LPI, mesh count, and degrees.

[NeroInferno]

Hello guys,
I'd like to print some gradients on fabric.

After days of search I've understood how to print gradients (at least on film)...but I need the help of any experienced guy to understand better.

I created some gradients with spot colors, then in the Illustrator printing dialog I selected just the gradient channels. Here I should set the angle and the lpi of the channels but I can't understand what to insert. However I printed the output to the Adobe Postscript Printer, having a file with .ps extension. (I've installed the Adobe Postscript Printer previously)

I installed Ghostscript and then Ghostview (you need Ghostview to open and use Ghostscript).

With Ghostview I opened the file, made by 2 pages, because 2 were the spot colors of the gradient. I exported one page in .eps, then the second page too, so I was able to mount in Illustrator all the plates for the film.

Now I'd like to know if there's any LPI / mesh count table to follow for a correct print and to avoid the moire effect.

I'd like to know also something about the degrees of the plates: how can I understand the right angle?

Thank you all
Fabio

[paul204]

Moire occurs when you are printing overlapping halftone patterns with semi-transparent ink, so usually only when printing four colour process. Don't worry about moire when you are printing spot colour gradients with opaque inks.

In Illustrator or Photoshop, you will have an image which contains objects that have seamless variations in value (or lightness/darkness, for example, text that is in a gradient from grey to black.) Your goal is to print this with a single colour of ink, so you want to transform it into a halftone pattern that tricks viewers' eyes into thinking there are many different shades of the same colours from a reasonable distance.

Manually printing, the highest LPI screen you can run on standard t-shirt fabric is generally 55. Autos run upto 65. After those numbers the halftone screens starts to be absorbed into the weave of the fabric and results are not as good. To hold a reliable stencil, a rule of thumb is to use a mesh count 5 times higher than your LPI--most textile printers use 305 mesh for 55 LPI halftone patterns. These numbers are in TPI/LPI and you are in Europe so you may have to convert to threads per cm.

The computer stuff:

Back to your image with a gradient. Your output device is what controls how gradients are turned into halftone patterns. If you print your image on a standard laser printer you can see that it automatically turns it into a halftone screen. It always prints at the same LPI, you have no control over any of this most of the time. Most upstart printers (like you and I) are using inkjet printers to print film. I am just going to assume you are.

Inkjet printers do not come equipped with a processor which uses the industry standard protocol/language/whatever to interpret gradients into files ready for output to film. This is why you need Ghostscript/view. It acts as an intermediary between Illustrator and your printer. With Ghostscript, you can create this post-script file and give the necessary information to your printer. You will need to do some experimentation depending on the image you are printing, but for spot colours most t-shirt printers will print 55 lpi elliptical dots at 25 degrees. Lots of reading about this on tshirtforums. From Ghostscript/view, you don't send the file back to Illustrator--you send it to your printer, looking exactly how you want it to on your film and therefore your screen. This will not be how the image looks on your shirt though--you'll quickly discover you have to accomodate for dot gain. This is the other important part of a post-script file--you can create transfer values in a halftone pattern, for example, if you want a 95% white area on your shirt, you will need to dial that back to 75%-90% depending on a variety of factors, otherwise when you print it it will just fill in to 100% white as the space between dots is so small.

Hope that helps as a start. There is much to read over at tshirtforums on this subject. Printers will never give away their four colour process angles. Spot angles don't matter so much. Most things like transfer/dot gain and screen angle as well as LPI and mesh count just depend on your technique, screen quality, press etc. When printing halftones lots you just have to run infinite amounts of tests. Try to print out a film that has 20 areas on it--5-95% black in 5% increments. Print it on a bunch of shirts and compare it to your film and you will see where you need to use your transfer function to dial dot size up or down. This function is built into Photoshop's print dialog, but it isn't in Illustrator--you need to go into the ps file and modify their numbers.

[NeroInferno]

This will not be how the image looks on your shirt though--you'll quickly discover you have to accomodate for dot gain. This is the other important part of a post-script file--you can create transfer values in a halftone pattern, for example, if you want a 95% white area on your shirt, you will need to dial that back to 75%-90% depending on a variety of factors, otherwise when you print it it will just fill in to 100% white as the space between dots is so small.

