Densitometer for linearization of DTS/CTS

[BorisB]

What densitometer is used to linearize RIPs used on CTS?
I just. tried with one Viptronic model and it really got no usable readings from printed T-shirt.

Or you do linearization by printing on paper and measuring dot gain on printed paper?

[Alex M]

We use an Xrite 301 and print onto film from the i-Image then measure that. No good way to measure at shirt that I have seen yet. Not to mention all the other variables like screens, ink, squeegee duro, speed for flood print... Etc you get the point.
When we linearize the i-Image every screen that comes out is using the same curve and should result in the best consistency.


Sent from my iPhone using Tapatalk

[ABuffington]

There are 2 points in the pre-press process for linearization which both control dot size on film.  First the film, it gains 10-20% on ink jet output.  X-rite 301 or341.  Print a tonal ramp on film from your inkjet, in Wasatch you can use calibration curves to input the readings, the RIP then outputs proper halftone percentage to film. Everytime you change image configuration you may need to apply this curve adjustment in the hafltone/calibration section. Go from 55 to 45, 45 may not have the curves applied, has to be done manually.  Save the curves so you can load them, run a tonal ramp, measure and run again.  The next linearization controls the dot gain in print.  T-shirts do not read very well due to the texture of the surface.  This can be done printing on paper and reading with a reflective densitiometer, some higher end reflective densitiometer (6800-9000k) may be able to read off fabric but quite often it varies with each reading.  You can print a black on white and subjectively compare to your film output and back off manually in Wasatch with 'Press Curves'.  In most cases dot control is iterative in print.  Subjective viewing often shows 60-95% tonals as solid in print.  To control you can add 5-10% to the tonal amount to enter into the curve dialogue box in your rip to open these tonals up so they aren't solid in print.  Wasatch has 2 linerarizations, Calibration and Press Curves, Accurip has one tonal control area, but I heard where they have removed this on updates?  Just downloaded a demo and it still had it in there.

In any case, save the curves you arrive at for use in other areas.  One brand shirt may dot gain differently, same with inks.  This dot control really improves 4/C process and sim process work.

Al

[ZooCity]

Alan, does Wasatch allow you to use the press curves even when separating in your design program?   Filmmaker has this feature but only for in-RIP seps, i.e., not useful for us textile screen printers. 

Boris, like Alan said I would linearize the CTS output to screen first, then adjust manually for the press gain.  IMO, press gain is subject to so many parameters- substrate, ink type, mesh, blade, and so on that I can see having a generic press gain setting that is used and then adjusting the channels themselves or using additional press gain linearization setting for various substrates/print methods.  Unless you print one way on one kind of fabric I don't see how to lock this down.

Alex M, I don't know if it exists but would it be possible to make an opaque white emulsion, coat then face coat so it's nice and smooth and use that to linearize a CTS?  Then you would have a more true representation of gain on the actual substrate. 

[Dottonedan]

You have gain from the type of printer used to output the image. That can be an inkjet with one amount of gain and a CTS with another amount of gain. Then, you have Wax vrs ink, and 1 head vrs 2 and 3 head, high sped, low speed, ink resolutions such as 1200x900, 1200x600 etc.  All of these are variables. This is controlled or adjusted/accomodated for in the RIP curve as others explained. All of these produce more or less gain but all devices gain before going onto your film or screen. Therefore, they must have some compensation done when producing halftones. How much depends on your preferences. I've seen some shops want to control gain exactly and I've had some that chose to leave it at default aka (non curve dot gain adjustment at all).

Zoo,
This is all done in the RIP. You SHOULD do this at output and allow your RIP to do the adjustments. Otherwise, it's all by eye in the program and individually per job per person. If Tommy is your guy in your shop that does all of this adjustment by eye and garment order type, and Tommy leaves for another shop or to become a rock star, you are left with no control.  The most efficient process is to allow your artist to do art and seps (as it should be), aka a 80% mixes in one color sep to a 50% fill into another sep, thus mixing, (stays) at that when sent to the RIP. Let the rip do it's thing (based on the correct or best results you've been able to test for and entered into the RIP dot gain control).

Once you determine how much you gain on the emulsion, exposure and wash out, you are still not done. Even though you cannot get easy readings from a printed tee, you can get averages.

From the printed tee you can use a transmissive and reflective densitometer and read multiple areas of a locked in % such as 5% and average them together. Then read a 10 or a 50% area multiple times and average that together to get a more accurate reading. THIS would be your final output adjustment.  Subtract what you are getting in your curves and then print it again and read/document again. Then, you should be very close to as best you can get.

