We want to print a halftone and are wondering the best approach. I know it will start with my file. Since I can destroy a file with the best of em please guide me to a path in photoshop or illustrator that can get the ball rolling. We also are not sure what resolution to drop down to or setting we may want to use. Our elements will be flowers and light detail objects. We plan to print on light card stock and text stocks. We want to make a copper die as we feel it will hold the best image but also wonder about photo polymer. We will take any advise or insults you have to offer.

Log in to reply   24 replies so far

Well, I think one piece of advice I can give is that since the cost of the polymer significantly undercuts the copper… And this may take a little experimentation to get ‘right’…. It may be better to start there.

Of course, this depends whether you have a base that works well for polymer, I am assuming you have a good base.

Things I have found helpful with halftones:

~55-67 LPI seems to work pretty well.
~”Stiffer” Ink seems to work best for these; combined with an makeready spot tissue under your top sheet of press packing which takes into account solids and deeper toned areas which may need a bit more pressure. This is especially true with platens.
I really, really like to use Carbon paper or graphite transfer paper to set this up;
My process-
Pack your platen out as if you were going to print- just add whatever packing you would normally add and account for at least two sheets of Tissue paper in your packing, even if it means dividing up a thicker piece of packing into three pieces of tissue. You’ll need to corner your packing so you can pull it out and put it back into the same place- this is critical to align the spot tissue you’ll do later in the same place.

So here’s the tricky part- Instead of inserting all three sheets of tissue, MIC the transfer paper and figure out how many sheets of tissue it takes to equal that. Leave the equal tissue out of the packing, and insert the transfer paper instead- put this sheet UNDER your top sheet, with the transfer side facing a piece of Tissue paper that is marked up or taped to another piece of packing. But set the ‘extra’ tissue aside because later after you remove the carbon transfer sheet, you’ll re-use it because you will need to account for the carbon transfer sheet being “left out”.

Pull an impression with no ink, just feed a sheet as normal.
The impression will transfer the carbon or transfer image to the sheet of tissue, which you will remove along with another sheet of packing. You use this image as a ‘guide’ for where to add your little cut out pieces of spot makeready tissue. Paste up your makeready to the spot sheet and then re-insert everything to the same orientation/corner.

~Inking up ‘slowly’ is critical; too much ink easily plugs the deeper tones up in short order. It’s minuscule, the differences between too much ink with a halftone and not enough; there’s a very sharp curve, so to speak, as compared to type or line images. Gain is a real big problem, so it’s something to be watched for.

~I really find it useful to ‘face’ any paper that is not pretty smooth. Basically run the sheets through the press with a blank plate that will ‘face’ the paper a little before printing. Make this plate a little larger than the halftone plates. Halftone for me is not about making a relief surface sculpt in on the paper - it is about transferring ink in a way that the image is clearest. So I tend to flatten out and smooth the paper before beginning to use ink on any rough papers.

Good luck!

^ this is utterly brilliant and has answered a question I didn’t even know I had. THANKS!!

What kind of press do you have?

Also it helps to have a good idea what max size you are hoping for, that can help the forum help you make sure your press is up to the task.

If your having plates made, often you plate maker can help you with getting the halftones right on the plate side.

Where are you located, i would be willing to loan you some half tones to play with if your close by, depending on what area your in you may find the same from others.


Hi Haven, thanks for the feedback. I had wondered about building up the packing for the image area and you have some great feed back. It’s kind of like building a counter for controlled pressure only under the image. We are running a redball windmill. We are looking to only run a 3 x 3 image area.

3x3 is fine on a windmill.

What HavenPress said on packing and ink.

It’s way easier to print halftones on coated stock, so try that first to make sure you have it down, then branch out on to other stocks.

We print halftones in the shop often, happy to help if you ever need any.


The lines per inch can be as fine as 95 if you are using a coated paper. The 65 line for coarse papes. Ink tack is also very important.You will need to experiment with that too. If too tacky it will pull on the form and slur will occur in the HT dots.

