Compensating TVI


Ok, now I am totally confused, I read the blogspot but I need it dumbed down a bit. If I have a press that is giving me a constant TVI of +25% in the halftones (50% gives 75% on sheet) and I want to bring it down to 18%, or a 50% that gives me 68% on sheet, why not have a compensation to bring the 50% tone on plate down by 7%?
Assuming the press is stable of course.
 
Ok, now I am totally confused, I read the blogspot but I need it dumbed down a bit. If I have a press that is giving me a constant TVI of +25% in the halftones (50% gives 75% on sheet) and I want to bring it down to 18%, or a 50% that gives me 68% on sheet, why not have a compensation to bring the 50% tone on plate down by 7%?
Assuming the press is stable of course.


The method described on my blog is pretty much standard and used by most vendors' TVI compensating software although the terminology might be different and it may be a two curve workflow vs one. Simply adding or subtracting % will not give you the correct required tone values.

Let's say that when we ask for a 50% in the native file we want to measure the result in the presswork as 68% (18% dot gain/TVI)

In the chart on the blog a 50% tone request in the file resulted in a target 68% tone in the presswork.
But in the new presswork the 50% file request results in 90%
If we did it your way we would subtract 50% from 90% and then subtract that difference from 50% and then map 50% to 10% (90% - 50% = 40%, 50% - 40% = 10% ). But a 10% request in the file isn't going to deliver your target 18% dot gain at a file requested 50%.

On the other hand, looking at that same target 68% you can see that it was delivered in the new presswork from a requested tone value of 30%. So a 50% request in the file needs to be mapped to a 30% tone request. This 30% file request will result in the 50% request in the file delivering a 68% in the presswork which is what the target is.

This all sounds complicated when you read it, but it's very simple once you understand it.

Basically you need:
1 - a target tone (the tone you want to match)
2 - a tone request (the tone specified in the source file (PShop, InDesign etc.)
3 - a press response (the tone that you actually got in the presswork).

With those three elements the software, or you manually (as per the diagram in the post), can create a compensating curve that is applied to plate imaging.

Note that the halftone dot value that's on the plate is NOT involved in the calculation. It's not involved because it is not needed to create the plate curve. All that's required is that the plate imaging is consistent.

This is a one curve workflow. And unless there is a specific need, this is the preferred method.

In a two curve workflow, a curve is applied to the plate to make it linear, then the same 1, 2, 3 steps above are used to create the compensation curve.
 
Last edited:
That method is for aligning the new press sheet to the older sheet or with a target, but if i have printed a new job or I didn't measured my TVI before and I get and measurement device and check my TVI and now I see that it is too much or out of spec. How will I do that?


The method is for whenever you have a target tone.

You need:
1 - a target tone (the tone you want to match)
2 - a tone request (the tone specified in the source file (PShop, InDesign etc.)
3 - a press response (the tone that you actually got in the presswork).

In your example:
but if i have printed a new job or I didn't measured my TVI before and I get and measurement device and check my TVI and now I see that it is too much or out of spec.

You say your response is "out of spec"
So:
1 - the "spec" is your target tone (the tone you want to match)
2 - you know the a tone request you made (the tone specified in the source file (PShop, InDesign etc.)
3 - you know the press response (the tone that you actually got in the presswork).

Since you know those three elements you can build a TVI compensation curve. It doesn't matter that you didn't measure your TVI before.
 
