GCR knob on or off.

This is a trade and not always a science. Terms can get confusing.


From: Chris Murphy, lists@colorremedies.com
Date: Tue, Apr 18, 2000, 10:30 PM
RE: [ColorTheory] clearing up the dot gain/linearization issue

>Lee may not have done magazines but I am fairly experienced in the field.
>Process control is a wonderful thing, but making a 30-year old press behave
>like a much newer model from a different manufacturer is a lot more
>difficult than one might think.

And a lot easier than some printers want us to believe.

> The client often doesn't know which
>specific press is going to be used, and assumes "SWOP", whatever that
>means. AFAIK every magazine printer uses some kind of dork to accommodate
>its different printing conditions in a manner transparent to the client.

OK wait a minute - now I see what's going on. I have seen the light of day. OK, now I'm thinking that Lee was referring to using transfer curves to CONTROL dot gain as a form of process control; to solve dot gain when it's too high. This cannot be done with transfer curves. You have to control the press.

Now in the case where SWOP is a specific thing, and everyone (we would hope) would make SWOP separations the same way, in order to get SWOP/TR001 behavior from a press it will likely be necessary to use transfer curves in order to compensate for that press's nature (not *too* high or too low, just DIFFERENT from the standard characterization - i.e. TR001). OTHERWISE you would have to make new separations for every freaking printing press on the planet. I'm all for that.

Geez, I was thinking that Lee was talking about transfer curves, and delinearized imagesetters as a means to fix a dot gain problem. OK, it seems pretty clear at this point that's not was he meant; but rather to compensate for the normal variations in dot gain that presses have - not problematic dot gain situations.

OKkkkkkk......


Chris Murphy

Adobe Photoshop training classes are taught in the US by Sterling Ledet & Associates, Inc.