Author Topic: Checkered patern?  (Read 5460 times)

Offline tonypep

  • Ludicrous Speed Member
  • *******
  • Posts: 5611
Re: Checkered patern?
« Reply #15 on: June 06, 2014, 01:26:42 PM »
You are experiencing both Moire in your screen and heavy dot gain from printing your black.

What LPI are you outputting at?

Follow the above directions for LPI/Screen mesh/angle.  55 lpi should expose just fine on a 305 mesh.

Colin riddle me this, why does this seem to happen mostly with black half toning? Something to do with the ink and how well it is round maybe?

Its the particle shape. If I remember right black is conical


Offline inkman996

  • !!!
  • Gonzo Member
  • ******
  • Posts: 3760
Re: Checkered patern?
« Reply #16 on: June 06, 2014, 01:31:52 PM »
You are experiencing both Moire in your screen and heavy dot gain from printing your black.

What LPI are you outputting at?

Follow the above directions for LPI/Screen mesh/angle.  55 lpi should expose just fine on a 305 mesh.

Colin riddle me this, why does this seem to happen mostly with black half toning? Something to do with the ink and how well it is round maybe?

Its the particle shape. If I remember right black is conical

Don't know about that but I do know black has always seemed like a dirtier ink. Ever notice when you let the screen run really low on black it starts to look crummy what little is getting flooded? I always like printing all blacks with Process black but it is a bit on the expensive side.
"No man is an island"

Offline blue moon

  • Administrator
  • Ludicrous Speed Member
  • *****
  • Posts: 6355
Re: Checkered patern?
« Reply #17 on: June 06, 2014, 02:48:26 PM »
You are experiencing both Moire in your screen and heavy dot gain from printing your black.

What LPI are you outputting at?

Follow the above directions for LPI/Screen mesh/angle.  55 lpi should expose just fine on a 305 mesh.

Colin riddle me this, why does this seem to happen mostly with black half toning? Something to do with the ink and how well it is round maybe?

Its the particle shape. If I remember right black is conical

Don't know about that but I do know black has always seemed like a dirtier ink. Ever notice when you let the screen run really low on black it starts to look crummy what little is getting flooded? I always like printing all blacks with Process black but it is a bit on the expensive side.

not sure if it's related, but black has lower viscosity than the other inks so it's easier for the shirt fibers to make it through the mesh and get sheared by the squeegee. So, easier to contaminate . . .

pierre
Yes, we've won our share of awards, and yes, I've tested stuff and read the scientific papers, but ultimately take everything I say with more than just a grain of salt! So if you are looking for trouble, just do as I say or even better, do something I said years ago!

Offline Colin

  • !!!
  • Gonzo Member
  • ******
  • Posts: 1610
  • Ink and Chemical Product Manager
Re: Checkered patern?
« Reply #18 on: June 06, 2014, 06:40:25 PM »
It doesn't actually happen more with black halftones.... you just SEE it more.

It's about visual contrast.  We - as Printers - also tend to put more pressure and try to lay down more ink when printing black so the dark areas have smaller amounts of fibrelation... Well that also causes dot gain :)  Right now, you are compounding dot gain with screen moire so it is all VERY obvious.

If you were to print your black ink with the same pressure and saturation as the rest of your colors, typically you would say your black ink looks faded/washed out.  There are ways around this....  But that goes into another level of color separations and are better left to another thread.

This is also why you have more VISUAL dot gain with darker colors.  You have approximately the same dot gain across all colors.  But you rarely SEE it with the lighter colors.

Oh, process black is just a low pigmentation black ink :)  They also use a different base.  You can make your own Process Black with your ink manufacturers process base and adding small percentages of black pigment... Maybe 3-5% depending.
Been in the industry since 1996.  5+ years with QCM Inks.  Been a part of shops of all sizes and abilities both as a printer and as an Artist/separator.  I am now the Ink and Chemical Product Manager at Ryonet.

Offline Dottonedan

  • Administrator
  • Ludicrous Speed Member
  • *****
  • Posts: 5875
  • Email me at art@designsbydottone.com
Re: Checkered patern?
« Reply #19 on: June 08, 2014, 03:27:55 PM »
It doesn't actually happen more with black halftones.... you just SEE it more.

It's about visual contrast.  We - as Printers - also tend to put more pressure and try to lay down more ink when printing black so the dark areas have smaller amounts of fibrelation... Well that also causes dot gain :)  Right now, you are compounding dot gain with screen moire so it is all VERY obvious.

If you were to print your black ink with the same pressure and saturation as the rest of your colors, typically you would say your black ink looks faded/washed out.  There are ways around this....  But that goes into another level of color separations and are better left to another thread.

