Dan Margulis Applied Color Theory

Offset: How Good Are You?

Offset: How Good Are You?
Posted by: "Ron Kelly"  
Wed Feb 7, 2007 2:40 pm (PST)

This is a question for those who prepare files for offset.

How many proofs do you expect to make per file prepared, on the average, to achieve good work?
Those of you who are among the best, would you expect to do the majority of
images in one proof? Two? Three or more?

I've got a calibrated screen and profiled proofer. I use the info pallet to tell me what's neutral, white and black. Still, to *know* if an image is right, I have to print it.

This is especially true for certain images. Why not all? I wish I had an answer for that one. Some images preview pretty closely to the printed result, but some are w-a-y o-u-t. This is especially true for CMYK work as opposed to RGB.

I'm confident, however, that when I make a proof, *that* is what I'll get. That's my gold standard. If I didn't have a proofer to guide me, I don't think I could prepare files for printing.

Are there any who are so confident that they don't proof? It could save you a lot of time and money, but it also could cost you a lot more if you aren't (like me.)

I have been preparing my own files for off-set for about 12 years but I do not consider myself *expert*. My business is seasonal, and I do get better with practice but I have long periods where I don't turn a wheel on the proofing RIP.

My minimum is probably about 2 or 3 prints per file. However, sometimes it goes to 10 and beyond.
I typically proof scaled down, to save paper, ink & time. If necessary I will print out a section at 100% scale to judge critical sharpening. It's not usually necessary, though, as sharpening is not nearly as challenging as color and contrast, which proof just fine on a half page.

In summary, I would expect that working at this 40 hours a week (or more) at this would make you pretty good. How good is that? If you are preparing magazine covers for big bucks, how hard is it to get it right?

Ron Kelly
___________________________________________________________________________

Re: Offset: How Good Are You?
Posted by: "John Romano"
Thu Feb 8, 2007 5:58 am (PST)

Ron

If your Monitor is Calibrated and Profiled correctly and your Proofer is as well.

You shouldn©ˆt need more than one proof on 90% of your images.

If your seeing some as close and some WAY out there is something wrong in Your set up.

Soft Proofing either works or doesn©ˆt....Not sometimes !

If I had to make 3 proofs per image my company would go Broke.....Approvals are expensive :)

John R
___________________________________________________________________________

Re: Offset: How Good Are You?
Posted by: "Ron Kelly"
Thu Feb 8, 2007 8:16 am (PST)

John:

I should clarify. I make successive iterations in part because the proof is different than I expected, and in part because the printed piece, while not entirely different than expected, is "not good enough" and the proof suggests ways to improve the image.

Basically, I find the subtlety of the image is lost on screen. The fine qualities, if print is the final goal of this process, can only be judged by a print. A screen image is coarse and rough.

The best monitors I've ever seen are better than the worst at showing the "true" state of the file, but not nearly close enough in my books.

I realize that others may be correcting for different purposes, such as "we're just trying to get this newspaper out" or somesuch where the quality is not as important as something that goes into your portfolio.

Ron Kelly
___________________________________________________________________________

Re: Offset: How Good Are You?
Posted by: "Andrew S. Webb"
Thu Feb 8, 2007 8:20 am (PST)

I rarely have to make a second proof. Most of the proofs I make are only to show my clients what I already know. "Way out" sounds as though something is wrong with your setup or your softproofing. Ten proofs? Something is very wrong. The time and money spent making ten proofs will buy you a half-day of Chris Murphy's expert RIP calibration.

I almost always proof at 100%, because that's what size ther rest of the world will see the work. I only scale for catalog work, and then I'm scaling to 50% of camera output size, which is often larger than the final print size. Proofing at sizes other than 100% can give you funny ideas about saturation and "weight".

Cheers,

_andrew webb
___________________________________________________________________________

Re: Offset: How Good Are You?
Posted by: "Ron Kelly"
Thu Feb 8, 2007 8:59 am (PST)

Andrew:

Could you elaborate. I don't see how the size matters in regard to saturation. I'm not sure what you refer to with "weight" but maybe you just mean the impact of size.

Ron Kelly
___________________________________________________________________________

Re: Offset: How Good Are You?
Posted by: "Lee Clawson"
Thu Feb 8, 2007 11:20 am (PST)

on 2/7/07 6:31 PM, Ron Kelly wrote:

...[snip]...How many proofs do you expect to make per file prepared, on the
average, to achieve good work?