This is getting to be an interesting thread, it would seem that the number of the beast is still a bit of a mystery even to those who work with this stuff.
Totally agree with you on the subject of a one curve solution and I think that I am compensating as described in the blogspot

(as in 1 - the "spec" is your target tone (the tone you want to match)
2 - you know the a tone request you made (the tone specified in the source file (PShop, InDesign etc.)
3 - you know the press response (the tone that you actually got in the presswork)

However this story has gotten a little twist.
I made a new set of test plates, this time I made my own and pretended that the other plates never existed and told the customer that the confusing results were caused by a gypsy curse. The new plates are perfectly liner and for sure without any Gracol, fogra, euroscale etc.
Sure enough the results tended to be dark in the mid tones with everything above 93% dot being totally black.
I measured the color bar and found that the SID`s were a bit too low for the inks used, however measuring the separate CMWK 1% to 100% scale I found that Cyan and Magenta were in more or less in spec, quarter tones showed 12% TVI, the mid tones 20% and the three-quarter tones 14%, nothing that a ICC couldn't handle, Yellow and Black however were a different story altogether, quarter tones 22% mid tones 32% and the three quarters 27% as measured by the trusty Gretag D19C ( I even cross checked that thing at another printers, it is in spec BTW) I have a nagging feeling that there is a underlaying issue on the Y and K towers on the press. What say Gordo?
I have made a new set of plates but now with a round dot instead of the Euclidan that they normally use and brought the ruling down from 175 lpi to 150 lpi.
If I get similar results from these plates then it would in my humble opinion point to the press as being the culprit.
 
I made a new set of test plates, this time I made my own and pretended that the other plates never existed and told the customer that the confusing results were caused by a gypsy curse. The new plates are perfectly liner and for sure without any Gracol, fogra, euroscale etc.
Sure enough the results tended to be dark in the mid tones with everything above 93% dot being totally black.
I measured the color bar and found that the SID`s were a bit too low for the inks used, however measuring the separate CMWK 1% to 100% scale I found that Cyan and Magenta were in more or less in spec, quarter tones showed 12% TVI, the mid tones 20% and the three-quarter tones 14%, nothing that a ICC couldn't handle, Yellow and Black however were a different story altogether, quarter tones 22% mid tones 32% and the three quarters 27% as measured by the trusty Gretag D19C ( I even cross checked that thing at another printers, it is in spec BTW) I have a nagging feeling that there is a underlaying issue on the Y and K towers on the press. What say Gordo?
I have made a new set of plates but now with a round dot instead of the Euclidan that they normally use and brought the ruling down from 175 lpi to 150 lpi.
If I get similar results from these plates then it would in my humble opinion point to the press as being the culprit.

Okay...
If you run linear plates and run the densities to their appropriate SID targets, you are basically printing the same dots on each of the four units. The only difference between the units are the SIDs.
Because halftone screens are bilaterally symmetrical (the 0%-50% dots are the same size as the 50%-100% dots) the only difference is that the 50%-100% dots is that they are the negative of the 0%-50% dots - then you'll get tone curves that will look something like these:

Tone%20curves_zpsnaeaje7g.jpg


However, if your tone curves look something like these:

Bad%20Curves_zps38k1xhbe.jpg


Or these:

TVIs%20graphed_zps0ui7dq0n.jpg


Then something is probably wrong press side.

Using a microscope like this:

Optex_zpsccrezckd.jpg


Can help in diagnosing what the problem is.
 
Last edited:
You mean like this?


Maybe. It's not ideal to look at just one element. You should smooth the curves to get a better idea as to what's going on and use the microscope to check halftone, solid, and over print integrity.
 
Maybe. It's not ideal to look at just one element. You should smooth the curves to get a better idea as to what's going on and use the microscope to check halftone, solid, and over print integrity.
Microscope has been ordered, at the moment all I have is a loupe and a gretag, however come next week I want to run a hundred or so prints on the first tower without engaging the others so that the press can settle then measure again.
However the test prints with 150 lpi instead of 175 and round dot instead of euclidan came out very well indeed, especially in the mid tones, almost but not entirely unlike a good print.
 

PressWise

A 30-day Fix for Managed Chaos

As any print professional knows, printing can be managed chaos. Software that solves multiple problems and provides measurable and monetizable value has a direct impact on the bottom-line.

“We reduced order entry costs by about 40%.” Significant savings in a shop that turns about 500 jobs a month.


Learn how…….

   
Back
Top