This is also why you have more VISUAL dot gain with darker colors.  You have approximately the same dot gain across all colors.  But you rarely SEE it with the lighter colors.

Oh, process black is just a low pigmentation black ink :)  They also use a different base.  You can make your own Process Black with your ink manufacturers process base and adding small percentages of black pigment... Maybe 3-5% depending.




All very accurate information.  I agree 100%.  To add to that, Cyan ink particles are also oblong shaped (pill shaped) and can tend to clog up in the screen faster due to the ink particles clotting (building bigger clumps) than the other inks. This is why in some CMYK jobs, your cyan screen needs wiped out more in the highlight areas.


As for the term moire, this (what we see in the images) is not exactly "moire", but rather (mesh interference). I'll explain the different later for those not familiar with the differences.
Artist & Sim Process separator, Co owner of The Shirt Board, Past M&R Digital tech installer for I-Image machines. Over 28 yrs in the apparel industry. Apparel sales, http://www.designsbydottone.com  e-mail art@designsbydottone.com 615-821-7850

Offline Dottonedan

  • Administrator
  • Ludicrous Speed Member
  • *****
  • Posts: 5875
  • Email me at art@designsbydottone.com
Re: Checkered patern?
« Reply #20 on: June 08, 2014, 09:06:48 PM »
Hello All,


I've done the seps on this job and I wanted to take this particular print result to point out some very common pit falls we all face in printing halftones. Every forum, multiple times a year, someone will post up the same issues. I'm hoping someone will bookmark this and reference it to others when they see the same issues. The same thing happened here. http://www.theshirtboard.com/index.php/topic,6087.0.html
As you can see in both post, (both images) have very soft fades out to nothing.


This is always hard to pin point, as it can be the result of multiple variables.
1, Mesh to halftone count relation. Do you have a chance of holding the smallest dot possible.
2, Stencil thickness, as it pertains to the availability of the ink to travel down through the stencil.
3, Screen angle. Does the threads in the mesh interfere with the angle of the screen pattern.
4, Moire and it's cousin, (Mesh Interference)


All four or just one of these can provide the results shown.


After speaking with Inkspot on this, one factor to look at is the choice of lpi halftone on a mesh needs to be heavily considered here. You need to know what can be held (in all scenarios) of that line screen. Let me preface this by saying that some people will say, "I've used Xhigh halftone on this mesh (or even S mesh) with no problems in the past". Thats great but that does not mean it's the right halftone for that mesh ( in all cases). You happened to get away with it before, (because the limits were not pushed to the extreme) in those cases. In other words, you were able to get away with it because (that job) you had success with, did not contain (the the smaller dots as this one does or the one in the linked post for example). In both cases, the dots fade off to nothing. If we continue on our way with using a (too high) of a line count on (every) job, without considering the dot size you're trying to hold (on that mesh), you will indeed one day get bit by this very same issue.


1, Mesh to halftone count relation.
The "general rule of thumb" does not always work (for every mesh). You can get away with 4 on some mesh (usually mid to very low mesh and get away with 4.5 on medium to semi high mesh,


I prefer to use 5 (in all cases).  I do that, because I don't want to be caught with my pants down, due to the fact that I've used 4 or 4.5 on all and one day, I've not been able to hold my 4% dot on a 305 mesh. The 5. will help guide you in THE BEST chance, of getting all dots....or the LEAST chance of getting these little checkered (mesh interference) squares. What's happening is that the small dots here are potentially getting blocked by the size of the mesh threads, but thats not all. For now, lets look at just how do we get the smallest dot possible on a mesh. What line count should we use?

305 mesh, divided by 4    =  76.25  (rounded to 75) line screen.  Not so good to use.
305 mesh, divided by 4.5 =   67.78 (rounded to 65) line screen.  Better, but not able to hold the 5% range well.
305 mesh, divided by 5    =   61     (rounded to 60) line screen.  BEST chance.

280 mesh, divided by 4    =   70 (rounded to 70) line screen.
280 mesh, divided by 4.5 =   62 (rounded to 60) line screen.
280 mesh, divided by 5    =   56 (rounded to 55) line screen.


Now, with that said, the visual difference of a 60 line screen versus a 55 line screen are not noticeable, but the production benefits of using the 55 outweigh the idea of using a smaller the smaller 60lpi dot. For this reason, I prefer the 55lpi on a 280 as well as a 305 mesh.


This applies to Murakami S thread as well. The S thread does not enable you to jump to another 10 line screens higher but enables you to pick up another few more dots in the line screen and decreases your chances of the threads blocking your smaller dots). This is due to the fact that is has a larger % of open area. Again, the S threads does not eliminate the need for using 5, but does improve your chances of having (less dots blocked) to a degree.

