G7 gray balance

[disbellj]

I made spreadsheet that automates the manual fangraph paper method (not that I recommend anyone doing it. I'd recommend IDEAlink Curve2, but I built my own and it works).

I got to where I'm about to do the calibration run to get to paper dependent solid and overprint Lab value targets (within tolerance), and then fix NPDC and graybalance (and TVI) after that, to be verified during the verification run.

Problem is:
I have paper dependent gray values I'm trying to find on GrayFinder target, then correct individual TVI to obtain that dependent gray balance when on press.

But at press, during verification run, the pressmen won't be able to use densities for All (VCMY), - Paper, and target EQUAL settings, will they?

They'll have to target Lab to see if gray patches are close to dependent gray values for a* and b*, and then maybe find the densities: All (VCMY, -Paper), that they'll run to for production?

Thanks,

Don

 

[meddington]

during verification run, the pressmen won't be able to use densities for All (VCMY), - Paper, and target EQUAL settings, will they?

They'll have to target Lab to see if gray patches are close to dependent gray values for a* and b*, and then maybe find the densities: All (VCMY, -Paper), that they'll run to for production?

Hi Don,

Not sure I'm following, but two school's of thought...target the solid/overprint Lab values for closest match to targets and rely on curve adjustments to tame gray balance and tonality, or adjust solids via inking, presumably within tolerances, to help optimize gray balance while relying less on curves. If you have the luxury of time and material, you could put on a test form to print low to high densities, recording the wet density values. Let dry then measure Lab values, deriving the optimal wet density value that yield optimal dry Lab. With this ammo, go back to calibration run at target wet densities and determine necessary curve adjustments for gray/tonality.

 

[disbellj]

Michael,

Thank you for responding.

I'm trying to find the best way to explain this.

You know how G7 method now has paper-dependent solid primary aims? And how G7 method now has paper-dependent gray balance?

Let's say that after I get my solids within 5 ΔE of the paper-dependent values, then the Calibration run is done.

After that, I let the target dry, and measure the P2P25 and GrayFinder targets. On the GrayFinder target, I see that to get to my paper-dependent gray values (not a* = 0, b* = 0), I need to move my Magenta and Yellow TVI separately from the Cyan TVI (since aim of Calibration run has been clarified as to bring solids and overprints within ISO tolerance, I didn't achieve gray balance during the Calibration run, and now have to use GrayFinder and draw 3 separate lines on CMY NPDC Fangraph 2009, one for C, one for M, one for Y).

I make these moves on plate that I need to for K, C, M, and Y.

Question is:
I know that some have said that you can just use Density set to All on spectro-densitometer and at press make sure that C=M=Y density (I believe the older G7 documentation said this also), and you have gray balance. BUT if I'm attempting a paper-dependent gray balance, I wouldn't aim for C=M=Y density in the gray patches at press would I? I would aim at a* and b* paper dependent values at the gray patches during Verification run, and then note the C,M,and Y densities using All setting for Density, and those would be the densities to aim for during Production?

I hope I have been able to express what I'm talking about better. Attached is a screenshot, using my proof paper as an example.

Regards,

Don

Hi Don,

Not sure I'm following, but two school's of thought...target the solid/overprint Lab values for closest match to targets and rely on curve adjustments to tame gray balance and tonality, or adjust solids via inking, presumably within tolerances, to help optimize gray balance while relying less on curves. If you have the luxury of time and material, you could put on a test form to print low to high densities, recording the wet density values. Let dry then measure Lab values, deriving the optimal wet density value that yield optimal dry Lab. With this ammo, go back to calibration run at target wet densities and determine necessary curve adjustments for gray/tonality.

 

[meddington]

I wouldn't aim for C=M=Y density in the gray patches at press would I? I would aim at a* and b* paper dependent values at the gray patches during Verification run, and then note the C,M,and Y densities using All setting for Density, and those would be the densities to aim for during Production?

OK, I think I get it now. Yes, gray balance is best verified via Lab measurement, then you can measure and determine the target density values for CMY, which may or may not be equal filter values.

