TSB
screen printing => Separations => Topic started by: drdot on May 06, 2011, 11:22:32 AM
-
This is a very, very common question and there's usually some "machoism" associated with it. You know, I can print 85 lpi all day, and so on. Here are some things to consider. Just because you can print a finer dot, doesn't mean you should. This will be the first of several posts on choosing lpi based on production considerations and related factors.
The choice of LPI for your halftone is determined by several factors. Chief among these are:
Substrate Limitations.
Viewing Distance.
Screen stencil resolving capabilities.
Production variables (mainly moire and dot gain.)
The coarser the dot you choose, the easier it's going to be to make an accurate stencil, control moire, dot gain, and tonal range reproduction.
On the flip side, the finer the dot, the more detail you can hold and the less noise the image will have (surface texture that interferes with the image.)
For textiles, the most common ranges fall between 45 lpi and 85 lpi. Some very experienced printers have gone as high as 110. Here are my recommendations to help you choose the best resolution for your purpose.
First and foremost, consider the "dot density". This is how many dots are in a given area. This is really enlightening and will have the single biggest impact on your choice.
45 lpi 2025 dots/sq inch
55 lpi 3025 dots/sq inch 49.4% increase over 45 lpi
65 lpi 4225 dots/sq inch 39.7% increase over 55 lpi
75 lpi 5625 dots/sq inch 33.1% increase over 65 lpi
85 lpi 7225 dots/sq inch 28.4% increase over 75 lpi
110 lpi 12,100 dots/sq inch 67.5% increase over 85 lpi
The dot density is the lpi squared. Since dot gain is uniform around the perimeter of ANY size dot, having the dots smaller and closer together radically magnifies the gain you will experience on press. For an ideally controlled halftone, normal dot gain(defined and explained in another post) will typically be:
45 lpi 22% - 28%
55 lpi 32% - 45%
65 lpi 38% - 50%
85 lpi 50%+
High dot gain is not necessarily a problem. It only has to be managed effectively. When you consider the relationship between lpi and dot gain, the most control occurs in the 50 lpi - 60 lpi range. This is why it is so often recommended.
Add to this the fact that there will be increased tone clipping on both the highlight and shadow end of the tone range as the smaller highlight and shadow dots will be clipped by the thread diameter of the mesh. Based on 34 micron thread diameter the smallest consistently (meaning all dots print with no eclipsing) printed dot will be:
45 lpi 2%
55 lpi 4%
65 lpi 6%
75 lpi 7%
85 lpi 8%
110 lpi 14%
With todays high resolution dual cure emulsions, you can resolve 1% dots for ALL of these linecounts, you just cannot print them due to the threads of the mesh being thicker than the dot itself thereby blocking the mesh opening.
I'm sure this is going to raise some questions, so let the discussion begin.
-
Good stuff.
-
On the flip side, the finer the dot, the more detail you can hold and the less noise the image will have (surface texture that interferes with the image.)
Great post Mark. If I am reading this right higher line counts help when printing on t-shirt fabric? I was thinking that if the line count is too high that some dots hang in the screen because they don't contact the substrate during the squeegee pass which in turn may cause more dot gain.
Also thanks for providing the data on the dot gain percentages and smallest printable dot.
-
I'm begining to see the value of larger dots as part of the design element.
For example... I really like the way the dots add to the design Artelf2xs posted last week. (see link... REDLINE ROCKETS text)
http://www.norwb.com/index.php?topic=301.0 (http://www.norwb.com/index.php?topic=301.0)
-
I think I'm going to have to read that a few more times.
-
I think I'm going to have to read that a few more times.
I know, right?
DING....school's in.
-
I think I'm going to have to read that a few more times.
I have and am still trying to wrap my head around it . . .
pierre
-
Mark,
I'm wondering something about the thread diameter and if those percentage limits could be be pushed. When we do simulated process, we work with 55 lpi output with 300/34 mesh, and I am happy to stay at 55lpi as customers are happy, so I'm happy. Considering your statement that on a 55lpi output that a 4% dot is the smallest consistently printable size on 34 micron thread, what do you think of using a 330/30 S mesh in terms of printabilty of smaller percentages? It appears to me from Murakami's mesh chart that a 300/34 and a 330/30 should behave similarly in terms of ink depost.
Pierre (bluemoon) mentioned this somewhere else and I've been intrigued by the idea ever since but haven't tried it yet. If it is the case that 330/30 might allow a printable dot a couple percentage points lower (or higher) how much differnce could it make visably on a print?
I'd love to hear Pierre's take on this as well.
TIA
-
I work closely with Pierre and know he's getting a better halftone because of the new mesh. His coverage is better and at the same time he's able to hold at least an additional 1% lower. We seem to get a good consistent 3% dot out of a 55 line screen.
I have what I think is another interesting topic that is somewhat related to this post. I would not want to begin to stir this one in another direction so I will create another post.