Yes, you will have other variables such as garment type, squeegee type, stroke speed, ink type, off contact. That's a lot more research and testing to do (if) you want to save a dot gain control for each variable of garment types you can come up with....but i wouldn't.  Stick with one, and then adjust on press (as needed).

Mark Coudrey and many of the science guys you read about in the articles may have gone to this length to test and save a dot gain control setting for each scenario, but lets get real. Much of those additional test may have been done (for testing and results sake only). They make good scientific articles, but not every scenario is feasible on a day to day production base. It's too much to be concerned with when you can adjust enough on press (by experienced eye) to get it to look right without changing all of your settings at output for each different product order. I mean after you have one master dot gain control, the most you would be changing is does the 95% print at 96 or 98?  Maybe the 50% that is cut back to 38% for X garment and mesh and ink, is now cut back to 40% for this garment, on this mesh, with this ink. The only time I would do additional sets of dot gain scenarios is if going from a tee to a sweat doing simulated process and keep one for halftone on 200 mesh and up and another for halftones on 156 and lower.  You can balance out the subtle differences on press with press skill.

Printing the test on basic black and white copy paper gives you one gain result. Printing the image on Photo paper gives you another gain result and printing on emulsion gives you another gain result. So I prefer to average the prints using black ink on a white shirt.  If you base the % of gain off of various types of porous/fibrous paper or emulsion only, there is still gain beyond that.

In a nut shell, save yourself some time. Trying to accommodate for every little nook and cranny that can affect dot gain can be like chasing your tail. If you get there in the end and have 30 different dot gain scenarios covered, the end result is that it can be all wiped.

Lets say the screen room used a 156 when your gain called for a 200 but they ran out of 200's. Maybe the screen room used a screen that was not fully dry. Maybe the screen rm had Buck out sick and Bill coated 3 times instead of the 2.  Sure, you, the owner or mgr is to have all of this running smooth and every one of your guys have everything down...and doing their part correctly all of the time right?

Ok, should we just not test for dot gain control at all and throw caution into the wind? of course not. Knowing dot gain IS paramount, but going to far can be useless. If it were me, I'd invest more time and skill into those press people and hold onto them.  How many know WHY you choose a 110 on X job over a 156?  How many of your press people truly understand how fast you can run a stroke before it becomes an issue 50 shirts in on sim process? What does stroke speed or off contact do to dot gain? Can you print the same color 2 times when printing halftones?

All in all, do enough averages on a printed tee, save that for your high mesh, do another for lower mesh, maybe do another for sweats...or any other unique substrate...and let that be that. Then rely on your press people.

[ABuffington]

Wasatch uses DCS EPS 2.0 files from photoshop typically, so all plates are pre separated for sim process, same with with sep studio.  The curve applies to all film halftone output for all files sent to the RIP. When well linearized it will output solid 100% black and will show a 95% tone as well down to a 5% dot.  Yes you can pre-curve grayscale value from photoshop for more punch and tonal control, or do this in Wasatch as well, this will add punch to a linearized system. (great for black shirt printing)  Controlling the dot gain for inkjet, or CTS helps textile printers out a ton.  As you mentioned each ink, or shirt type, or halftone ouput may have it's own curve needs.  But regardless, if you start with a 10% tonal value ink jet output reading 20-30% with a densitometer measurement it can only improve your tonal prints to have it linearized.  At least the film is at the correct tonal value.  For print dot gain that is totally subjective and is the place an additional linearization can be performed.  Proof of all this control is my mentor, Mark Coudray.  His 4/C process is so flawless due to controlling the printed dot value controls used in pre-press.  We can only print what's on the screen.  If the tonal value is 10-20% higher, sim process and process will not be as good as it could be.  Sharper prints, much more subtle colors, better neutrals in 4/C process, less muddy prints makes linearization worth the effort.  The recipe can be a pursuit for months or years, or a never ending evaluation with each ink system, shirt, halftone shape, and tonal value all slightly different in curves.  We had one for each ink system, shirt brand, 14,18,24 singles, cotton/poly.  The list is endless.  However Wasatch allows as many as you want to create and you can save them all, and print it out image configuration on your films for reference.  Nice to know when your 4/C process is muddy and you check films or screen and find out a linearization for a coarse shirt using viscous sim process inks was used for films.  As great as all of this sounds, all it takes is using a dull squeegee, a different durometer, different viscosity inks, A real cold day in the shop, a real hot day in the shop, too much pressure, speed, flood and on and on can screw up the final print as well.  Pre-press can only do so much, but at least you know it isn't the reason for a muddy print, just know it is controlled so you can move on to the other nth variables that make this a hands on process.