Can you use a higher line screen on a low mesh?  Sure, many do and do it well, if they eliminate all dots smaller than what would get blocked by the threads.


A sure shot measurement, is to create a sheet of film with various line screens, using a square of a 1, 2, 3, 4, and 5%'s to view your smallest dot on film, under a magnifying glass laid up on your screen... It's been written in trade mag articles that we should use one mesh opening and two mesh threads to determine the size of our smallest dot (to play it safe). This to me, is overkill, but does work. It makes sense, that no matter the angle, no matter the mesh, (if you do this), you will know for sure what you can and cannot hold.


The first picture sample submitted by the original poster BinkSpot, is an extreme case, and the 2nd improved sample is an example of teetering on the point of (just not enough) to hold those finer dots. On this 2nd sample, (if he were using a low mesh) and too high of a line count, I'd say that is the mesh is blocking out the last few percentages of dots.  This is so common that it's accepted by many as par for the course.  Since he's indicated that he's now using a 280 (and it's looking better as we see in the sample), I now fall to the next variable. I'll note that I think #2 is the real culprit.  I say this, because in both cases, he's using a mesh size that should work well with the halftone (assuming he's using a 55 line screen)...but does coat 2:2 on these mesh selections. In most shops, using 55 line screen on either a 280 or 305 mesh, you should be able to hold at least a 5% dot. Some shops hold smaller, such as 4% and even 3%.


In the 2nd image below, the original seps show an 8% range (that should be achievable) using a 55 lpi on both of the mesh used to print these samples. In the improved sample pic, you can see where these dots are lost right at this 8% range. Indicating that it's not necessarily a mesh to line count issue but something else. This is, of course, unless he's using higher than 55 line screen, like 60-65 therefore decreasing his chances of holding those smaller dots (as the mesh threads may be blocking those dots below 8%). I believe it's the stencil thickness.



2, Stencil thickness,
After emailing BinkSpot, and his printing specifics, I also see another area of need that would make this type of job (fine halftone printing) work all that much better. His common practice is to coat 2:2 and I love that!  I love it ...on lower mesh.  Mesh below the 200 area. Anything higher, would benefit more, from a 1:1 and even better, would be a 1:1 (with a face coat). The face coat is put on (after) the first 1:1 has dried. This fills in that waffle texture we often see when a mesh is coated thin. This face coat creates a better seal or (gasket) that helps lock in or confine the transmitted ink deposit to it's intended shape. This thin coat 1:1 can be used every day for general work on high mesh, but the result of not using the face coat, can be that your edges of the images and dots may spread or wick out, preventing a sharper imaged line or dot. The thinner the ink, the harder the squeegee pass, the duller the squeegee, the more chances you have of the ink spreading or (creeping out) from underneath the stencil image (hole) that the ink passes through. A face coat is a good thing to use in CMYK printing and even sim process printing. Better image definition, more confined ink drop shape. The face coat is all well and good and should be pointed out when using a thin coat of emulsion, but it's not BinkSpots issue. I believe his issue lay more in that the 2:2 stencil he does use is (too thick) for this ink and detail.  The lower end dots are getting blocked out (not able to push through) those tiny and long or deep holes with ease.

Once you know you're using the best choice of mesh (for the halftone used), you can still get moire from other variables such as the emulsion thickness, ink type, screen angles, etc.

3, Screen angle.
22.5 is not (THE) correct one to use for (ALL) shops. There is no "correct" angle. Lets say, your shop is a little more unique and are using very thick ink on medium mesh, (thats one contributing factor), than add using a screen stencil that has a very low EOM Emulsion Over Mesh ratio, such as 1:1 of a thin fast pass of emulsion on a medium mesh or the opposite of that would be 2:2 of a very slow pass of emulsion on a medium mesh. Each of these can add to your chances of taking away from or adding to creating mesh opening to dot size/angle interference and results in what "could look like" moire...but is actually screen mesh interference. Each shop is different but in general 22.5 does work for most, UNLESS, you're using the popular S mesh. That mesh worked with various angles (depending on) the mesh size you are using. For that bran, they have a chart with many different angles that work best for each mesh size. Not many use the same exact angle.



4, Moire and it's cousin, (Mesh Interference)
Moire is the result of two screen (angles) directly intersecting/over crossing, resulting in an unpleasant or unwanted pattern. Think miss aligned cmyk with different screen angle on each color.


Mesh interference...is everything I spoke of above in # 1.  This (could be) part of the issue but I doubt it as the (dot size and mesh thread) are vastly more common. To illustrate the differences more clearly, take out the idea of "screen angles" all together. With no screen angle on a dot, we have no (what looks like) moire issue correct?  Not so. Mesh interference can come from many directions. It's probably harder to identify the source than true moire.