 

[disbellj]

Thank you again Michael. This is what I thought. Painstaking process that GrayFinder target is getting a* and b* close to the target (actually just working with it testing it's getting better/easier).

Yeah, I'm kind of wondering now that you said something:

I was planning to just go to press and get pressman to make density adjustments (knowing how increasing density makes L* go lower while making a* and b* go higher) until we hit in tolerance on all 7 (plus paper), then get him to run a 500-1000 sheets, make sure we're still in tolerance, and then pull me some sheets to dry and measure in prepress.

But you said something about basically making 3 runs for each paper type. One to get sheets run from low to high density, letting them dry, and seeing which hits the Lab solids when the sheet is dry.

One method goes for lab solid aims when sheet is wet. The other goes for lab solid aims when the sheet is dry. Does it matter much, or is there a correct way?

Regards,

Don

 

[meddington]

But you said something about basically making 3 runs for each paper type. One to get sheets run from low to high density, letting them dry, and seeing which hits the Lab solids when the sheet is dry.

Nice thing about this is you don't need a really controlled pressrun to perform, and if needed you can send more than one paper type through the press for measurement without a lot of set up time. You are after all only looking for a nice range from low to high density to measure wet density/dry CIELab.

One method goes for lab solid aims when sheet is wet. The other goes for lab solid aims when the sheet is dry. Does it matter much, or is there a correct way?

Whatever way gets you within target tolerance is the correct way. If aiming at "stadard" LAB targets for ISO, etc, those are dry values, so accounting for dry back is an assumption made by those writing the standards/specifications.

 

[disbellj]

Michael,

Thank you once again. If I do it that way, it might be an extra press run, but most likely civility will be better kept between press and prepress.

Regards,

Don

Nice thing about this is you don't need a really controlled pressrun to perform, and if needed you can send more than one paper type through the press for measurement without a lot of set up time. You are after all only looking for a nice range from low to high density to measure wet density/dry CIELab.




Whatever way gets you within target tolerance is the correct way. If aiming at "stadard" LAB targets for ISO, etc, those are dry values, so accounting for dry back is an assumption made by those writing the standards/specifications.

 

[TerryWyse]

Personally, I don't think using "All Densities" and shooting for equal values in C, M and Y for gray balance works all that well, at least not within about +/- .03 or so....

.....but to answer your question, you can sort of use RGB (C, M, Y) filters to get in the ballpark of paper dependent values but you MUST zero out on paper first. This will make the gray balance "relative" to the paper tint which is basically the G7 method for getting gray balance.

The reason I say "it mostly doesn't work" is this:

* If you measure gray balance using absolute density, proper gray balance will not be equal densities....it will be in the direction of the paper tint (GRACoL gray balance will measure a bit too cyan).

* But if you measure gray balance RELATIVE to the paper tint, the opposite happens. In the case of GRACoL neutral gray, the yellow density will measure a bit high relative to cyan and magenta densities because you've removed the -b* of the paper.

I would say to use either L*a*b* (roughly half the paper a*/b* will give good gray balance at the midtone) or simply do it visually. Given our visual system's sensitivity to neutral grays, it's pretty darn good at figuring out what's neutral and what isn't...although I'd tried having a good neutral reference nearby, such as a Color Checker chart, as a reliable reference. Depending on your black ink, it can't be trusted as a neutral reference on the press sheet although it's very tempting to do that.

Terry

 

[disbellj]

Terry Wyse,

And you said you never met a prepress person that ever set up G7 on their own without a G7 Expert. I guess you have now LOL. For real though, I've asked questions and been given input by multiple G7 Experts, so I do NOT consider that I did it by myself, even though I did have to practically do it myself here where I work.

Solid densities to obtain in-tolerance solids (with lowest Delta E) was written down for pressman after the pre-calibration run that Michael described. These are our production densities. This did indeed help me to stay out of the pressman's way, get what I needed (wet densities and dry Lab values), which allowed me to give them the densities to run to for the calibration/verification/productions runs. After calibration run, I found all solids still in tolerance, and overprints in tolerance, although papers are not.