Usually one. If I'm working with a new printer more. The one's that give me the most problems contain lo-res JPEG's that have been "worked" to appear somewhat ok for press.

...[snip]...Still, to *know* if an image is right, I have to print it.

More than anything I find its that my eyes give a "pass" to memory colors to quickly and those are the ones I need to be especially careful with.

I'm confident, however, that when I make a proof, *that* is what I'll get.
That's my gold standard. If I didn't have a proofer to guide me, I don't think
I could prepare files for printing. ...[snip]...

I thought we where talking about proofs from the print house. Are we instead talking about desktop proofing ??

I take proofs from our printers as as reasonable expectations of what will happen in production. Some of our print vendors are better than others at holding to these targets. On the other hand I've had reprints made and called the printer to say I like these much better than 1st job.

...[snip]...Are there any who are so confident that they don't proof?

With jobs that are set up like ones in the past (same printer, same paper stock, same page layout file type) I'd go to press no proof if time dictated the need to do so.

My minimum is probably about 2 or 3 prints per file. However, sometimes it
goes to 10 and beyond. ...[snip]...

I have run at least that many on a desktop proofer. Usually its design considerations that push the count high. Sometimes its color changes made by a client.

The printer we work with most often is making stuff ready for press by the time I receive a proof. We're all betting on saving time. There's times we have copy/photo edits but there's no way I or a client could afford to go more than 2 proofs. With catalog work press time is reserved way ahead. We plan proofing in the early stages of the job so as to not miss that date by extending proofing requests.

...[snip]... working at this 40 hours a week (or more) at this would make
you pretty good. How good is that? If you are preparing magazine covers for
big bucks, how hard is it to get it right?

I have a retired photographer (78) who comes in once a week looks at what we're doing and makes some comments, helps me with problem photos, has lunch and leaves. He "see's" color much better than I do. Afterwards I feel like you; it's hard work to get good color.

Lee Clawson
2/\V/\7 Studio
___________________________________________________________________________

Re: Offset: How Good Are You?
Posted by: "Lee Clawson"
Thu Feb 8, 2007 11:22 am (PST)

on 2/8/07 12:02 PM, Ron Kelly wrote:

...[snip]...where the quality is not as important as something that goes into
your portfolio.

Ron,

That's different expectations than I thought you'd been asking about.

Lee Clawson
2/\V/\7 Studio
___________________________________________________________________________

Re: Offset: How Good Are You?
Posted by: "Andrew S. Webb"
Thu Feb 8, 2007 11:23 am (PST)

Ron,

A smaller version of a given image can look both denser (tone) and more saturated than a larger version. For example, a seascape with beach, sea and sky can look like a very saturated bluegreen smudge from a distance. You don't see the sand until it is at a sufficient size. A black and white checkerboard will look like a medium gray at a small size. Judging light/dark and saturation is easier and more accurate at the final output size.

By "weight", I meant tonal weight; as in typography "page weight"; the distribution of tones and sense of mass one gets from a printed page.

Wrinkles around people's eyes often disappear when prints are scaled down; after all, if you reduce the size by 50%, you're printing 50% of the information in the file.

Cheers,

_andrew webb
___________________________________________________________________________

Re: Offset: How Good Are You?
Posted by: Dan Margulis
Thu Feb 8, 2007 4:33 pm (PST)

Ron Kelly writes,

Basically, I find the subtlety of the image is lost on screen. The
fine qualities,
if print is the final goal of this process, can only be judged by a
print. A screen
image is coarse and rough.

It is, but with practice you adjust to it. After all, people who have been used to correcting on CRT monitors now have been able to make the jump to LCD, which is a much rougher look.

I would suggest carefully restudying the proofs you have had a problem with against the screen images, so that you can see exactly what it is that's fooling you. If it's a color accuracy issue, that's one thing, if the color and weight is in principle accurate but you aren't interpreting the screen correctly that's another.

I agree with John Romano and would probably go even farther--I think that a second proof is very rarely necessary, unless the client and/or printer requires it. I don't even bother to proof most of what I do because I'm pretty confident of how it will print. For example, the only time I made the magazines get proofs for my articles was when I wrote about moire, which is something you can't always evaluate from the screen. Most of the images in previous editions of Professional Photoshop had no proof.

