changes in line screen

Got a problem that I can’t understand…

I have a DI press. When it was setup, we linearized it and then selected the dot gain curve we wanted. So basically we made sure 50% was 50%, then asked the rip to make 50% become 70% for example. All this was done at 150 lpi and we made profiles.

We want to move to 175 lpi so I linearized the press at 175 lpi and applied the same dot gain curve as we used for 150 lpi.

The natural dot gain of the press showed me that 175 gained more than 150 if I remember correctly. This told me that the same image printed at 175 would appear darker compared to 150.

My proofs were a pretty good match to the press at 150 lpi.

Now, when using the profiles I made at 150 lpi, with the press running at 175 lpi, my images appear lighter compared to the proof. I don’t understand this. I figured that since I’m asking for (and getting to my knowledge) the same dot gain at both line screens, the profile should still be ok. At worst, I expected the output to be a little darker but it appears to be lighter. (I’m running to the same densities)

My proofs are basically useless for 175 lpi now which is obviously a big problem.

Is my logic wrong? I don’t understand what’s going on so I have no idea how to fix this problem.

Any help is appreciated.


I forgot to mention that at 150, dot centered rosettes are being used. At 175, clear centers are being used.

I didn’t expect that one dot to make much of a difference but maybe I’m wrong?

That very well could be the difference…it’s one dot per rossette…times that by all those rossettes and that may make a big difference.

That being said, are you measuring the dot gain of your DI device at 175lpi? You could also measure simply the density of your tone scales and either attempt to match it to the proof, or the proof to it by plotting print density curves. Its not all that different from measuring dot gain curves, but without the gotchas you can get with measuring only dotgain…such as a waterless process where density and dot gain don’t quite correltate in the manner that they do in a traditional offset process.

How is that, Michael? I’ve experienced this phenomenon at our facility but have never been able to explain its cause. I’m waiting for PIA/GATF to publish a paper on Waterless now that their DI is installed, but no luck as of yet.

Your comment on the clear centered rosettes is right on. Our “square spot” proofer was set up and installed with clear center rosettes and bingo, nothing matched. Flipping the switch, as it were, to dot centered rosettes made life grand again!

Well, please don’t infer that I’m an expert on the subject, but it has to do with plate structure and ink emulsification. Ink sits on top of a traditional offset plate, but a waterless plate has “sidewall suport” that reduces lateral dot gain upon impression. And waterless inks are usually higher tack, and therefore less suseptable to gain while traditional inks undergo more emulsification which makes them more readily absorbed by the paper.

there’s a good article here: … i_20633661[/url]

Ah, I see the reference you’re making.

What we experience is that we cannot affect dot gain at all (maybe 2% or so) by increasing SID, until a point where we apparently overcome that “sidewall” support and the dots just explode. At this point we add 8 to 12%! Backing off and cleaning up yields the exact same gain as before we reached that threshold.

Thank you for that article… it pointed me back to the WPA, which I had apparently forgotten! If you do not have the “Complete Guide to Waterless Printing” book offered by the WPA, and you’re interested in furthering your knowledge, it’s a terrific reference. I had a chance to review it at the PIA/GATF Process Controls Boot Camp recently.


Are rosettes created for everything except solids? The color I had in mind when I posted this was something like orange-red. Under the loupe, I didn’t see any visible rosettes like I can in other colors/areas. Are rosettes always there? Does the dot always come into play except on solids?

Not really. I have no reason to believe that we are getting very different dot gains than what we asked for though but I could be wrong. The press sheet is so small there usually isn’t room to include dot gain patches but I’ll try and get something on when there’s space.

Can you explain this a little more? I thought you can’t really compare density and dot gain between press and proofer because the technologies and inks are quite different.



Do you have a link for this book? Price? More info? Ordering? Sounds like something I might like to look at.

Does it explain that thing with dot gain and density behaving differently in waterless than traditional offset that Mike mentioned earlier?

Forgive me for butting in…

Our product has no opportunity for a color bar of any shape or size; in researching this problem we found that a 50, 39, 39 patch as small as our instrument would read gives us enough information… we hide it in our product logo under a patch of 100% black.

Knowing the values as 50, 39, 39 we can infer gain information and also check gray balance by ensuring that the densities are within .01 of eachother. In fact, a solid patch of each primary and three gray patches (25%, 50% and 75%) should be everything you need, assuming you aren’t using an auto-scanning device.

The book is at:

If you’re not a member of the WPA it’s $89. It’s a 200 page book so at least it’s not priced like an ISO doc!

The article Mike linked me to explains it as well as anything I’ve read on the subject. It’s on the second page about halfway down.

The rosette pattern will show up predominantly at around 50%. I really can’t say what a clear or dot centered rosette will look like at different percentages, but I would assume it plays a role throughout the tone scale.

It depends on the proofer. For a Kodak Approval…no problem. Match the tone scales density for density and you proof will likely be fairly close to your target. For an inkjet, this is less likely to result in a good proof. What type of proof are you using?

You don’t remember Mike? :slight_smile:

I’ve got an HP 5000 and I’m using a combination of profiles and tweaks in the rip (same applied to every image) to get the color right.

I don’t remember jack lately… :wink: Yeah, in that case print density curves aren’t the way to go. I’d try imaging with dot-centered rossettes and see what that gets you. Or created a profile from your current 175lpi set up and go from there.