Went ahead and made gray balance moves at all gray balance percentages in GrayFinder, and moves all along NPDC scale. The resulting verification run looked horrible. Now I knew I had implemented solid Lab aim correction for paper error, and paper dependent gray balance also, so I was wondering where I went wrong since I was accounting for the latest developments that I knew of.

It turned out I was doing way too many calculations at too many points which made my verification run look bad. So I went back to my calibration readings, used 50% patch only for gray balance M and Y moves (this was very important), and actually ended up just using the 50% dot in my calibration program that makes plate curves, letting the program smooth the data.

Got on press and we're in tolerance on NPDCs now for coated and uncoated.

So gray balance is what I need to check to make sure it's close to what I aimed for, but since the loaner 528 had to get sent back as soon as we did verification run (the first one that looked horrible), and we haven't had our 508 upgraded with Color, then I'm unable to do that yet. The 528 did help us get to this point no doubt though. It's just that X-Rite wanted us to test a device before we ordered the functionality to be added to the 508, to make sure it was indeed what we wanted. It was.

As for gray balance, with paper dependent gray balance (altough I get the pressman to take density readings without paper, of 50K and 50CMY patches for production control), then I'm just having a little trouble figuring out what the exact readings should be if using All densities, but I can say that at this point all readings are within .02 of each other when using All densities and measuring the paper first, so I'm not too worried about it (although .02 density = 4% when talking TVI aka the horizontal scale of the NPDC FanGraph).

Although nobody really wants to describe it as TVI, that's exactly what the NPDC is: a tying together of "gray" tone curve, density, gray balance (when used in conjunction with GrayFinder), and dot gain (aka TVI) IMHO. For every .01 density moved up, it equals 2 "dots" (TVI/percentage of 100) moving right. That's the NPDC chart as I see it, and I have studied it to try and understand it the best I can. Looks pretty cool and simple now. I actually really like how Don Hutcheson tied it together. Makes perfect sense honestly.

I like the fact that I can at any point (or my boss, etc.) measure 2 spots on an already-ran and dried press sheet, and tell if the pressman was in tolerance of what he was supposed to be. It gives us quality control, which we haven't had this precisely before.

And you want to know what I REALLY love about it? The fact that I can now tell our customers what to use to build their jobs if they are interested, since the "press" profile that was made years ago of our custom press was made by proofing software, never meant for soft-proofing or conversion of production files, and the paper when soft-proofed looked either bright yellow or bright magenta or black. Not really something I could have given to our customers and said "Use this". Now I can say "Use this official (for GRACoL2006_Coated1) or beta (for uncoated G7 beta) ICC profile, depending on what paper you'll be printing on". I can also now say "You can use these profiles for conversion of their Lab, RGB to CMYK (I can give them a grayscale profile that is just the black separation of the CMYK ICC profiles for grayscale printing), soft-proofing (and I can give them settings for both what color temperature we aim to: 5300K, as well as Gamma 2.2, and recommended luminance for LCD or CRT monitors, and can also give them PhotoShop settings for soft-proofing), and also hard-copy proofing. I love how these profiles can be used for everything needed. And until I get customers that understand, they can continue to use SWOP images and it'll be just fine. I would recommend highly that IDEAlliance make a coated profile with lower TAC like Europe did years ago after I said that it would be good so that the coated separation could be used on uncoated paper also. They must have seen the importance of what I was saying, but until this also happens in the U.S., then practically I don't see myself switching between 2 ICC profiles in my Adobe and Quark programs all the time. I just use their defaults (except I do replace QX 6.5 and under's color profiles and use Adobe's default North American General Purpose 2 profiles instead, so no blues go purple. Quark's old ICC profiles were junk IMHO), and have no problems.