Now, it's become so cheap to gang things up on a 24x36 that I pretty much proof everything that's of any importance, but it's more for design reasons than for color. If I'm running a series of several versions of the same image it's very difficult to evaluate how they work as a package on the screen. I also find proofing useful when there is some minor defect (like pronounced noise) that I expect to be nearly invisible but don't wish to take a chance.

I don't recall having redone any images in PP5E after the proof except in cases where the overall package didn't work. I redid a few more in Canyon Conundrum because I was less familiar with the particular press.

But even in the cases where you legitimately decide to make changes after the first proof, they should be minor enough that a second proof isn't needed. You just mark the first proof, "this proof is no longer accurate for color; the fleshtone has now been lightened" or whatever. Most printers have no problem accepting such a thing for guidance.

Dan Margulis
___________________________________________________________________________

Control during a press check
Posted by: "mfedecky"
Fri Feb 9, 2007 1:45 pm (PST)

My question relates to how much control can be exerted during a press check by the designer. Assuming I have done everything by the book i.e. I've color corrected by the numbers. I've made my color separations taking into account dot gain, total ink limit etc etc. I have even submitted contract proofs (matchprints or an accurate digital proof) and told the printer to match them. And yet when I'm at a press check, and I think the midtones look heavy or the fleshtones look too yellow, can significant changes be made at this point, or is pretty much etched in stone. Can you pin a numerical value on this? What is the maximum a color can be altered on press? Whenever I ask for a change, I feel the pressman is humouring me and nothing changes.

Much thanks
Michael Fedecky
Fedecky Design
___________________________________________________________________________

Re: Control during a press check
Posted by: "Louis Dina"
Fri Feb 9, 2007 2:45 pm (PST)

Michael,

First, I usually give my printer a digital proof as a guideline (using good custom profiles, an appropriate proofing paper, etc.). But I always have them generate and provide me their own proof, and assuming it is acceptable, I sign off on their proof, not mine. That is my contract with the printer, and he is obligated to come close to his own proof. If his proof is very different from mine, I'd stop and figure out what is wrong early on, but I really haven't had that problem before. SWOP guidelines allow +/- 3% dot gain and +/- 0.10 for solid ink density on each individual ink, which is quite a bit of leeway. A good printer will do a lot better than that. If you can measure that they exceeded these limits, then they "should" reprint the job at their expense.

The printers I have worked with (and had a good relationship with) were pretty flexible and accomodating at press checks. I tried to focus on the important stuff and not be overly picky. I always asked the advice of the pressman and we discussed making various adjustments to get as close as possible to the proof. They wanted to please me, but they also wanted to run the job and get onto the next one, so it was a matter of getting close, but not spending hours nitpicking every little thing.

For images separated with less GCR, adjustments to color are easier on press (but colors can also vary more during the press run). In my experience, yellows are fairly easy to adjust on press, since they contribute less than C, M or K to overall density and small adjustments won't unintentionally turn your job into a low key or high key job. Heavy GCR is designed specifically to reduce variation on press and to hold neutrals, so press adjustments to change color can be more difficult or sometimes impossible if you want a big enough move.

Unless you have a bulletproof black plate, you should be able to make reasonable adjustments to color and density on press. Of course, you need to balance type, blacks, images, etc, and an adjustment for less black in an image might leave you with weak looking type, so it is a compromise. If the printer cannot meet their own proof within acceptable limits (sometimes a tough thing to establish or prove) then they have a problem. If they are intractable and difficult and unwilling to make changes during a press check, then perhaps it is time for a serious discussion and possibly a new printer if they aren't responsive.

You mentioned that you are submitting contract proofs to the printer and asked them to meet them. If so, you probably need to know more about what guideline they are print to and make sure they are ABLE to match YOUR proof.

Hope some of this is helpful.

Lou Dina
___________________________________________________________________________

Re: Control during a press check
Posted by: "Lee Clawson"
Sat Feb 10, 2007 1:01 pm (PST)

Michael,
If you're bringing proofs to them you may indeed be "humored". Usual practice is the contract proof is submitted to you from the printer. And then you can expect a reasonable effort to be made to match as close as possible. You do have minor leeway in shifting color on press. Its important to keep in mind that an alteration can affect everything printed.

