Predicting/Preventing Moiré

[chevalier]

This is a long read but the details are important.

For most of you moiré this hasn't been a problem since CtF came around. CtP pretty much completely eliminated the problem in commercial litho. In the packaging world this is still a headache and problem. As I came in into the wonderful world of prepress with CtF and very minor film and pasteup shooting (ITEK for duplicators and ancient repeat stuff that didn't have digital files) moiré is not something I have much experience and wisdom with.

I made the jump from commercial litho to packaging over a year ago and now feel that I have a full grasp on the minor differences between the two (personally I don't think it's really much more complicated than pocket folders). The one thing that is massively different between the two niches is the utilization of spot colors. I now work with jobs that are up to 8 inks and many times every ink is a spot color. Sometimes 6+ colors are intersecting with halftone screens. The Heidelberg Prinect workflow I have is somewhat deficient with spot color angle control as it is really designed around commercial litho and patched to work with packaging.

I am limited to 4 angles and 2400DPI/175LPI. I use an elliptical dot. It would be nice to be able to screen some colors at a higher LPI as this prevents/hides moiré. My current angles are Cyan-165º, Magenta-45º, Yellow-0º, Black-105º. I have the option to shift all angles 7.5º, or leave CMK angles alone and move Yellow to 30º or 60º. I can also utilize document controlled screening but I am unaware of any way to set these angles manually. Most of my files are natives handled in Illustrator. I also get PDFs sometimes fully prepped and trapped to my specifications sometimes not. The PDFs generated are then color/converted and trapped (if needed) in PrinectP, imposed with Signa StationP and RIPed with MetaDimension.

I have a Fuji FinalProof halftone-dot proofer but it is rather useless for predicting/preventing moiré as it simulates spot inks with multiple donors at a higher DPI/LPI. I have used it with wrong-color direct donors but when you are attempting to see something that is a visual problem this doesn't help you much. Not to mention the cost!

My current method: When I process a potential problem job I build a spreadsheet listing each ink then listing whether it has screens. If it has screens I then list each screened ink the screens intersects with. I then very thoughtfully start assigning angles and create a brand-specific workflow in my RIP with manually set angles for each ink. I do not use the yellow angle unless the ink is yellow as I learned the hard way the yellow angle LOVES to moiré.

Problems:

• No way that I am aware of to accurately proof moiré short going to plate and press or films. Am I wrong?
• No way to predict moiré other than intuition.
• When a brand introduces a new design or new colors the whole custom angle assignment workflow is shot.
• When a brand uses multiple designs sometimes multiple workflows are needed.
• Getting imposition operator and RIP operator to remember to use the right brand specific workflows as the number is exponentially growing.

Any suggestions, comments, and criticisms are welcome. I am leaning towards pushing for a workflow that is designed around packaging.

 

[Luc St-Pierre]

First, the 7.5 degree shift would help. I also would suggest to always cheat the yellow linescreen by 5 (job at 150 lines, yellow at 155). In flexo, when the expanded gamuts were introduced (such as 7 color Opaltone), we still used only the 4 basic angles, building color pairs that are most likely not to touch (ex: magenta and green that would give a muddy color anyway). Very dark tones (dark brown, deep unsaturated greens etc) combined on the black angles, etc. We had screen viewing capacity that allowed us to look final rip'ed files and also proofing capacities from the rip'ed files, that's pretty much how we checked for moirés. In packaging, flexo especially, a one-size-fits-all workflow is something you can forget. Ease your life by documenting each situation using a job database or modify yours accordingly if you already run one.

 

[chevalier]

I should have stated that we are a folding carton offset litho only operation. I once worked at a commercial litho shop that utilized CMYK+7.5º 2400dpi/175lpi and we had horrible moiré issues with green and brown CMYK builds. It was the most bizarre moiré I have ever encountered as "solid" strip of brown or green on one side of the press sheet was prefect and the other side was wavy looking and inconsistent and on the next run the opposite side would have the issue. It didn't help we printed mostly images of horses and pastural scenes. I was new at the operation and it was one of the first troubleshoots I was assigned. Nobody there could explain to me why they were using that screening-system as their unused DI units were setup using "classic" CMYK angles. Every time they'd have this moiré happen they'd blame the new platesetter and fire up the working DI units. The platesetter was switched to "classic" angles and the problems ceased.