But as I have said multiple times before and will say until the day I die, I would not recommend anyone going through what I went through. Get a G7 Expert to deal with bosses and pressman. Stay out of it if you're prepress as much as you can. Pressman don't like prepress telling them what to do. It wasn't until I got done and was able to take a couple readings and tell the pressman what he needed to do to bring things in tolerance, that I actually got cooperation I needed, without the BS to go along with it. So to re-iterate, I wouldn't have done it myself if I saw another way. There's no reason for that much pressure to be put on prepress, when G7 is really about what's going on at press, and with IDEAlink Curve 2, the pressman can tell prepress what changes need to be made to plate curves - that's all prepress has to do with G7. Proofing: We use Certified proofing to get to standards/specifications. Workflow as far as Adobe/Quark/Microsoft: G7 don't care about that and doesn't really address that much, other than SWOP images look fine printing on other G7 paper/ink/press combos. I use Adobe's North American General Purpose 2 default profiles in Quark and Adobe (Quark 7 and above uses basically Adobe RGB(1998) by default, although ICC profile named different in Quark, but since QX7 the profiles are so much better in Quark than 6.5 had, that I go with defaults in QX 7 or above if not using legacy profiles. I won't use Quark's legacy profiles), and I don't see them changing anytime soon.

Regards,

Don

Personally, I don't think using "All Densities" and shooting for equal values in C, M and Y for gray balance works all that well, at least not within about +/- .03 or so....

.....but to answer your question, you can sort of use RGB (C, M, Y) filters to get in the ballpark of paper dependent values but you MUST zero out on paper first. This will make the gray balance "relative" to the paper tint which is basically the G7 method for getting gray balance.

The reason I say "it mostly doesn't work" is this:

* If you measure gray balance using absolute density, proper gray balance will not be equal densities....it will be in the direction of the paper tint (GRACoL gray balance will measure a bit too cyan).

* But if you measure gray balance RELATIVE to the paper tint, the opposite happens. In the case of GRACoL neutral gray, the yellow density will measure a bit high relative to cyan and magenta densities because you've removed the -b* of the paper.

I would say to use either L*a*b* (roughly half the paper a*/b* will give good gray balance at the midtone) or simply do it visually. Given our visual system's sensitivity to neutral grays, it's pretty darn good at figuring out what's neutral and what isn't...although I'd tried having a good neutral reference nearby, such as a Color Checker chart, as a reliable reference. Depending on your black ink, it can't be trusted as a neutral reference on the press sheet although it's very tempting to do that.

Terry

 

[TerryWyse]

Terry Wyse,

And you said you never met a prepress person that ever set up G7 on their own without a G7 Expert.

And I still haven't. I'd like to meet the prepress person that did a press G7 calibration (and got it right the first time) without A) having watched a G7 Expert to it once-upon-a-time or B) having picked the brains of a G7 Expert(s) enough to figure out how to do it themselves. In any case, they'd never end up with a proper G7 Master Qualification because that requires either the help of a G7 Certified Professional or G7 Certified Expert.

Regards,
Terry

 

[disbellj]

Terry,

I did the best I could with what I had at my disposal. I read all the G7 documentation years back, and got caught up with the latest documentation. If using too many points for gray balance and NPDC correction is SUCH a bad thing, then oh well, I'm guilty of making a mistake. Guilty of fixing it too.

If you think my boss cares about G7 Master Qualification, I think them not getting a G7 Expert out here and not buying IDEAlink Curve 2 says a lot about what they think about G7 Master Qualification. My boss instead putting confidence in me to do it (although it was stress I would have rather not went through and have stated as much) says a lot of what they do think about me. They knew I have studied this whole thing pretty deeply back then and now.

We've matched G7 proofs through using Bruce Lindbloom's freely available LabDotGainCalculator, and through using G7 (me automating the manual FanGraph 2009 method, including correcting solids Lab aims due to paper error, and correcting gray balance AFTER the calibration run, also done with freely available math if one knows how to use google and excel, they could do it too). Doesn't matter to me which is used, but G7 does give process control with a couple clicks so it does have the upper hand.

I knew as much as most G7 Certified Experts even before my practical implementation, and even if you don't agree with me, what I've done speaks enough to me to say I did it, and have done it correctly.