I suggest meeting with the printer between jobs and asking them these questions. As a starter you might compare your by-the-book settings to theirs.

Lee Clawson
2/\V/\7 Studio
___________________________________________________________________________

Re: Control during a press check
Posted by: "mfedecky"
Sat Feb 10, 2007 1:01 pm (PST)

I'm not sure I grasp why there is a plus or minus 3% for dot gain and only .1 per cent for solid ink density in each individual ink. I don't think I myself could see the difference.

Another problem I've had crop up is setting up images for a web press only to have it switched to a sheetfeed press, which can hold more ink and needs less allowance for dot gain. And yet the job ends up looking absolutely great. So this is what has led me to start questioning about how much manipulation of ink occurs on press.

Michael Fedecky
___________________________________________________________________________

Re: Control during a press check
Posted by: "Terry Wyse"
Sat Feb 10, 2007 3:08 pm (PST)

On Feb 10, 2007, at 1:12 PM, mfedecky wrote:

I'm not sure I grasp why there is a plus or minus 3% for dot gain and only .1 per cent for solid ink density in each individual ink. I don't think I myself could see the difference.

I'll bet you could! A 3% midtone variation can have a profound effect on overall "weight" and gray balance especially. As far as +/-.10 solid ink density, you may not notice this as much in the solid primaries/secondaries but the side effect of a density change is more or less dot gain/TVI which, again, can be quite noticeable.

Another problem I've had crop up is setting up images for a web press only to have it
switched to a sheetfeed press, which can hold more ink and needs less allowance for
dot gain.

Not really all that true. A sheetfed press can have similar dot gain to a web press by virtue of it's (typically) higher solid ink density. An increase in density raises TVI levels unless compensated for by plate curves.

If you frequently have jobs that switch between web and sheetfed, you should look into creating separations using the (just released) "G7" family of ICC profiles. This is basically a set of 3 profiles, GRACoL2006_Coated1 (sheetfed on ISO paper #1), SWOP2006_Coated3 and SWOP2006_Coated5 (web on ISO papers #3 and #5) that all have very similar tonality owing to the fact that the "G7" method was used to generate the press data and thus the profiles. Since the tonality is designed to be the same, you can basically re-purpose images separated with, say, SWOP2006_Coated3 and use them on either sheetfed or web. Highlight to midtone tonality and gray balance will be virtually identical with the main difference being different solid ink density aims and total ink settings in the profiles. Since the two SWOP versions are set for 300%, this is the safer one to separate to rather than separating for GRACoL2006 which has a 320% ink limit. The reduced ink limit of the SWOP profile will still print fine on a sheetfed press while the opposite may not be true. You can find this "G7" set of profiles at www.gracol.org .
___________________________________________________________________________

Re: Control during a press check
Posted by: "Terry Wyse"
Sat Feb 10, 2007 3:13 pm (PST)

On Feb 9, 2007, at 3:11 PM, mfedecky wrote:

My question relates to how much control can be exerted during a press check by the
designer. Assuming I have done everything by the book i.e. I've color corrected by the
numbers. I've made my color separations taking into account dot gain, total ink limit etc etc.

Did you take into account the dot gain and total ink limit of the SPECIFIC press your job is going to print on? While total ink limit isn't so critical (a range of around 280%-320% can be handled by any decent web or sheetfed press run by a competent pressman), dot gain can vary over a fairly wide range depending on a whole host of mechanical and chemical issues, not the least of which is whether your printer has made any attempt to adjust their plate curves towards a specific dot gain target for their various papers and screen resolutions.

I have even submitted contract proofs (matchprints or an accurate
digital proof) and told the printer to match them.

Not to pick nits, but it's not a contract proof if the printer hasn't agreed that your proofs can be used as a "contract" between you and them. I would dare say that true analog "Matchprints" cannot be matched by most printers today unless they've adjusted their plate curves to achieve "legacy" dot gains of 20-24%, to say nothing of the fact that analog Matchprints and most digital dot proofs have perfect "wet trap" characteristics that cannot be matched on press unless compesnatre for via color management or some other means. Most linear CTP systems these days have dot gain closer to the teens (13-18%) so it's going to be virtually impossible to match a Matchprint or any digital dot or inkjet based on a Matchprint simulation. Suffice to say, you can't just supply them with a proof that YOU consider a contract proof and expect them to match that unless you've communicated with them before hand and you both agree on what the proofs should target. And it's not enough to agree on a "SWOP" proof because that can be interpreted pretty broadly. I would suggest having the printer produce the contract proof for you OR have them provide you with very specific information, preferably in the form of an ICC profile or perhaps a colorimetric data set, that will allow you to produce a proof matching their print conditions. Even then, be prepared to VERIFY that your proof does indeed match their specifications to an agreed up tolerance (I would suggest 3 delta E or less).