[BorisB]

Thank you all.

The measurement of film printed with I-Image helps getting correct information from RIP to film. Not to screen, which can suffer undercutting and doesn't have as smooth surface as film. As I have no experience with linearized VS. non linearized RIP I don't know big improvement it is.

Off course biggest dot gain is generated when pushing ink through squeegee.

If we don't measure this step, is it of big help to linearize imaging on film which isn't producing same dot gain as imaging on screen with rougher surface and different surface energy and such different spreading of dots?

To us most important are 305 screens with 55 lpi halftones on good quality T-shirts. We use same ink, same squeegees, same angles, same off contact, same emulsion, same exposure.  All others combinations of screens, halftones... are not important. Would differences in substrate/shirt brand and other minor variables influence results as much, that measuring is waste of time?

As I'm unable to measure t-shirt with our reflective densitomer, I will take transmissive one and measure print on film with plastisol to see where that takes me.

[mk162]

I measured a shirt with our reflective densitometer and everything came out around 35-38% gain on a non-linearized RIP.

We use an X-rite 404.  We didn't go crazy with it because we were still testing it out and wanted to make sure we were on the right path.  We adjusted the curves and it made an enormous difference.

[jvanick]

we have an X-rite 418...  I will echo brad's results and say that we are around a 30-35% gain as well... our 'curve' ends on the '95' mark at 70%.  There's probably still some tweaking, but I'm extremely happy with the halftones we've been getting.

of course, now that we're dialing in our emulsion and thickness, I'm sure I'm going to have to recalibrate again, but doing the initial curve adjustments made such a huge difference.

[mimosatexas]

I hope I'm not retreading ground here or misunderstanding stuff, but when you are adjusting for 30% gain, are you essentially making your halftones 30% smaller across the board on your film.  The reason I ask is if people are talking about holding 5% dots at 55 LPI (uncalibrated?) and those suddenly become 30-35% smaller, aren't you now printing dots you can't hold?  Is gain less of an issue at the lower end or the range than the higher end?  I guess I'm just curious if linearizing to adjust for gain ends up meaning you now have really accurate range from 95 to 30% dots, but all your lower range dots are basically gone, unless you lower your lpi, which would of course necessitate another calibration of some kind I assume.  Just seems like a rabbit hole and would love to have it explained.

[jvanick]

you linearize in different parts of the curve, adjusting for gain based on what is being output (and printed) vs. what's in the art.

example for my densitometer to read 5%, I need to output a 8% dot due to a 5% dot not holding.

then the curve 'curves' from 10% to 90 and then to 95% where the output for a '95% halftone' is actually 70%. When the screen gets rinsed out the 70% is actually nearly completely open due to the inkjet dots 'spreading'.

it's not a linear -30% mapping.

[mimosatexas]

got it, thus the "curve".  I thought that's how it worked, but I was reading some of the above differently when it came to reading different percentages on the print and adjusting.  So essentially you are compressing the 100%-0% to 70%-8% correct?

[Dottonedan]

A 5% can have 30-40% gain on average. Let's call it 30.

5% + 30% gain = 1.5.  Totals 6.5   In my rip there is no .5 so you either rounded up  or down leaving it at 6 or 7.  That's what your output would be, a 5% becomes 6-7% and so on.

1% May be looked at as coming out as 1.5 or even 2% etc.  one thing you don't want to do is mess with the 1% area too much. It makes pho shop files with outside image are fragments show up and sometimes hold rasier. ASI do not put 1% in the 0 area. That fills your entire document with 1%.

(25% area input) + 30% gain = (32.5% output). That's a difference of 7.5  so you subtract 7.5 from 25 to get your compensation of 17.5 in the 25% area. So when you read your output and you get 32 or 33% in the 25% zone, you got 30% gain.

[jvanick]

So essentially you are compressing the 100%-0% to 70%-8% correct?

99%-1%

100% and 0% are left alone as like Dan said you can have wierd things happen if you do that.

[Alex M]

Boris
If you have an ST series I have true curves to film from the unit for a set of standard settings.
If you want to email me Alex.mammoser@mrprint.com I will send you the curves and settings.


Sent from my iPhone using Tapatalk