Look at stochastic or even random (diffusion dither) index seps.  You can print (one screen only) that has a random halftone dot (square dot) on paper, shirt or pellon....and can potentially see what people often consider as "moire",  but  how?  no line screen (angle) was used to interfere with the mesh or any other color printed down with a line screen, just the one color. The one color that was used, had no line screen in it. You only have the mesh that contains any "line patterns" of sorts. The answer lay in the (size) of the dot and on the mountain tops of the mesh threads.


It may be no coincidence, that (if) the resolution of the indexing or stochastic (random dot) file was (too close) or matching the thread count, (e.g.) 200 file resolution used on a 200 mesh, the threads begin to (block) the dots that fall on top of the threads. For this reason, savvy indexers and square dot users steer clean of anything close to (same rez as the mesh they want to use).
For 200 mesh, they may use 183-233 (or some odd number higher or lower, but stay away from (in between) that.


These (holes) will follow the shape or path of the (stretched) mesh. This is what is seen on the print and in the screen. You can see large arches, rainbow's or even spirals near the outside of your print area and mesh...and think to yourself, how the heck did I get moire?  It wasn't moire. It's mesh interference. To test this, if you were to go back to your mesh (cleaned out) of course, and use a #2 lead pencil and let the point lay in the valley of a thread. Run it down the path of the mesh lightly...and you will begin to see the direction of that mesh (where it was pulled more tighter on one side or in one direction) than another...jump over a few threads and do the same. You will begin to see arches, rainbow shapes and even spirals. These paths will for sure, be more noticeable on the outer edges and hopefully, not so much in your print area, but I've seen it deep inside the print area. This is where you get mesh interference issues (with random dots). The Stochastic and random dot indexing is always promoted for lack of moire, but this "technically" is true as you don't get moire, but rather you (can get) mesh interference.  Here again, you must know the limitations of your dot size (on the mesh) you want to use.  That pencil test is also another way to see the quality results of your stretching techniques. Rainbows are not so pretty in these cases.


Lastly, the final image shows a composite (on a black tee) as this would print on various garment colors. It's beneficial to note this as the black ink plays a part with the white base and top white. The top white also covers slightly over some other areas that is intended to be pasty pastel in color tone (as would be on the spatulas). ON the white print examples, the seem a bit punchy since no top white is overlaying.


In this post below, Mark Coudrey talks a little about the relationship of dots versus mesh.
http://www.theshirtboard.com/index.php/topic,295.0.html
« Last Edit: June 09, 2014, 02:25:55 AM by Dottonedan »
Artist & Sim Process separator, Co owner of The Shirt Board, Past M&R Digital tech installer for I-Image machines. Over 28 yrs in the apparel industry. Apparel sales, http://www.designsbydottone.com  e-mail art@designsbydottone.com 615-821-7850

Offline Binkspot

  • !!!
  • Gonzo Member
  • ******
  • Posts: 1108
Re: Checkered patern?
« Reply #21 on: June 08, 2014, 09:26:10 PM »
Thanks to everyone including Dan. We have done simulated and 4 color before but never anything this challenging. This is an in house work in progress for us. Due to scheduling I had to strip the job Friday and won't be able to try agin until later this week at the soonest. In the mean time I will coat screens 1/1 and have Megan print a set of films to match as per instructions provided. Once I set it back up I will post pictures again with the results. Hopefully better next time.

Offline TCT

  • !!!
  • Gonzo Member
  • ******
  • Posts: 2869
Re: Checkered patern?
« Reply #22 on: June 09, 2014, 12:08:28 AM »
Holy crap, great technical info there Dan! And Binkspot, I really like the concept for the print!
Alex

Hopefully I'll never have to grow up and get a real job...

www.twincitytees.com

Offline Underbase37

  • Hero Member
  • *****
  • Posts: 790
Re: Checkered patern?
« Reply #23 on: June 09, 2014, 12:46:23 AM »
Dan, that was all great stuff thank you. That old post by Mark is exactly what I have been looking for too, read it a while back & couldn't remember where I read it.;)

Murphy37


Offline Dottonedan

  • Administrator
  • Ludicrous Speed Member
  • *****
  • Posts: 5875
  • Email me at art@designsbydottone.com
Re: Checkered patern?
« Reply #24 on: June 09, 2014, 01:40:31 AM »
Thanks. That last illustration of ink held up in the screen was pretty bad,but it gives you the idea and thats the important part. :)
Artist & Sim Process separator, Co owner of The Shirt Board, Past M&R Digital tech installer for I-Image machines. Over 28 yrs in the apparel industry. Apparel sales, http://www.designsbydottone.com  e-mail art@designsbydottone.com 615-821-7850