-
I'm wondering something about the thread diameter and if those percentage limits could be be pushed. When we do simulated process, we work with 55 lpi output with 300/34 mesh, and I am happy to stay at 55lpi as customers are happy, so I'm happy. Considering your statement that on a 55lpi output that a 4% dot is the smallest consistently printable size on 34 micron thread, what do you think of using a 330/30 S mesh in terms of printabilty of smaller percentages? It appears to me from Murakami's mesh chart that a 300/34 and a 330/30 should behave similarly in terms of ink depost.
It is possible to pick up that extra 1% by dropping to a 30 micron thread. It gets more complicated, but it is doable. I need to put up several more posts about the relationship of actual thread count to dot area. It gets pretty deep, but it will explain a lot of the behavior that takes place in the very low percentage areas.
There are so many related behaviors, it's hard to make a general post like this and get the full picture of what's actually going on.
-
@Orion,
It's a tradeoff between detail and dot gain. I've got all sorts of graphs that show the optimium possible. It comes down to how calibrated (level and parallel in the same plane) your press is, how sharp your squeegees are and how careful you are. The single biggest player is the ink. With the right thixotropic profile, you can do amazing things. Here is a really cool trick. Add 2% - 5% high density base to you colors and you will dramatically improve (reduce) the dot gain in the midtones and significantly open up the shadow detail. This allows you to use a finer line count to capture more detail and still maintain full tone range.
-
High density base? Not high density clear? Same thing or different?
-
Mark...... I'm pleased you mentioned blade sharpness. Quite often one of the most overlooked interdependant variables. I know of many flatbed printers that dispose of the squeegee material when finished with the run and simply replace it just for the secure knowledge that it is a stable variable. Correct me if I'm wrong; I know of no tool to measure this.
On a quick sidebar often when printing straight process on white we often create a "wet white" (non flashed)underbase that helps to control staurated hues as well as soft pastel shades. Also improves fibrillation. Sort of goes against the rules and some may call it a bandaid but I have to say it works quite well.
Best tp
-
On the high density base, I believe Mark is talking about an additive. I have some made by Wilflex.
http://www.polyone.com/en-us/about/businessgroups/Inks/Literature/Wilflex%20High%20Density%20Additive.pdf
-
Mark is making an excellent point about how the "body" (the rheological properties, how much it flows/pours vs. keeping it's physical shape) of your ink has a large impact on how the dots hold their shape when deposited on the substrate. Be it the garment itself or a white base plate.
It is VERY important that the ink be just ridgid enough to maintain it's dot shape both during the initial ink deposit (will not flow out under the edge of the stencil) and when it is "steped on" during the wet-on-wet process.
Some ink lines will need an additive to acheive this "controlled ridgity" while others already exhibit these properties.
Loving the discussions here by the way! I've really missed this level of interaction...
-
Thanks. I have HD Clear on my shelf. Guess I better see if I can squeeze someone for the additive.
-
Mark...... I'm pleased you mentioned blade sharpness. Quite often one of the most overlooked interdependant variables. I know of many flatbed printers that dispose of the squeegee material when finished with the run and simply replace it just for the secure knowledge that it is a stable variable. Correct me if I'm wrong; I know of no tool to measure this.
Tony, you bring up an interesting point. I have never trusted brand new squeegee material. The reason is not the uniformity of the edge, but rather, the ability to clamp the new blade so it is absolutely flat. I used a Svecia squeegee sharpener for years. I rigged up a dial indicator that allowed me to parallel the guide rails to .001". This was mostly for very high res graphics printing (133 lpi and 150 lpi). The blade had to be perfectly sharp and perfectly smooth.
To determine if your blade is sharp enough, drag your finger tips lightly across the edge of the blade. If you can feel the individual ridges of your fingerprint, the blade is good to go. If you can't, the blade is too dull.
On another note, we sharpened our blades at the very least weekly. This meant that we were only taking off .001" -.004" of material. It was very fast and we could do it in 1 or 2 passes. Kind of like a butcher using a steel to bring the edge of the knife back. If you wait too long, you have to take too much material off to return the edge to printing condition. It also takes too much time and this is the main reason printers do not want to sharpen their blades.
-
Pierre (bluemoon) mentioned this somewhere else and I've been intrigued by the idea ever since but haven't tried it yet. If it is the case that 330/30 might allow a printable dot a couple percentage points lower (or higher) how much differnce could it make visably on a print?
I'd love to hear Pierre's take on this as well.
TIA
well, here is the deal . . .kick back, it will be long!
Before trying the S mesh, I was printing using static frames. They are in pretty decent shape for statics, and I buy more on regular basis. The ones I have go through some pretty harsh cleaning and coating, so they tend to delaminate over a period of time. They surprisingly do not lose tension very much as we print with very low off contact (about a 64th of an inch) and rather low pressure (if we can). So decent statics vs. the EZ frames with "S" mesh. Statics at 15N-18N and EZ's in mid 20's. This for 305 statics and 330/30 smart mesh.
After a couple of visits by Dan, 5% was solid on statics. Some days are better than the others, but I can count on 5% without much trouble. 4% will open, but not all the way. It seemes that I am hitting some kind of a wall at 4% and anything lower just does not open. This is at 55lpi.
