Halftone file for DIY films/screen testing and some pointers on calibration

[mk162]

linearization is for film output, dot gain compensation is for printed garment.  My old RIP (that Dan now has) had both.  That thing was awesome.

I don't know why you would need both since I would think that if you tested your dots at 5% intervals 5%-95% that you could test the printed shirt and work from that.

[Inkworks]

Gotcha, I knew I must have been missing something.

[JBLUE]

Isn't the idea to linearize to the printed ink deposit, not the film? Or is that next to impossible to do on a shirt?

You need to do both. Your inkjet printer is going to have dot gain. There is no way around it. The only way to compensate for it is to linearize the printer. This is the first place to fight dot gain accurately. Correcting it before it even gets to the screen is important.Trying to compensate for it in the art or on press is just going to affect the quality of your print.

If you are getting 15% or more gain on the film add that to the percentage of what's going to happen on press and you end up with a mess. You can only do so much in the art before you start making a negative impact on press.

Most would be surprised at how far off their printers are once they linearize.  I was. That 5% dot that you thought you were holding perfectly now is hard to resolve because it was really a 18% dot the whole time.

[Screened Gear]

Ok I said I would post my FilmMaker dots. These are the same printer (still have not fixed the out of ajustment nozzels need to get photo paper or somthing better then I have for that.)

These are 55 LPI 50% dots. The dots themselves are super dark. If you hold the film up to a light and put your finger between the light and the ink deposit you can verily see your finger (in a solid black area of the file). The ghost rip ink density was lets say 50 to 60 percent black when looking at the light. The Ink with the adjustments in MF is about 90 to 95 percent.  What does that mean? The dots have a really good chance of not being over exposed by light penetrating the ink. This also means I can burn the screens longer making sure the emultion is fully exposed with out the worry of overexposing my design. The darker film also makes washout really easy.

[ABuffington]

Just my 2 cents:

If you want incredible dots get a real film image setter.  Those dots are as crisp as you could ever want them.  They are cost prohibitive in both equipment and film costs, but ask some of your top end printers and they will go with image setters, not ink jet for the ultimate films.  However inkjets are the most common film making method, nothing wrong with them, just need to know how to tweak the output.

The stray dots seen in the ink jet halftones are the pico liter dots, typically 7.5pl to 12 pl that the printer uses to build up the halftones..  These dots are too transparent and too small to image on emulsion.  They will either be undercut, or too transparent to avoid burn through.

For textile printing halftone printing is a balance. For me I get diminishing returns as I increase lines per inch.  At 85 line the ink wants to jump across the negative dot of 70-95% and the tonals disappear. Higher mesh counts an help control this but ink opacity suffers.  Increased squeegee pressure to get the halftones to be more opaque just kills halftones above 80%-90%.  The balancing act: Any halftone print should have the following attributes: Good print opacity, reproduction of as many tonals as possible between 5 and 95%, the least amount of squeegee pressure possible to yield good 100% solids and tonals within the base.  With S mesh there is no need for print/flash/print
to get a base.  I do like a dummy screen after the flash to press the tonals flat to accept other halftones printing on top.  Under a microscope a plastisol base plate halftone is a pointed peak of ink, not a nice flat dot to print on.  Flattening them with a dummy screen after the flash means you can get away with less squeegee pressure on the colored halftones.  Discharge White base avoids this altogether, but does suffer dot gain and dry in of finer dots.

For me the optimum halftone baseplate depends on other factors.  a 14 singles low end T is better off with a coarser halftone.  A 20 singles smooth faced t can handle a 65 line. I like 55 or 65 through a 225S mesh baseplate on smoother knits, or a 45 or 55 line through 150S or 180S on coarse weave or darks.

I also like a linearized output.  No tones below 4%.  You can't see them anyway in most prints!   And it is better to print a base plate with minimal squeegee pressure to control dot gain on press.  For me that is a 150S, 180S or 225S.  Tones below 4% do get blocked by mesh threads, going up on mesh count to print dots you can hardly see kills opacity and causes the 85-95% to print solid.  Rather than trying to image a 2-4% dot, try to print a 90% dot.  This is the tonal area that begins solid art to tonal, a crucial tonal transition that looks better when smooth and not an abrupt transition and where dot gain kills most tonals.   

Attempting to hold less than 4% is not easy.  Threads block some of the halftones, an inkjet 2% is a really ugly dot with the pico liter dots barely forming what we would call a halftone.  Plus the shirt fabric as mentioned often can't image it.  Consider that UV poster printers often eliminate these percentages due to the problems in moire they generate.  They print with 380 and 420 mesh, which are hardly t-shirt friendly base plate meshes, and are really, really expensive.

The key test:  Output 45, 55, 65 line 5-95% tonal steps using the correct angles, (see below) .  Image on your choice of baseplate screen(s).  Set up and print on your typical shirt brand or multiple shirt brands.

Analyze all prints for best opaque tonal output capturing as much of the 5-95% as possible.  From here it can be fine tuned for linearization of the film and print.  For film you need a transmissive densitometer that can read halftones.  x-rite eye one is good for this.  For the print, reflective densitometers can get real pricey.  Comparing the 5-95% film to your print and adjusting squeegee durometer, angle, speed, ink viscosity, ink type can save thousands.  That's the beauty of this industry, sometimes all it takes is just making it look good!  Compare the different line counts visually, use what works best for you.