It is good to know. For some reason I thought the letterpress could not print on coated stock. This will be interesting to try.

Actually you can probably go for far finer screen resolutions than mentioned above. I believe ATF indicated that their Kelly presses could acheive or even exceed 150 lines per inch. Coated stock, excellent makeready, and consistent and exact ink control are key.

Here in the UK I have printed from Polymer at 120 lpi
on art and coated card stock recently. In the past we went up to 200 lpi on very good coated stock printing on Heidelberg cylinders.
If you are printing on a platen you will need to have the help of grippers or as we call them in the UK Friskets to help with the lifting of the sheet from the plate(sometimes)
With a cylinder press the problem does not arise on small plates. For make ready it might be worth looking at the Heidelberg platen manual for guidance ( can be downloaded from letterpress commons ). have included a picture from the 70’s showing halftones being pirinted a British Miehle. paper stock was a good quality matt coated art and the subject matter was antique furniture etc.

image: proof readers020.jpg

proof readers020.jpg

This is nuts. Printers used to run 150-200 line screen halftones on letterpress routinely, mostly on gloss-coated and dull-coated paper but also on supercalendered paper. The problem is more that the imaging process has deteriorated — a pre-screened image can not be reproduced as accurately by a platemaker as can be done using a quality halftone screen in a graphic arts camera. For the letterpress revival someone should revive one of the old Robertson cameras and the film processing gear to go with it, and obtain a selection of good cross-line screens, and learn to make quality halftones again!


I’ve got to agree with Ad Lib Bob…..

When I was first learning to print, the shop I worked in routinely printed 165 line halftones on copper plates and the results were superb. Back then, it was normal to use a graphic camera for all cuts and if a shop didn’t have a camera, there was always a shop in town that would do the work for them.

Producing a good halftone plate is not easy, and it requires a bit of skill but it is well worth the effort.

Unfortunately, many printers nowadays (even the hobby printers who don’t have the same economic constraints) have traded technology for skill and prefer to use electronic imaging methods rather than other possibilities. The end results are sometimes cheaper and faster, but they are not quite as high resolution and seem to lack the same crispness on the printed page. The truth is you just can’t beat a good copper plate when it comes to crispness.

I still use old-school film negatives for some of my work, and posted about it here on Briar Press several years ago….. and the response was lack-luster, at best. Apparently, nobody was interested in learning how to do it so I stopped writing about it.

aka Winking Cat Press

by the way: HavenP is correct about inking. A little difference on the press makes a big difference on the page.

With respect, ??? Is it just possible that instead of talking.>LINES PER INCH< the terminology should be D.P.I. = DOTS PER INCH.

Lines per inch, always used to be applied to Television, as in
- - 405 Lines, . . for early Black/White,!
- - 625 Lines, . . for early Colour,!
-1,025 Lines, . . for High definition,!

Even modern *Pixelation* appears to speak in terms of D.P.I. rather than *Lines per Inch* for resolution.?

Pending. ?

Mick- “lines per inch” is the terminology use for older halftone plates. It comes from the gradations of the screen used to produce the image. It basically corresponds to the term “dots per inch” in modern use, but is only counted linearly and should not be confused with “dots per square inch”.

Winking Cat, Thank You, perhaps as it was a long time ago, (during apprenticeship days `54 - `60) do not remember too clearly, but in the Comp Room do seem to remember, the Journeymen comps, referring to D.P.I. on Halftone, (stock) plates when used as Background Tints etc. drawn from *In House* store and cut/trimmed to size, per job.

Thank You, again, possibly, variation(s) in terminology, different sides of the Atlantic, maybe. - Mick.

Mick… you are correct in your remembering. Background tones used in copy-prep were sold as “Dots Per Inch” back then.

