Thursday, May 14, 2009

Grey Balance Unbalanced – An inconvenient truth, part 2 of 2

Sorry, this post has moved.

Click

HERE

to view.
Posted by
Labels: , , , , ,
Your reaction: 

16 comments:

  1. OK Gordo.

    I think I follow your basic point that a 3/c gray in a color bar isn't going to actually represent anything else on your press sheet (unless you're printing test forms or something).

    But if gray balance is not a suitable process control metric for the reasons you outline, what do you feel we should be looking at? Solid densities? Something else?

    Your sample form with the fruit is a great illustration. Let's assume your prepress is all dialed in and you don't typically need to manage color on press (big assumption, I know). Are you saying the gray patch will vary edge to edge due to press
    mechanics or the psychological factor of the customer saying "lets bump the cyan on this side and the magenta on the other"?

    Great topic, as usual.

    -sean

    ReplyDelete

  2. To play devil's advocate, if you have to pump up the color in the blueberries and tomato, you have a flaw in your proofing system. If the colors aren't vibrant enough on press but are so on the customer approved proof, you have issues more involved than hitting a grey balance target.

    And there are industry standards for lighting as I am sure you know.

    I believe these two points are valid in perhaps a textbook and hit the fan in the real world, but none the less, the issues you bring up are dealt with.

    I don't believe grey balance is the holy grail of printing but in general, if your proofing device and your press can both hit the same grey balance (be it industry standard or not), you are in pretty good shape as far as being able to match your proof. And matching a proof is really what everyone is trying to achieve here.

    Even if we all see color differently, except for the color blind, there are obvious mismatches in greys and we all fall within a certain tolerance. And where we differ (we being the printer and the client) can usually be resolved on press. Assuming your proofing system is set up right.

    Very interesting posts. Keep it up. I'm always happy to see a new posting on your site. A great source of information. Thank you.

    ReplyDelete

  3. Thank you for taking the time to comment.
    Once the solid ink densities for the press run have been established, they become the primary metric for maintaining consistency through the run. The reason SIDs have that role is because that is the primary function presses are designed to perform - i.e. lay a 1 micron film of ink on a substrate.
    Yes, the gray patches in the press form with the blue berries and tomato images will vary edge to edge. This is because of the extreme ink coverage differences across the width of the sheet, not because someone wants to play with color on press. It's a mechanical ink transfer issue that presses have but does not happen on a proof.

    best, gordo

    ReplyDelete

  4. Hi Gordo,

    Great stuff...sure to have lots of feedback. ;) Regarding the fruit sample. It has not been my experience that image content affects gray balance as drastically as implied. In fact, looking through draw after draw of print samples here, I'm finding image content has had very little if any impact impact on the either SID or gray balance on our press forms. Where densities are hitting their targets, gray balance is being achieved regardless of content (and though we're not printing A3 size tomatoes, there are some very saturated content).

    Regarding CMY gray not equating CMYK gray, I agree. Live seps will be more stable that a CMY control element. But the gray balance patches of 50C, 40MY are still relevant for a quick reference to integrity of the print. Its akin to the canary in the coal mine.

    Totally agree with issues regarding visual analysis. Measurement is better, though not perfect.

    ReplyDelete

  5. To Len (the Devil's Advocate :-)),

    The reason the press operator would pump up the color in the blueberries and tomato example is because of their extreme ink usage coverage. It's the result of how ink is transfered on press vs a proof. The result would show up primarily in the 3/ gray patches (because of their over sensitivity) but not so much in the SIDs (because they are high saturation colors).
    In general presswork you may not see this ink usage is as strongly - my example was extreme to illustrate the issue.
    Yes, there are industry standards for lighting, and they should be used whenever color is discussed - however, print buyers are not always so understanding.
    I agree that matching your proof is what everyone is trying to accomplish - however I disagree that hitting the same gray balance on proof and press sheet will get you there. If your primaries are different, your overprints are different, you can have gray balance in both but the colors will not match. Interestingly, this fact is also noted in the G7/GRACoL documentation which is gray balance-centric.

    Thanks for taking the time to comment - it's much appreciated!
    best, gordo

    ReplyDelete

  6. Great follow up post Gordo!

    Touching upon the differences in color perception with age and gender differences and the influence Metamerism has on color rendering are topics too seldom dealt with in my opinion.

    Great comments and discussion all around!
    Otherthoughts

    ReplyDelete

  7. I wonder if a side benefit of the G7 / gray balance approach is that it forces you to collect and evaluate hard data in the middle of the tone curve?

    -sean

    ReplyDelete

  8. To Anonymous/sean
    I think the benefit of G7 is that it gets printers to actually look at their process from a data-based rather than an intuitive point of view.

    thanks for your comment - gordo

    ReplyDelete

  9. The traditional 3/c gray values - and their intended purpose – are still valid if you use "good old" skeleton black separations. Many repro/printing houses still use them for various reasons.