Here is an article on moire and how to avoid it.  Also pages 4 and 5 have a chart with typical Murakami mesh counts and the angles to use.  Choose from the middle of the green blocks of OK angles.  This will allow your mesh capture to be off be a few degrees and not matter.  For any halftone angle to work the mesh must be straight and square to the frame.  Or you can try biased stretch screens stretched to 22.5 degrees or 61 degrees, two common non moire angles.

http://murakamiscreen.com/wp-content/uploads/2012/07/HalftoneAnglestopreventmoire.pdf

Al

[GaryG]

Excellent!
Wow, only 28 posts and a wealth of information!!
Keep it coming. 

Yes, I remember that chart with 100 different angles- someone was busy.

[Dottonedan]

Excellent information.  Very good food for thought and should be book marked. I think this is great info for 100% of the shops. Highly professional, well thought out information.

At the same time, (while not taking one thing away from your contribution), there are those that make up maybe 5% of the industry that do use the 380 and 420 and higher mesh, that do shoot for the 1-2-3% in a 65 line screen and these shops typically have and use 10-14 colors excluding flashes and cool downs. A few shops have more colors than that available if needed. These higher color auto's make use of extra "shades" of color often used to beef up the print saturation. With so many colors going down for sim process, what is a thin coverage is often intended and built into the separation process. For the average good shop, I often provide seps containing 4% dots in a given area x 3,4 or 5 colors to create a shade.

Lastly, I do believe that those small dots (while confined to a pattern) in traditional halftone can produce hit or miss areas, when using the typical 305 on a 65 line screen, but out of 10-14 colors, much of that is hidden anyways. If using the higher mesh (on a 65 line screen), you can hold more of the smaller dots.  So the key for holding those smaller dots...is using that finer mesh higher than the tried and true 300 mesh range.

In addition, the (as was mentioned earlier in another thread), DTG spits out (like you have said), 7-10-15 Picoliters of dot sprays. (THAT) can be seen on the shirts and when combined x 7 other colors/shades, produce a full fledged detailed vibrant print.  So, (these small dots) can be seen, can work and are worth trying to achieve. We will not get (that small of a dot) as seen in a 7 picoliter spray, held in even the finest mesh, but if you do use a mesh that works correctly (with the right sized small dot) they can be held in print. Of course, this is not for everyone, but it can be done. There are customer target audiences who do desire this level and there are people putting it out there. Again, I agree.  95% of the shops don't need to do this or even try, but I do believe that they should be aware of what is possible and what some shops can do.

That said, I do know that, the contribution you provided is indeed true, accurate and 100% right for most all shops. It's invaluable and a gold mine of info.

Thanks for chiming in on the conversations and contributing your obvious vast experience. I know I can sound like I disagree and I really don't. I simply think there is more than that.  I respect you and the info and would encourage you to continue to let this good food (info) flow as I know all of us do.

P.S.  I am aware that I could be wrong on some things and welcome the opportunity to be proven wrong. After all, it's the "correct" info that is the goal. Not the goal to be right.

Thanks
Dan

[ABuffington]

Hello Everyone,

Yes Dan you bring up a good point about base plates.  They don't have to be white, or opaque.  Sometimes it can take 2 baseplates of differing colors to pull off a design.  Subtle tone on tone printing works better without a super bright white base plate.  We had a client years back that was into vintage rock n roll t-shirts.  They wanted to mimic old worn t's with missing ink and faded colors.  We used a clear base plate for their work.  It gave a vintage soft look to the print making the ink slightly transparent and 'worn'.  Add some stretch additive to the clear base for super soft prints.

On the subject of dots we haven't mentioned stochastic or index printing.  The one area I like to use these dots instead of halftones (or you can combine with halftones) is in a design like a band where the size of the faces is small, or where I have a small area in the print like chrome details of a guitar that are quite small.

A fast method for separation is to take a duplicate file>put art on black background>greyscale>level adjust for image punch>Invert>output in a 55 line halftone.  Then for the color plates take the original art>Mode>Index>choose 2-4 colors more than the press can hold since index will choose the shirt color, clear, and duplicate some colors. With both the original master file open and the index file open you can double click on a color in the color table for the index colors to open up the color picker.  Then take the eyedropper over to the original color file and pick out the primaries, secondaries and get the color table for index to match closer to the original.  I never understood why index mode shifts the color table to darker duller colors.  To be able to know how to combine dots to form secondary colors is not too difficult, but to combine multiple colors is a skill, even harder IMO for halftones, takes lots of practice.  We would clean up stray dots that didn't belong.  Like a erasing blue index dot lightly scattered across a yellow gold to clean up the image.

Separation skills are really an art unto themselves.  It is a never ending quest.  Channels, Masks, offer so many options in combining colors that it is a wonder the job is ever done on time or on budget!  There are some clunky auto sep programs available that can make ok seps, and they do a nice job, but to really capture details and smooth color transitions and predictable prints takes skills. 

Screens: 150S, 180S, or 225S for the baseplate - 350-380T for the over prints.  Mesh callouts for the index method above.

Al

[Dottonedan]

These types of discussions are the brain food for people in this industry.
I love this discussion.  I have used index and stochastic on this level and overall, I prefer the stochastic for the blend ability compared to the side by side dots of index printing. Also, as you've mentioned, multiple colors work best.

I'd love to hear more later.

[dahmitdesigns]

We run our bases on 173.

[Dottonedan]

That should be printed at about 35lpi then.   You can go a little higher, such as 45, but don't expect to hold saoft fades out to the shirt color with any smoothness. It will drop off. The mesh threads are actually blocking the smaller dots at that point.