I'm not recommending anyone go through what I did, but I'm happy with my results, which are measurable as being in tolerance for those measurements which include tolerances. Having said that, I will repeat that I recommend people (just like I did to my boss) to get a G7 Expert and IDEAlink Curve 2. Do I consider what I did to be wrong? No. Followed all instructions, except initially overlooking how many points were needed for press calibration and gray balance calibration. Re-reading the documentation once again helped me on that. IF someone would have not tried to sell a G7 Expert when my boss inquired about IDEAlink Curve 2, he probably would have bought it. He was inquiring after all. They just didn't have the money for an Expert when they already have me is my understanding of their thinking.

Edit: BTW, this is disingenuous as hell Terry:
"And I still haven't. I'd like to meet the prepress person that did a press G7 calibration (and got it right the first time) without A) having watched a G7 Expert to it once-upon-a-time or B) having picked the brains of a G7 Expert(s) enough to figure out how to do it themselves."

Maybe you could find a prepress person that did a G7 calibration right the first time by getting everything out of thin air without training. If that's what you're looking for, good luck. You'll never find him. I'm the closest you're gonna find my friend.

I've paid my dues, read the G7 materials multiple times each, and no I didn't just go about picking every Experts brain I could find. Besides talking to you, I talked to 2 other Experts who are both on this forum (so I didn't have to look far, and it was politely asking questions on a couple things - no big deal), and they BY NO MEANS gave me ALL the information I needed to do what I did. I went out and got math that is not in G7 documentation to do what I did. If anyone doesn't respect all that, I consider it their problem, not mine.

Don

And I still haven't. I'd like to meet the prepress person that did a press G7 calibration (and got it right the first time) without A) having watched a G7 Expert to it once-upon-a-time or B) having picked the brains of a G7 Expert(s) enough to figure out how to do it themselves. In any case, they'd never end up with a proper G7 Master Qualification because that requires either the help of a G7 Certified Professional or G7 Certified Expert.

Regards,
Terry

 

[J]

And I still haven't. I'd like to meet the prepress person that did a press G7 calibration (and got it right the first time) without A) having watched a G7 Expert to it once-upon-a-time or B) having picked the brains of a G7 Expert(s) enough to figure out how to do it themselves. In any case, they'd never end up with a proper G7 Master Qualification because that requires either the help of a G7 Certified Professional or G7 Certified Expert.

That's what happens when you have a consultant creating specifications and methodologies.

J

 

[TerryWyse]

Don, I'm not trying to be a jerk about this and I (mostly) applaud your efforts. All I'm really saying is look at the amount of time and effort you've put into this over the past couple of years. Maybe you did it all on your own time or maybe not, but I guarantee the way you went about it was WAY more expensive in the long-run than hiring an expert in the first place.

"J", I'm not sure what you mean when you say:
"That's what happens when you have a consultant creating specifications and methodologies."

I guess all I would say is that I'd rather have a professional who's well versed in specifications and methodologies...be the one to come up with specifications and methodologies. Who else could do it and do a credible job?

Regards,
Terry

 

[disbellj]

Terry,

But you act like I had a friggin' choice. Yeah, I coulda quit. I sure as hell felt like it because it was NOT my responsibility to get my company on board and even get the pressman and everyone here on board. If I could have afforded to hire a G7 Expert myself and get it over with, I would have. I had no such luxury.

I Spent. Not my boss. Me. They paid me the same as they would have no matter what. Yeah, they may have put a half-day on all 6 press runs together. Not a big deal to them if they have the time and already made sure the machine was checked and brought into great working condition (new blankets, etc., etc.). They saved money, and if you say they didn't, I can't even believe how you could justify that.

