Transmissive or Reflective Denistometer for CTS imaged screens?

[Orion]

Interesting print philosophy Danny, we do the same.

[blue moon]

Interesting print philosophy Danny, we do the same.

as do we, but am working on making things significantly more predictable. I have few ideas that should allow less experienced printers to dial in much quicker.

pierre

[jsheridan]

Didn't know what a densitometer or linearization was before 2005.. it was all eye till then.

Like Danny said.. it's art and when it looks good. there you go.

 

[Dottonedan]

It's an art form and THAT can be the dilemma depending on the type of art your shops handles.


When you are a custom shop that does not create most of your own art or if you send out a lot of your art and seps, to freelancers, then having your output more "streamlined" or consistent internally is appropriate.


If you are a shop that controls all of your art creation and your seps, then the art form can work very well for you since you kinda already know what you need in your seps (as you build art and seps).

[Orion]

True Dan, our formula works well with our seps, output, mesh counts and inks. In another shop it could easily not work as well.

[Full-SpectrumSeparator]

It's an art form and THAT can be the dilemma depending on the type of art your shops handles.


When you are a custom shop that does not create most of your own art or if you send out a lot of your art and seps, to freelancers, then having your output more "streamlined" or consistent internally is appropriate.


If you are a shop that controls all of your art creation and your seps, then the art form can work very well for you since you kinda already know what you need in your seps (as you build art and seps).

Ok I will try to keep this as simple as possible...   

Mixing Ink....    as in you have a job that calls for a custom color / pantone color etc.... and you need to mix it up from your set of inks,  possibly a mix system designed to allow you to reproduce certain values with certain mixes of the base components, or maybe you just have a few inks or colors to work with...     

I have experience with mixing inks both with scientific systems and measurements and tools, and just with basic colors "by eye".   For different circumstances I will choose one or the other or use a mix of both at times to get to the target color.   

A simple question, is ink-mixing only an art-form, or are their scientific ways to do it and scientific tools and systems to make it so you don't even need a human operator? 
Just punch in the number and you get your desired ink mix?     This link video attached seems to show some sort of scientific machine that can mix colors... still has a human operator as in they press some buttons, but is the ink-mixing process and the variables and values used just chosen "artistically" by the operator?  Or is he making decisions based upon the science and entering the correct program numbers to get the desired ink color mix?

-Have someone pick some random colors and mix them, that is creative, pure art - an original swatch color.

-Pick some random colors just totally without even thinking about or seeing the color swatch, and you match it exactly with some other set of random colors perhaps or you just picked the exact ones to match, and you did it without looking, and you do this every time perfectly?   That would be genius superhuman ability, and the "artistic" way of matching an original color someone gave you to match.

-Match that color by mixing not knowing what they used to mix it, but you can look at it, and guessing and trial-and-error until you get close and have it... thats the eyeball/artistic way, still uses science because of measurement-observation-experiment-change-adjust-trial-and-error etc, using your eyes and understanding, experience etc.... mixture of art and science to make it happen.

-Match that color swatch they gave you, by the reference they wrote of how much of each percentage of colors and what colors they used and how they mixed them etc, using devices to measure percentages, math and calculations and weights etc... that is the almost entirely scientific way of doing it, apart from lack of certain probable attainments of absolute perfection in a certain sense, but you call that "tolerance" - but essentially you mix and match a color the same way someone else did and referencing their recipe and experiment to get the same result -- that is science and the scientific method of matching/mixing a color.   

All of that can be applied to the discussion of color separation and halftones and linearization for screenprinting.



https://www.youtube.com/watch?v=zfByVlzucTc

[Dottonedan]

Let's say that we kept it simple ...but didn't. You assume way too much. Try to just post interesting things.  You have a lot of good to offer...somewhere deep inside your post, so just leave direct comments towards individuals out of it.

[Full-SpectrumSeparator]

Let's say that we kept it simple ...but didn't. You assume way too much. Try to just post interesting things.  You have a lot of good to offer...somewhere deep inside your post, so just leave direct comments towards individuals out of it.

I apologize Dan, I went and removed the personal references and assumptions,  comments directed towards you, it was not my intention.   Have a good weekend.

[Dottonedan]

No harm, no foul. I was a bit thin skinned tonight. My apologies for responding with the self worth thing. I have no basis for that comment.

Thanks
Dan

[ZooCity]

So we've ran a decent amount of halftone work at the stock settings on our CTS and things are going extremely well.  I don't think that further linearization at the output device will be necessary but expect I'll need to do some tweaks to that linearization.  At this point our 1-5% dots are clearly about 3% chubbier than their "true" size which is something that concerned me at first but it's been working well.

One thing I didn't expect was the extreme improvement in the % range we can hold with the glass and film out of the picture.  Even though our low % dots are as I said a little beefed up we're now imaging 1%, and I mean truly holding it on 225 mesh at 55lpi.  That's nuts.  It's not really functional as that micron/tpi mesh's knuckles are still blocking most of those dots from transferring ink of course but it's been impressive to see.  Holding true over 99% of the tone range has, so far, given me what I wanted in that I don't need to trim back seps like I used to.  Much more experimenting to come but it appears we can kick the other lpi's we've traditionally used- 45, 50, etc. and just run 55lpi on 95% of jobs since all our screens can image it now and print it with excellent results on press.   We did a dc fade the other day that was gorgeous, pure tone and nary a dot to see with the naked eye.   I expect it will improve even more once we replace the warped array on our expo unit. 

[ZooCity]

Update on this. 

We had a situation that I should have planned for better- head replacement on the CTS.  Not only that but a change in droplet size and resolution. 