And yet when I'm at a press check, and I think the midtones look
heavy or the fleshtones look too yellow, can significant changes be made at this point, or is
pretty much etched in stone. Can you pin a numerical value on this? What is the maximum a color can be altered on press? Whenever I ask for a change, I feel
the pressman is humouring me and nothing changes.

Fact is, an offset press is NOT a very good color correction tool! Basically, the only DIRECT adjustment they have at the press is solid ink density (technically, ink film thickness) and, indirectly, dot gain (TVI) and gray balance. In other words, there's nothing like a curve adjustment on the press itself or a reasonable way to affect local changes without affecting the entire job. It's a mistake in my opinion to rely on a "press check" for final color correction for what should've been caught at the proofing stage and adjusted in prepress.

Regards,
Terry Wyse

_____________________________
WyseConsul
Color Management Consulting
G7 Certified Expert
704.843.0858
http://www.wyseconsul.com
http://www.colormanagementgroup.com
___________________________________________________________________________

Re: Control during a press check
Posted by: Dan Margulis
Sun Feb 11, 2007 5:14 pm (PST)

Michael Fedecky writes,

Another problem I've had crop up is setting up images for a web press only to have it
switched to a sheetfeed press, which can hold more ink and needs less allowance for
dot gain. And yet the job ends up looking absolutely great. So this is what has led me to
start questioning about how much manipulation of ink occurs on press.

There's a major difference on press between files that are initially too dark and those that are initially too light. In the first case the pressman has to cut back inking. This may make a difference for the better, but it will definitely look worse than if the file had been prepared correctly in the first place, because areas that are correctly dark, such as the shadows, will print lighter than they should and contrast will suffer.

If (and this is normally the case when files prepared for web are printed sheetfed) the incoming file is too light, the pressman compensates by adding more ink. Now, it *is* possible for the job to look as good as you would have wanted to. If there is critical detailing in the shadows it may get lost, but otherwise it should look just about as good as if you had prepared the file properly.

This is an important lesson for those without a lot of CMYK experience who are preparing files for press: you're better off in the long run if you submit files slightly lighter than your best guess of how the press will actually print it.

Dan Margulis
___________________________________________________________________________

Re: Control during a press check
Posted by: "Laurentiu Todie"
Sun Feb 11, 2007 9:29 pm (PST)

that,.. and you'll avoid hickies from keeping the ink level too low : )

Laurentiu Todie
___________________________________________________________________________

Re: Control during a press check
Posted by: "J Walton"
Sun Feb 11, 2007 11:30 pm (PST)

Good point! What'd my old boss say...

"Lots of ink and lots of squeeze helps a printer print with ease."

Does anybody have a rhyme for "shadow detail?" I'm trying to finish that line.

-----
J Walton
___________________________________________________________________________

Re: Control during a press check
Posted by: "Laurentiu Todie"  
Mon Feb 12, 2007 11:07 am (PST)

On Feb 12, 2007, at 12:37 AM, J Walton wrote:

Does anybody have a rhyme for "shadow detail?" I'm trying to finish
that line.

"squeeze too much and you won't know where the shadow detail is"

(not my best ever, but has a twist : )

Laurentiu Todie
___________________________________________________________________________

Re: Control during a press check
Posted by: "mfedecky"
Mon Feb 12, 2007 4:12 pm (PST)

Not really all that true. A sheetfed press can have similar dot gain
to a web press by virtue of it's (typically) higher solid ink
density. An increase in density raises TVI levels unless compensated
for by plate curves.

So I take it there are no standard 'dot gain values' for Sheetfed presses (coated or uncoated stock), especially if the Total Ink Limit can be anywhere between 320% and 340%. (Especially in view of Dan's contention that some printers don't know their own dot gain numbers.)

Thanks to everyone who has replied for the great information.

Michael Fedecky
___________________________________________________________________________