    Heavy GCR is a very effective way to stabilize gray balance and reduce ink consumption, but that means one needs a very high quality (very dense) black ink to get really black text along the color images. We had many issues of this phenomenon, and the final decision was to go up to 320 % TIL with a more traditional UCR separation - on a web press. On sheetfed we still use 340-350 % TIL for the same reason.

    Lighting should be perfect all along the prodution line - in fact this is the cheapest investment, it is a complete no-brainer to spare here. And with good quality papers/proofing substrates (with low or zero optical brightener) the metameric effect should not rise problems.

    Regards,

    Peter

    ReplyDelete

  10. Gordon,

    If gray balance is not the best tool for process control; what do you think we should use?

    Thanks for sharing your knowledge,
    Curtis

    ReplyDelete

  11. @Curtis - IMHO
    #1 - solid ink density and hue
    #2 - RGB overprints and hues
    #3 - dot gain/TVI

    ReplyDelete

  12. Gordon,

    "...the benefit of G7 is that it gets printers to actually look at their process..."

    Wow, that says it all.

    stephen

    ReplyDelete

  13. It think you may have missed the point. The grey balance target is a control strip that offers an easy visual check that your densities and TVI are in balance across all 4 plates. (Yes all four. The black only patch should match the lightness of the CMY patch.)

    It assumes that prepress has things properly calibrated so that it your TVIs are correct along the full scale.

    Achieving neutral greys in the live areas is not the point - that is just a side effect of having the correct densities on a properly calibrated system.
    The CMY grey isn't meant to be the same as the live image grey. As a control it is meant to be as sensitive as possible.

    The ink-usage issue is a problem, but in the end the control is only reflecting (in a more sensitive manner) the same problem that you will have in the live image areas.

    Metamerism isn't an issue since it is only designed to be evaluated under a single controlled lighting condition.

    The grey balance target is not a magic bullet that will instantly make everything fall into place, but it is certainly a very useful tool in helping to get there

    Regards

    ReplyDelete

  14. @Ryan,

    Nope, I don't think I've missed the point.
    I've explained in some detail why the grey balance patches in the color bar are of little to no value. You can disagree with me, however, you need to explain the mechanism by which the grey patches provide relevant information.
    You say that "The grey balance target is a control strip that offers an easy visual check that your densities and TVI are in balance across all 4 plates."
    I think that's a myth as far as the grey balance target in the color bar is concerned. And there's no research (that I'm aware of) that demonstrates that the grey balance target in the color bar reveals anything about the densities an TVI being in balance.

    The purpose of the color bar is to provide a consistent target across different print projects that can be measured as a process control tool. In that context, the grey balance patches provide very little useful information. They may be of some use forensically when something goes terribly wrong, however, as a guide during actual production printing the gray balance patches are for the most part either irrelevant or useless for the reasons I described.

    One of the reasons that the grey balance patch is useless is precisely because of its sensitivity to minor SID variations. It's telling the press operator that variation is happening in the grey patch - however that variation has no relevance to the live image area which is what is important.

    ReplyDelete

  15. Gordon,

    I explained the mechanism. Just because you think it's a myth doesn't change that. I'm not sure what needs a study. Assuming the CMY grey patches are set up to whatever standard you are printing to (even if it is an in-house standard) then by definition they will match the K patch when the densities and TVI are in balance. (Of course they could all be out by a matching amount making the print overall too dark or light.)

    Being sensitive is not a problem, it just means it will not always look perfect and as a press operator you need to learn how far out it can be before it starts to cause issues. It's good to have an indication that the press is starting to stray *before* it moves far enough to cause noticeable problems in the live image area.

    Regards

    ReplyDelete

  16. The grey patches in the color bar indicate changes in the grey patches in the color bar. While that may indicate that the same change is occurring at the same tone value in the live image area (e.g. the 40% M in the grey patch reflects what is happening to the 40% tone in the live image area) my contention is that information is, in the vast majority of presswork, irrelevant.
    Put another way, a 2% shift in the magenta in the grey balance patch will be very noticeable in the grey balance patch itself, but will not be visible in the live image area. So the press operator will see the grey balance patch vary but no action is required to correct it. So the signal the grey balance patch is giving is irrelevant.
    The press operator really only has one control - solid ink density/film thickness (and subsequently water to stay in balance). S/he only has a narrow window of SID before lithographic problems like scumming, tinting, slinging, set-off, etc. begin to manifest.
    Maintaing those SIDs within tolerance will maintain the color integrity of the live image area irrespective of what shifts are occurring in the grey balance patches.
    When a press is running there are two types of variation that one sees - the normal, natural, cyclical variation of SIDs which will show up in the grey balance patches but for which not adjustment by the press operator is required. I.e. they might as well ignore the grey balance patches. And the other, a slow directionally change, or drift, e.g. plate wear. This kind of drift may not show up in variation in the grey balance patch but it will reveal itself during press pulls as the the job is running.

    ReplyDelete

Subscribe to: Post Comments (Atom)