I do understand of the strengths and weaknesses of the various screening models as I have read and reviewed a physical copy of this through 2-3 times. Perhaps upon our next G7 evaluation I will generate the test form with both sets of angles and have a profile created for both.

Unfortunately, I do not have the ability with my current RIP to override LPI per color. It's 175lpi, 150lpi, etc. across the board - all or nothing.

 

[tmiller_iluvprinting]

Is FM screening an option?
Regards,
Todd

 

[chevalier]

AM/FM Hybrid screening is probably an avenue I can explore for some of our customers but I do not currently have licensing for it or FM/stochastic. I'm very interested in what Esko has to offer. But asking for that kind of expenditure is going to take a lot of selling and talking to management. FM/stochastic is probably not an option. We had a competitor utilize pure FM/stochastic screening and they completely lost a huge multi-national consumer products company as a client because of it (we gained the client). The issues were solid logos with very fuzzy edges. The screening system they were using was subpar or the dot size too large.

One of my biggest problem customers is a multi-national pharmaceutical giant and they won't even let us use 175lpi screening. We have to plate their carton art at 150lpi. This being because there are still other folding carton manufacturers out there using equipment so ancient that they can't print 175lpi. They want the flexibility to move between vendors without differences of quality. It really is a joke and insulting to compare us to them but you can't tell your customer that.

 

[gordo]

Some points of clarification as there are some common errors in this thread.

Offsetting screen angles by 7.5 degrees does nothing to help with screen angle conflicts that cause moiré. It does not matter whether the application is offset or flexo. The 7.5 degree offset is designed to eliminate single channel (i.e. one color) moiré in the screen that can happen with halftones that are at non-rational angles (i.e. not 45° or 90°). The 7.5 degree rotation is a solution to a digital screening problem that does not occur with analog screening.

So, angling by 7.5 degrees will not help you.

Except for the size of the 1 and 2% dots there is no difference between an AM or Hybrid AM. Dots are positioned the same way on the same grid.

So, switching to a Hybrid AM screen will not help you as far as dealing with inter-screen moiré.

All screen angles create moiré. The least visually intrusive for 4/C color work is when the screens are angled 30° from one another. That causes a high frequency moiré called a rosette. Unfortunately 30° offsets only allow for three colors - so the fourth color will cause a visible moiré. To make the 4th color's moiré less visible it is reserved for the lightest color - usually Yellow. To further reduce the moiré the Yellow is typically screened at about 108% of the frequency/lpi of the other process color. This happens whether you ask for it or not. It's designed into the screening algorithms.

FM is certainly a possibility as there are hundreds of thousands of labels and packages being done that way everyday world-wide. You still need different screen patterns for each color or you run the risk of visible "clumping" in the over prints. It appears though for your situation you may not be able to use FM/stochastic though.

To your problems:

No way that I am aware of to accurately proof moiré short going to plate and press or films. Am I wrong?

Contact Michael Jahn at Compose USA ( michaelejahn(at)composeusa(dot)com ) He has an inkjet halftone proofing solution that may work for you.

If you have the ability to view the RIPped halftone bitmaps in color on your computer display you might see the moiré if you view the bitmaps at 100% and you stand back from the display. Convince your boss that you need a 70" LED TV to do this. At about $3800 it's chump change for all the headaches it will solve.

No way to predict moiré other than intuition.

Math works but believe me it is brutal. Basically if two halftone screens are less or more than 30° apart - moiré will be visible. If the conflicting colors are dark - moiré will be strong. If they one is a light color and the other dark then the moiré will be less visible. If the areas of overprint are small then the likelihood of a visible moiré will be reduced.

best, gordo

 

[chevalier]

I should mention that before I posted this I refreshed myself on your blog Gordo. It is a great resource! Thanks for the clarification about 7.5º angles. From what I am reading second and third generation FM screening has come leaps and bounds and is worth exploring. We already have two 50" LCDs on the wall here but the resolution is below HD and the color isn't very good as they were bought years ago before I came onboard. I suppose I might be able to convince replacement of them (Did I mention I LOVE the company I work for in this regard?!?!). I use CGS ORIS Color Tuner for a proofing RIP which has a nice dot simulator that I can license or I can license 1bit TIFF output from my plating RIP and ORIS Color Tuner will recombine and proof that way as well. I explored dot proof capability when I bought ORIS and our Epson 9900 with Spectro. The dot proofing works very much the same as our current solution only utilizing FM screening rather than AM screening. I can also buy FM screening licensing for my halftone proofer as well.