For the competition print, I re-leveld the press, used new emulsion, new platten tape, new squeegees and a new bulb in the exposure unit. Also all the screens were shielded from the light before exposing (I walk through the production area with unexposed screens on my way to the exposure unit. Normally I just walk through quickly and let the light hit the screens. In this case, screens were double bagged on the way through the lit area). We also did a comparison between two RIPs and concluded that the Ghostscript halftones were much darker and cleaner shape. We used 58 lpi for this print. 65 was having some issues and 60 divided evenly into 330 so we had to chose between 58 and 62. 58 seemed like a better bet.
So there were many changes between the regular prints and the high end stuff. But here is the kicker!!! The 3% dot was completely open with no visible interference from the mesh. 2% was open better than 90% and 1% was open better than 50%!!! Unlike the regular screens, there was no sudden drop off point and all the dots were opening. The obvious issue with the 1 & 2 percent was the mesh interference. We checked with a microscope and could tell that some of the dots were blocked. In most cases the blockage was not complete, but rather partial, but since it repeated at predictable interval, it generated a moire.
Another interesting (read: worthless trivia) part was that there were various levels of 1% in the art. Anything below 1 was not visible on the screen, but there were at least three different shades that were less than 2%. Setting the color wand on 0 tolerance, I could go and pick different shades within 1%.
We actually spent two days discussing how to get rid of the 1 and 2 percent halftones as they were clearly printing on the shirts and were showing a really bad jagged patterns. We could not and still do not know how to eliminate everything under 3% cleanly and had to resort to masking tape to block those off and create smooth edges. Dan thought I was crazy for doing it!
And now to throw another curveball, knowing what I know now, the 3 percent dot that we are holding is actually a lot bigger when printed. Even though the dots on the shirt are so fine (they are barely visible), it is nowhere near 3%. Yes, the photoshop calls it so, but if I had to guess, they are probably around 10%.
As far as the difference on the shirt, I don't know if it is visible in this particular example. I think holding a 5% with a good screen would have yielded and impressive print, but I also feel that the 2% difference is probably what made it an award winner rather then just an outstanding print. I don't think, actually I am pretty certain that an untrained eye would not know the difference.
That 2% does come into play with 4CP though. There the art is heavily dependent on the mixing of inks and many colors require a very small quantities of other inks to achieve a certain color. The example I often cite is a print for one of my customers that has a gold medal in it. The gold color is 100 yellow and about 3-4% magenta. Before getting the new screens, I had to bump the percentage up to 5-6% so it would open and this in turn would make the medal rose gold rather than yellow gold. By being able to hold the 3% I can actually make it be the correct color. We don't print as much 4CP any more, but if we do now, it always goes on the new screens and it prints beautifully with very soft transitions and pretty good color.
pierre
-
You had to tell everyone about the masking tape. ::) Ha! I forgot about that. Probably intentionally. :)
-
This is a very, very common question and there's usually some "machoism" associated with it. You know, I can print 85 lpi all day, and so on. Here are some things to consider. Just because you can print a finer dot, doesn't mean you should.
Not only do I agree, we have additional points to consider.
Anyone who thinks printing on garments with HTD above a range of 40-60 (at least doing it well) may well be fooling themselves.
Here is a photo of a popular brand (6.1 100% cotton) in black with white plastisol ink of 42 LPI 62 angle (microscope with 50 power magnification)
(http://i60.photobucket.com/albums/h40/douglasgrigar/White_on_black_42lpi.jpg)
None of the dots pictured here are even close to 2% and that is not the issue but rather the capability of the substrate to hold the dots in position.
This has not been washed, it was directly out of the dryer.
Note the left circled dot - broken (again this is not a washed garment).
Now check out the right circle - only a very fine thread shows any ink... the dot is there, it exists on the screen and ink was pumped out... Where did it go?
It is actually stuck to the face of the screen stencil.
When will that ink get deposited - we do not know, and when it is finally deposited a ?double hit? of ink (or more) will end up in that position. There are several examples of missing dots that fell into the ?thread caverns? and of course stayed on the face of the stencil.
Now the print is far from perfect and is not intended to, it is simply a sample of something you would see in any HT print from ?joe printer? to examine the effects of the substrate on HTDs.
Basic stuff - but points out the massive numbers of variables any printer even ?joe? has to contend with.
Change the substrate and we change the possible or potential dot.
-
Add 2% - 5% high density base to you colors and you will dramatically improve (reduce) the dot gain in the midtones and significantly open up the shadow detail.
Mark,
Just tried this idea of yours on a sim process job where the black was printing too dark and was on the verge of needing an edit and a new screen. We tried everything including up to an 85 duro sharp blade before the HD additive, but with the additive the gain was cut considerably and made just the difference we needed.
Thanks for the tip!
-
. . . .this is such an important thread, worth another look, so I'm bumping it. . . .I am going to print this sucker out too.
-
Good job Jay. I'm printing it out also. Mark is one of those guys I can listen to for hours. If anyone has the chance to see him, it's worth your time and money. Every time!
-
I have learned a lot from Mark over the years, I wish he'd drop in more often, this thread alone helped me out of a couple jams. Just as I'm rereading it I'm thinking it's time to sharpen the squeegees, we've gotta keep up better on that.