I should have stuck to my original guns and linearized at the print head to a consistent, easily readable substrate.  Had I done this I would not be working today to tune our new printhead back to the output of the first.  As it is, there is no way for us to get back to our prior output values because those were adjusted using the wrong method.  I say "wrong" because while the curve worked great for us, it was entered in the incorrect way.

I've learned there are 3 main steps to this process:

• Output an uncalibrated/raw target to your device, read it, enter values to allow RIP to adjust and generate a calibrated target.
• Once calibrated/truly linearized above, use Tone Curves to apply a boost or a cut to various % areas as needed for your process.
• If desired, and if your substrate(s) and presses allow for this, use intended and actual press values to further adjust for your process.

At least in our RIP workflow (Harlequin), this is the correct procedure. 

Using film adhered to a screen I went through the correct linearization process at the head, tuning all values to within 1%, or the accuracy of the screen reader I'm using and as best as the output resolution will allow.  i.e., at 600dpi, one cannot get a true linearization in the 1-4% range, not enough pixels available to assign to those values and make them different.  The point is that I have something that I can return to on a different print head or device.

So yes, the "true art" of this process comes in at the Tone Curves and/or the actual/intended press adjustments.  Same process as initial linearization but for textile screen printing I'll be using intuition and my eyes to tune and compress the curve to allow for what our screen can image, how our separations are typically adjusted and how our prints gain on press. 

An exception would be if I one can get their screen reader to accurately work on resolved stencils.  Then the Tone Curves part will also be measurement based and the press adjustments will be where the art form kicks in but I doubt any device can measure a transparent pink stencil over dyed yellow mesh very well at all.  If this were a reality then all our mojo as printers and artists would have it's correct place at actual/intended press adjustments.

So I just wanted to clarify that yes, this is an art and your final curve adjustments will reflect that but do not skip basic linearization at the print head or you won't have a baseline to start with and will have to start from scratch each time you change output devices or even change a feature of your current output device.   I know RIP workflows vary but so long at least two layers of linearization adjustment are available you should be able to achieve both device and then your preferred on press linearization. 

[Dottonedan]

I've learned there are 3 main steps to this process:

• Output an uncalibrated/raw target to your device, read it, enter values to allow RIP to adjust and generate a calibrated target.
• Once calibrated/truly linearized above, use Tone Curves to apply a boost or a cut to various % areas as needed for your process.
• If desired, and if your substrate(s) and presses allow for this, use intended and actual press values to further adjust for your process.
• [/l][/l]

• Re #2.  when you say "tone curve".  Are you referring to inside photoshop, and narrowing down a "standard " tone curve that you will use all of the time or, will this be an arbitrary (unique each time) curve applied based on the art?  To me, a standard would make controlled sense (inside photoshop) but, I would imagine you would really want to gather this info based on your "averaged" end results on press.
  
  
  If I had to guess, I'd imagine 1/4th of your gain is on press.  That's 1/4th, but still provides some gain. So when you control your gain in the art (#1) and in the output (#2), and (#3), at your screens exposure and washout, you still have a good portion (1/4th) of gain on press.
  
  
  These can be adjusted with art skills of the pressman like you mentioned.
  All of these, as you know, are contributing factors to that 1/4th.  Mesh selection, Inks, angle, off contact, stroke speed, Duro,. Any one of these that changes greatly, can affect that last portion. Therefore, I find this area to be vital.  Vital only to the point of total control.
  
  
  This is where you might choose to have 3-5 different saved settings for gain (on press) for different order scenarios.
  
  
  1, 55-65lpi high mesh, light garments.
• 2, 55-65lpi high mesh, colored/dark garments
  3, 55-65lpi high mesh, light bleeding garments (underbase) such as comfort colors.
  4, 45lpi low mesh, light bleeding garments (underbase) such as red sweat shirts.
  
  
  Having the first 3 portions are fantastic and far more than most shops do at all.
  Having the above scenarios (averaged out) using a densitometer to read the shirt should help.
  Some say "you can't get a good read" off of the shirt due to the garments loft and weave and sporadic readings...but you can average. Take 5 reading of a square filled with various levels of tone (5%, 15%, 30%, etc.  and plot those into your curves. Thats what others do (for the on press) section of control.
  
  
  Having a base tonal adjustment from the default (not curve applied at all), is smart.

[/list]

[ZooCity]

No that's all in the RIP. 

The first is the true linearization at the print head of your output device, i.e. 50%=50% on a densitometer reading.  It's purpose is to be a baseline to start with.  If we changed imaging devices down the road, all I need to do is get the new device linearized at output.

Tone Curves are boosts and cuts applied in the RIP.  They work on top of the initial calibration.
I did these yesterday, working toward a 1-10% that was inflated enough to image at 600dpi the data we want on our screens of choice, an 85-99% that didn't gain out to solid black and a smooth curve in the middle.  I ultimately tuned this in on press, printing black onto a variety of substrates at the same angle/pressure/speed.   So far so good with 225 on through 330/350 mesh showing a very nice curve.  I don't think this part could be done very effectively without being a printer, artist, separator, screen tech and all that.  It really takes all of your knowledge.   

You are right in thinking that there is a "correct" tone curve for every job and substrate. This is their function in our RIP workflow.  We'll default to this curve and can toggle to others as needed.  Sadly for us, we get many runs with wildly mixed garments so we'll never really be able to dial in for a certain fabric type, it'll have to be a happy medium.

Decided to skip intended/actual press adjustments.  Seems like that is better geared toward color printing on flatstock.  Two levels of linearization should suffice for our needs.

What I find very tricky here is avoiding the urge to linearize output to an ideal, i.e., 50% on screen outputs to 50% on substrate with gain.  If I were to do that, none of our prior seps would output correctly, they'd be "airy" as I've heard you say before Dan.  All our seps were made with a lot of nasty, sloppy gain in mind and pulled back accordingly.