Overall what I'm hearing is that I am pretty much doing the best I can with what I have. I'm going to look deeper into Esko. Perhaps I'm confused or mislead but from what I understand with Esko I can RIP one-ups and set the screening at the preflight level then proof that file and (hopefully) see moiré in my inkjet proof as well as on screen. This rasterized file can be stored/archived when approved. Then when needed be imposed and passed through to the CtP as binary data. Making the plating a rather "dumb" system as no actual RIPing occurs at that point. Webcenter also looks very promising for proofing and communication as we specialize in stupid-fast turn around of mostly promotional product cartons.

 

[gordo]

Yes FM screening is much improved but you wrote that: "One of my biggest problem customers is a multi-national pharmaceutical giant and they won't even let us use 175lpi screening. We have to plate their carton art at 150lpi." so I understood that FM was not workable for you.

What you want to be able to see on your 50" LCDs are the bitmaps that will be imaged onto the plate. If you view at 100% they will appear to be enlarged on the displays. Each dot will be about 1/8" in diameter. So your LCDs have more than enough resolution. You walk away from the monitor so that the dots appear smaller - and if there's a moiré you'll likely see it.

If that doesn't make sense then make a 1" square in PShop at 72 dpi. Then do a mode change to make it a bitmap at 2400 dpi using a round dot at 150 lpi and 2400 dpi. Then zoom in with PShop to the result until it's displayed at 100% and you'll see the big halftone dots.
That's what you want to see in your workflow and I'm sure that Esko would give you a temporary license so that you could try it to see if it works for you. If it does then buy it.

best, gordo

 

[chevalier]

That same 150lpi customer refuses to provide anything but DCS files as well. I won't let one very important but antiquated customer hold us back from meeting and preferably exceeding my other customers needs.

 

[gordo]

That same 150lpi customer refuses to provide anything but DCS files as well. I won't let one very important but antiquated customer hold us back from meeting and preferably exceeding my other customers needs.

HEY! Don't knock the DCS :-D

You can do some neat stuff with DCS that would be hard to do any other way. For example, you can prescreen one of the DCS files (in PShop or other application). When the DCS hits the RIP the prescreened channel will be untouched while the other three channels will be screened by the RIP.
Or you can screen each of the channels a different way.

I made great use of DCS when I was designing FM screens back in the day.

Lot's of fun LOL

best, gordo

 

[TheProcessIStheproduct]

Is it possible that what you are getting is possibly doubling or slurring on press? At one point you said a brown would moire on just a part of the form, if moire I would think it would be more consistent, but a press problem, say gripper problem before mag unit could cause multiple jobs with a brown build to have a moire type pattern?

 

[steppy]

Esko offers stochastic yellow in standard screening. Not as an option. Totally eliminated moiré in our process work

 

[gordo]

Esko offers stochastic yellow in standard screening. Not as an option. Totally eliminated moiré in our process work

If it's 4/C process and you are getting moiré from the Yellow printer then using FM/stochastic for the Yellow is a typical work around (typically 35 micron for Yellow with 175lpi for KCM). That being said, one should investigate why the Yellow moiré is visible since that may reveal some other problem on press. Usually it's a contaminated/dirty Yellow or the wrong hue of Yellow. In that case it would be better to fix the Yellow problem than mask it with an FM screen.

best, gordo

 

[chevalier]

I'm no longer at the operation that had that issue but as I said changing screening angles to classic angles from +7.5º angles fixed the problem. I setup parallel workflows so if they came across a repeat job that off with the classic angles they could easily revert to the +7.5º angles. Ink sequence had already been played with as yellow contamination, press slur, etc were all thought to have been the cause. We had plates flown in internationally, plates flown out internationally for testing, plate specialists, platesetter specials, prepress specialists, etc. in to try to resolve it. The screen angle change was an instant and permanent fix.

Side-note:
As my current workplace we use a fade-resistant process yellow and I learned the effects/limitations/issues of this with our first G7 calibration here. CMYK isn't my issue it is packaging products utilizing 6 - 8 inks with intersecting screens.

DCS had its day just like postscript, appletalk, win32 among other technologies. Personally I now think it needs to give up the ghost. The biggest problem with it is every vendor has a different standard for it. PDF/X4 is a life-saver and has eliminated countless problems for us.

 

[gordo]

I'm no longer at the operation that had that issue but as I said changing screening angles to classic angles from +7.5º angles fixed the problem. [SNIP] The screen angle change was an instant and permanent fix.