Western411, you will find that the dot gain on letterpresses is higher than on offset presses and your halftones are likely to come out too dark without some modification. Depending how you are getting your plates, there are a couple of routes you can choose. In either case, build a step scale file in Photoshop, or download one. This is a set of at least nine rectangles, each with a different screen percentage from 10% black to 90% black. Be sure to select CMYK mode, and use ONLY black. Now, have a plate made and print it on press, getting the ink up to your desired density. I like 1.80 for the solid black.

Option 1) If you have a reflection densitometer, measure the dot area of each rectangle, then go back into your Photoshop file and drag the curve until you’ve lightened each patch by the amount needed to get the desired dot area, if you subtracted the gain. For example if your 10% patch actually printed out at 15%, you’d drag until the patch in Photoshop showed 5%. This procedure is a bit tedious and will take a couple of tries, but you will get a curve that will yield a plate that pins much closer to your intended values. Save the curve, then apply it to a photo, make a plate with both the raw file and with the curve-applied file and print it. You’ll clearly see the difference between the two halftones.
Option 2) Send your printed sheet and your step-scale file to your platemaker and ask them to adjust their output to yield halftones closer to your desired end.

I print Halftones as Fm Screened in Letterpress from Polymer.


Type nut, are you running film? If so, what filmsetter and RIP are you using?

To David R Miller

Panther Capstan with Panther Rip 10

Linotype Hercules Pro M with Xitron Rip 8.08

My standard Plate is a Flint Digital Security Steelbacked

I’m limited to a Plate size of 24 x 30 inch.

Before this starts another Discussion:

Panther Capstan is 13.3wide by x

Linotype is Drum 24 x30 inch

Historically, PROS who have a lot of experience with halftone printing were running high LPI 150+ Linescreens on coated stocks and ultra-calendared sheets. Yes.

But you all, come on.
The Original Poster, Western411, asked about their FIRST halftone. Not “What was possible with the finest robertson camera”. So, Bob/Adlib, with respect- I do not think giving solid advice about a starting point is “Nuts”. I think it is actually very “Sane”.

We are not having a conversation about the limits of the industry, we are having a conversation about a printer who has a Heidelberg windmill who is going to use it for the purposes of printing their first halftone image.

This is not to discourage attempts at difficult things, but frankly it is entirely possible that the stock they might wish to run would be un-suitable to the task of higher LPI halftones. I am talking about what is able to be accomplished with just about any stock, universally- from coarse/rough cotton stocks to even the very fine stocks, and coated stocks. Not what you can do with the slickest setup and plates. Additionally, coated stocks are fine to use and people print with them with LP all the time- but if you let the pile get too large, you’re going to run into a problem with offset from sheet to sheet. Frankly, if your windmill is equipped with a powder sprayer, you’re gonna wanna turn that on when you run coated stock to minimize this. Otherwise that shit will- as another printer once put it to me- “Brick in the chute”.

My advice stands, and another thing I will ad is that I suggest you use the hardest packing you can find behind the tissue- do not use anything soft; use red pressboard and be very meticulous with your spot underlay beneath the top sheet, as described.

Thank you all for the input. My background goes back to creating reflective flatbed negs in a dark room. We ran a full 7 color process job on a Kord back in the 80’s, more mask flats than I care to remember, so crazy is my middle name by todays modern standards. We have also done modern full color 200 line using platesetters with ink densitometer support. I have always look at DPI as photo fidelity callout but LPI as an imagesetter or plating callout. We have done imagesetter rip setup so color curve control and dot gain are something we actually understand. Our thought process was to print a 75-100DPI halftone art at the prevailing screen bottom film rate or DPI, generally these days 133-150 Line for an imagesetter. Our thought was to each dot as large as was reasonable in order to get a usable plate and better control You all had great feed back. We had wondered about color curves and dot gain and did not know the best way to approach it. I could make sense of printing with little or no hit, but could not figure out what would happen as the strength of impression was increased. That is what got us wondering about a lower DPI photo being able to take a stronger hit without everything just merging into one big blob.