Who paid? Me. Who spent hours? Me. Do I have knowledge now that I wouldn't have before? Yes. Is it a total loss or waste of time? No. Will I need the calculator again? No. Just like I wouldn't need IDEAlink Curve 2 or an Expert after it's set up right. But if I do, it's easy-cheesy to do it again. Much easier than the first because I've already done it. And this time, it'll take one press run - if that. If our solids are within tolerance of new specification that may come, I might be able to do it without a press run (since I have all the numbers already from the calibration run, and would just choose a new source to aim to, plug in some numbers for the new solid and overprint aims, and my calculator gives me paper-dependent corrected solids aims, and if I'm in tolerance still there, then I have paper-dependent gray balance numbers I'm aiming to, may have to, in PhotoShop with new specification profile, Absolute Colorimetric Intent, use 6 patches on P2P25 target, 1 patch on GrayBalance target, to get 16-bit L* (and a* and b* on GrayBalance target), which when plugged into my calculator, converts to 8-bit L*, then to density, subtracts paper, and gives me the density aim values for my CMY gray patches, a* and b* converted to 8-bit, moves for M and Y given, and my calculator spits out changes to K,C,M, and Y plate curves needed, and I'm done in no time).

For a person that has time on their hands because I sit here almost all day every day with almost nothing to do and get paid for it, I don't mind doing it really. What else would I do? Actually, my boss has even let me work on my other business while I've been here not having much to do after I got done with GRACoL study years ago (until this year), and in the mean time, I've been working on my own thing when jobs are not in here. How many places would allow that? So on one hand, I want more freedom. On the other, I already have more than most. I just have to sit in a building where I'd rather be in a location I want, with location freedom and time freedom. So I work towards it when I can.

But me having gotten it done finally puts it behind me and we have verifiable quality control and aiming to an ISO standard on coated paper - per the method G7 for press calibration - per the specification GRACoL 7 - via the official ICC profile GRACoL2006_Coated1v2. As far as uncoated, we're as up-to-date as the beta in the U.S. is, and within tolerances of it.

So there's really nothing anyone should say negative about it, or act as if one doesn't go about it just the way that's recommended, that it's no good. I actually did go by all the math (then some) and automated the manual method. So I still did the manual method, just I can repeat it much faster and do it faster than the manual method. The paper is out of spec because of optical brighteners (lower than ISO standard paper -b) on coated and uncoated. But everything else is in spec.

I can now put this chapter of my life behind me and concentrate on what I consider my next objective/goal I have been working on for about 2.5 years.

This is what you and others may not understand about small business. They've been taken enough over the years, and the margins are tight enough, that something has to be proven to them basically before they will decide to do it, and then they still may not do it for years. They have to see the benefit. And my boss is no different than any other. If you have an employee that looks like they can do something, and look to mostly know what they're talking about, would you spent the thousands extra, or give the guy a chance to prove it to you. I guess my boss wanted me to prove it to them without all the extra money being spent. We already spent a good amount on an ongoing basis, and like I said before, even when we were out of spec, our printing still looked close. Now it's indistinguishable from the Certified Kodak MPI GRACoL proof that we proof.

Maybe my boss is different and extra hard to get on board. But I doubt where I work is that much different than most small businesses where they make due with what they can. And I guess they think how my saying goes: An expert can do for $0.10 what anyone can do for $1. As far as real money spent, I proved that, and got an education in more than G7.

In time and effort spent, I guess it depends on how people look at it. I've spent 2.5 years on something (nothing to do with printing or color management) that if it doesn't pay off, I've never gained much money from it yet, and some would consider I wasted 2.5 years of my life. But if it does pay off, I could possibly, by the Lord's grace, be a millionaire or even multi-millionaire within a year, and never have to work a day again in my life unless led by the Lord to do so, since He will be my only Master. I know what's possible. Have I achieved it yet? No. But I work towards it as if inevitable. Let's say I'm a much better Excel user and MQL4 programmer than I was a couple years ago, and I don't mind putting in the effort to get to a goal, as you see.

I hope you continued success in whatever you endeavor,

Regards,

Don

Don, I'm not trying to be a jerk about this and I (mostly) applaud your efforts. All I'm really saying is look at the amount of time and effort you've put into this over the past couple of years. Maybe you did it all on your own time or maybe not, but I guarantee the way you went about it was WAY more expensive in the long-run than hiring an expert in the first place.

Regards,
Terry