It may have fixed the symptom - which is great - but it could not fix the problem - or it was a different problem (e.g. single channel moiré). Moiré occurs as a result of the angular relationships between the screens. Since rotating all the angles by 7.5° does not change the angular relationships between the screens - it cannot fix the problem if the problem was moiré caused by the angular relationships between the screens.

Side-note:
As my current workplace we use a fade-resistant process yellow and I learned the effects/limitations/issues of this with our first G7 calibration here.

Yes, fade-resistant process yellow is like a dirty/contaminated yellow - it will increase the visibility of the existing yellow/magenta or yellow/cyan moiré.

DCS had its day just like postscript, appletalk, win32 among other technologies. Personally I now think it needs to give up the ghost. The biggest problem with it is every vendor has a different standard for it. PDF/X4 is a life-saver and has eliminated countless problems for us.

I agree - but, as I said, I did have a lot of creative fun experimenting with halftone screening using DCS back in the day that couldn't easily be done any other way. Nostalgia

best, gordo

 

[chevalier]

Don't angles slightly change gamut? My thought at the time was that the builds having issues were out-of-gamut/edge-of-gamut with one set of angles and not the other. When you looked at this thing under a loupe it looked pretty sound but as an optical illusion at normal magnification/distance there was an issue.

 

[gordo]

Don't angles slightly change gamut? My thought at the time was that the builds having issues were out-of-gamut/edge-of-gamut with one set of angles and not the other. When you looked at this thing under a loupe it looked pretty sound but as an optical illusion at normal magnification/distance there was an issue.

Yes, screen angles can change gamut very slightly (click here for an explanation with examples: Quality In Print: Why use halftone screen angles? ) but the effect is very very small with standard screen angles and would likely be within the normal noise of instrument error. It also would cause or eliminate moiré.

I strongly suspect that your old moiré problem was caused by single channel moiré. I.e. One (or more) of the screens already had a moiré pattern in it. This is caused by digital screens run at angles other than 0° and 45°. At those other angles, a halftone dot at a certain lpi may end up with a few extra pixels. If those extra pixels on each dot repeat they form a pattern. If that pattern has a frequency and angle that harmonically beats with the underlying screen then you have a moiré within one screen. To avoid this problem, some vendors utilize shifted angles of 7.5°. What this does is introduce "noise" around the edges of the dots. This break up the pattern created by those extra pixels and therefore eliminates the visibility of single channel moiré.

You wouldn't see single channel moiré under a loupe but you would see it at normal viewing distances.

A well designed AM/XM screen shouldn't have single channel moiré - but if it does then the 7.5° offset is a perfectly reasonable work around.

Another subtle cause of single channel moiré in CMYK work is "wire side" moiré. When paper is manufactured it has two sides - the felt side (top) and wire side (bottom). The wire side is a mesh screen - a grid. That mesh screen can cause a very subtle pattern in the paper. If the wire side pattern has a frequency that harmonically beats with one of the halftone screens then you'll see a moiré pattern in that screen. And because you can't really see the wire pattern in the paper - the moiré will look like single channel moiré since there's apparently no second screen to create the conflict.

Again, you wouldn't see single channel moiré under a loupe but you would see it at normal viewing distances.

Again, the 7.5° offset might alter the halftone screen angle/wire side pattern conflict enough to eliminate the problem.

best, gordo

 

[rich apollo]

Stochastic would fix what you're facing. The group that lost the account you referred to didn't lose it because of FM screening. They lost it due to poor implementation. FM screening is great, but is very different than AM screens. I can imagine several ways that the other printer might not have run FM screening optimally. It IS NOT a simple "drop in" solution.

The sales person is gonna' need to sell it. There is a lot of value in stochastic screening - for the printer and the client alike.

 

[gordo]

Stochastic would fix what you're facing. The group that lost the account you referred to didn't lose it because of FM screening. They lost it due to poor implementation. FM screening is great, but is very different than AM screens. I can imagine several ways that the other printer might not have run FM screening optimally. It IS NOT a simple "drop in" solution.

The sales person is gonna' need to sell it. There is a lot of value in stochastic screening - for the printer and the client alike.

Amen!

gordo

 

[meddington]

I am sometimes amazed at the amount of work separators and printers will put in time & time again to make complex 6-8 color jobs work with AM screens, then balk at the suggestion of implementing FM screening because of the effort to introduce it.