We have run into an issue and I am looking to see what others have experienced in the real world.
Just had a job with color photos as well as flat screen tints. The client was going from a Pantone book that has the PMS color next to their suggested cmyk tints. We were unable to get there on press without damaging the photos.
This was a question I had initially that no one really answered for me. If you have a flat tint where the magenta wants to be 70 percent, when the job goes through the renderer, the tint gets changed to 67.9, based upon the determined curves.
I am wondering what other printers do when they come upon this issue. Obviously I could manually change all the tints to compensate for the curves but this could be time consuming. But, maybe this is what I have to do.
Would really like to hear how other printers deal with this issue.
A 2.1% change is negligible. Plate readers are repeatable to +/- 1%, same with platesetters if not worse. 2% is hard to see on press, except in something like a 50,40,40 gray.
I would determine if the PMS swatch matches my proof. Pantone cannot predict how a PMS color will convert in your rip nor can they predict how a cmyk tint will look on your press. We do not honor outside samples. If our internal proof doesn’t match the provided reference then we edit color in prepress and pull a new proof that hopefully matches. Even if it matches your proof you cannot expect to match every color on press. If a “fair” match is required then proofing Gracol and printing “towards” Gracol will get you through day to day operations. If a perfect match is required, then you will want to profile your press and pray.
If Curve is cutting a 70M down to a 67.9 then that is the requirement for NPDC and graybalance, based on your gray threshhold settings. If your solids are aligned with Gracol and your grayscale is good then you get what you get. The more ink channels used in a tint the less predictable the color will be on press, especially with larger dots. If your 300CMY patch is not neutral then expect distortion in dark saturated colors. I have also found that what’s good for the grayscale isn’t always good for the red, green blue ramps. Curve may tug on a channel to force a neutral grayscale and introduce hue error to the RGB ramps.
Focus on getting your RGB hue angles in alignment with your proof. Keep your KCMY under de5, with a priority towards minimizing hue error. Calibrate your grayscale and that’s it. Color matching is hit or miss. An excellent Gracol setup on press will have IT874 max dE over 5 with 8-10 being common, even if the ave dE around 3. If your PMS/4cp is one of those high dE patches, then it’s time to scramble.
Consider reseparating all your pages with GCR and you may find have an easier time matching more colors.
I am kind of amazed that there have been over 350 views and only one response. No one else is having to deal with this at all?
To try to make my point better I will give a different example. This effect is most noticeable when one of the colors is a solid. Suppose the tint called out is 80c and 100m. Based up the curves we came up with, the cyan will actually end up being 73.8. If the customer has a chip in front of them showing how 80c 100m print my print will not match their chip unless I change the tint combo to 86.2c and 100m. This really has nothing to do with gray balance as far as I can tell. Using the our G7 curves gives us great results on photos but is giving us headaches on flat tint combos.
It sounds like that would be more of a trapping related issue. How were your solid primaries and overprints compared to the Gracol characterization data?
I tend to pay a lot of attention to the 2/c overprints, and also, the 300% CMY patch. The more neutral it is, the the better my overall results. I had a problem years ago, where we had reduced the strength of the cyan ink to get better gray balance before applying curves. This hurt the green and blue overprints, a gave us a strong cyan cast in the shadows of the 3/c neutral ramp. Magenta and yellow were pushed into the 3/4 tones and up. It caused the green and blue tints to shift. You will probably also see a large hook in the ink hue/chroma graph for the green and blue.
Bret, thanks for the response. So your thinking is it is not what the NDPCs do to the tints but how the inks are trapping? My first thought was, nuts he doesn’t understand my problem. But as I think about it I realize you might be on to something and that is maybe why I am not getting more responses to my question.
Okay, I went back to the original data and looked at the spider plot of the primary and overprint ramps. Some look pretty good and some not so good (primarily the red). So what do I do to tweak something like this? Is it the tack of the inks? Or are we talking about the hue of one or more inks? Or is it something else? I would really like to get to the bottom of this as it will certainly simplify production.
Hope to hear from you again soon. On the surface it seems to go against logic and yet over the years I find that logic and color management don’t always seem to go hand in hand at first look.
I would agree with the trapping issue, I don’t know however that its something you can adjust for in prepress and here’s why: We print books in volumes by years with color tabs on the spines that denote each year, those tab colors were picked from a Pantone swatchbook and cmyk seps made by converting in Photoshop and getting the values, years ago, some colors match close while others are a little off, 2 or 3 that we print that have high C and M percentages with very little K give us big color shifts on press at times. The densities can be perfect on the scanner but the color changes from purple to blue to everything in between and our press operator never sees it and it doesn’t get caught until coming off the collator, but that’s a different story, just passing on what we’ve experienced with those colors.
We print G7 and use a maxGCR. Those same colors, only darker, with some K to them, don’t have those color shifts.
Just went thru some testing last week and made up new seps for those problem colors by converting it to how we run now with the high GCR and those colors now have some black in them and didn’t exhibit the problems. One of them was even 78 points less ink separated that way. It was our “1959” color which was originally chosen as PMS#448 and sepped as C-87, M-72, Y-100, K-0, when reseparated using G7 profile and maxGCR it came out as C-39, M-16, Y-57, K-69, so you can see our CM trapping became a non-issue. Not saying this works for all colors but you may try it and see what happens.
Some colors as you know will not be a perfect match no matter what you do. I’d say most will not be and some will be worse than others and some won’t be close enough to use, so those are unmatchable colors, unless you customize the cmyk combo to get closer. My experience is I couldn’t do better than all this expensive software can do.