Hope that helps as a start. There is much to read over at tshirtforums on this subject. Printers will never give away their four colour process angles. Spot angles don't matter so much. Most things like transfer/dot gain and screen angle as well as LPI and mesh count just depend on your technique, screen quality, press etc. When printing halftones lots you just have to run infinite amounts of tests. Try to print out a film that has 20 areas on it--5-95% black in 5% increments. Print it on a bunch of shirts and compare it to your film and you will see where you need to use your transfer function to dial dot size up or down. This function is built into Photoshop's print dialog, but it isn't in Illustrator--you need to go into the ps file and modify their numbers.

Hello Paul,
thanks for the detailed reply.

Have you the way to measure the dot gain on your prints? At least for the things I've read here on Gigposters is better to do not compensate the dot gain if you haven't the tools to measure it.

However I'd like to make some tests with the 5%-95% increments and see what happens.

Thanks!
Fabio

[bdix]

You can totally get moiré just printing one halftone on a shirt, regardless if the ink is opaque or transparent. Moiré is just two "grid" systems overlapping each other and causing interference at certain angles. Hell it doesn't even need to be a halftone; it could be a column of lines rotated against another column of lines. Want to see how it works? Print out a halftone pattern onto film, then lay it on an uncoated screen. Rotate the film clockwise or counter-clockwise. You'll see that at certain angles the halftones and the mesh of the screen interfere with each other a cause moiré.

Screen mesh is based on a grid where one set of threads is angled at 0˚, and another set at 90˚. Therefore you should avoid setting your halftone angles to anything that may interfere with that pattern. Basically you're avoiding 0˚, 15˚, 30˚, 45˚, 60˚, etc. So what do you set your halftone angles at? The rule of thumb is to split the 15˚ difference in half, which would be 7.5˚. Since you can't enter half a degree, you round down to 7˚. So instead of a 45˚ angle you add 7˚, making it 52˚. Instead of 15˚, you add 7˚ to get 22˚, etc, etc.

These aren't trade secrets anymore. Haven't been for years. Been discussed here on GP numerous times.

As far as dot gain, you learn to adjust as you learn your capabilities as a printer. I've done seps for shops that had as little as 10% dot gain, to some shops that had as bad as 40%. Keep samples of all the halftone work you print and you'll start to see where you need to adjust your artwork.

Something that will help you in the short-term is to print your halftones on a separate screen than your solid spot colors (even if they're the same color). That way you can print the solid black text with a heavier hand (one screen), and print the black gradient with a softer hand (second screen).

[squeegeethree]

Moires can happen anytime two or more patterns intersect. I'd go further to add that they always happen but are more noticeable with certain combinations of patterns. One can even get a moire from two flat colors with no dots printed on top of each other if the warp/weft is not the same between the colors. The screen itself is a pattern after all. Bigger dots on finer mesh = less moires (I usually divide my mesh by 4 to determine the lpi). For whatever reason 22 degrees is the easiest for me to shoot without getting a dot on screen moire. So if it's one color, I'll output my dots that way. When doing 4 colors I use increments of 22 degrees, Cyan 22, Mag 44, Yellow 66, Black 88. Also, if I'm worried about a screen moire I will use a lighttable to line up the film and screen before coating.

[RichieGoodtimes]

Those 22, 44... angles seem practically designed to moire, but whatever works. I think we've had this discussion before. Carry on.

[mikeage]

My head hurts now.

[squeegeethree]

Those 22, 44... angles seem practically designed to moire, but whatever works. I think we've had this discussion before. Carry on.

They never moire with each other. Sometimes they moire with the screen but rarely. We just finished 20 CMYK images, 50 lpi, 230 mesh and never had a moire.
We did have this conversation, I mark it as the beginning of our glorious friendship.

[RichieGoodtimes]

I'd almost count on moire at some LPI/mesh count combos with 44 and 88. They're practically 45 and 90.

[squeegeethree]

You are right. Finding the right LPI is the key. The most I've done was 100 lpi and you never had to worry about a screen moire but the dot on dot moire was fierce.
Personally, I'm more concerned with dot on dot moire as I can change the angle the film in relation to the screen before coating. A t-shirt printer wouldn't have this luxury as they can move there image around so wantonly.