Compensating TVI

[Slammer]

Ok, so maybe I am missing a trick here, Customer complaining about too flat midtones, swampy three quarter tones and too high highlights, (like the patrons of a Dutch coffeeshop they are very high) customer is running perfectly linear plates out of the CTP but on press they get over 20 percent TVI, I then used colorflow to create a compensation, the first test prints were really good, clear highlights nice midtones and open up to 98 percent, customer happy. Then, a live job with the same settings... aaaaaand crap, the color bar was nice, the SID´s in spec and the dot gain in check, (50% = 57%, 25% =30%, 80% =82) but the image was flat, flat, flat, pale and washed out, totally yuck, wet rag, you get the picture, it´s not the first time I have done TVI compensation from the press result not by a long shot, generally on print I go for gray balance, nice SID´s and TVI´s that are roughly in the same area where the original has them.
At the moment I am in my cubby hole licking my wounded pride, what glaring obviousness am I missing here???

 

[schenkadere]

Not nearly enough dot gain. At 7% everything will surely seem flat.

 

[buckeye]

I'm with schenkadere. Not nearly enough dot gain. You should be in the 16-18% range in your midtones. What workflow are you using? Is it possible a double compensation curve is being applied somewhere?

 

[gordo]

Ok, so maybe I am missing a trick here, Customer complaining about too flat midtones, swampy three quarter tones and too high highlights, (like the patrons of a Dutch coffeeshop they are very high) customer is running perfectly linear plates out of the CTP but on press they get over 20 percent TVI, I then used colorflow to create a compensation, the first test prints were really good, clear highlights nice midtones and open up to 98 percent, customer happy. Then, a live job with the same settings... aaaaaand crap, the color bar was nice, the SID´s in spec and the dot gain in check, (50% = 57%, 25% =30%, 80% =82) but the image was flat, flat, flat, pale and washed out, totally yuck, wet rag, you get the picture, it´s not the first time I have done TVI compensation from the press result not by a long shot, generally on print I go for gray balance, nice SID´s and TVI´s that are roughly in the same area where the original has them.
At the moment I am in my cubby hole licking my wounded pride, what glaring obviousness am I missing here???

You appear to be wanting the TVI compensation to make the presswork linear i.e. 50% in the file measures 50% on the press sheet. If that is correct Then the comments you received are correct and you are targeting the wrong...target.

 

[Slammer]

You appear to be wanting the TVI compensation to make the presswork linear i.e. 50% in the file measures 50% on the press sheet. If that is correct Then the comments you received are correct and you are targeting the wrong...target.

Hit the target but missed by a mile… You are correct however I was not aiming for 50%=50% but a bit higher, around 10% all the same I am going to tone the compensation down, waaaay down. What bugged me is that the one percent on press came out as 15 percent, 50% was 75% and above 93% darkness reigned, then, after applying the curve the one percent was at 3% and the 98% was 99% 50% was at 57%, so the system works as intended (first time dabbling in Prinergy BTW) if I take it down to compensate for say 10% TVI instead of 25% my 50% should then be at 65% with the 3/4 tones open to almost 100% and the highlights where they should be.
Hopefully
I did a Slur and doubling test on the press with a good enough result so I take it that the blanket and rollers are ok-ish and the ph in the fountain was is spec. later on I saw that the artwork in question had a Fogra 39 profile embedded but I can't imagine that it would make such a huuuuge difference.
Tomorrow I will be making the next set of plates with the toned down curve, may the force be with me.

 

[Slammer]

Ok so how did this pan out…?
Well got the TVI to 16 to 18% and what do you know, amazingly the results were fine.
Good job and a big round of applause.
Everybody is gagging to see the print masterpiece of the next job.
I mean it is a job with pearls and diamonds and rubies, aka a little shop of pastel horrors and the printshop is not where the plates are made, is not the place where the proof came from, is not the prepress, is not the design studio… all are separate.
What can possibly go wrong?
Same settings and the prints are too dark, that's what can go wrong.
Curses, curses, curses!!
Back to the drawing board.
Ah well it's something to keep my mind at work for the weekend

 

[Slammer]

If I could have 0% dot gain coming off press, I'd take that any day. It's always been my understanding that since every press prints differently (some VERY differently), plate curves can never be standardized. Dot gain is present on every press but it's my belief that many in this industry misunderstand the true meaning of that, the amount that's actually occurring, and how much compensation is necessary.

I'm not implying anything about your current dilemma, just posting my (moderate) experience related opinion on the subject.

Amen to that, I had aimed for a Dot gain of 10% around the mid tones, but there are so many things that can influence dot gain, i.e. baked or unbaked plates, the inks used, in this case they come from a supplier that gets them from a supplier somewhere on the mainland, black that is mixed from left-overs, then Xinggraphics plates, not too sure about them, when I have measured them a few days they seem to read high. But what I have just found out this weekend is that the Gretag D19C I use seems to give me a reading that is 10% too high when I cross reference it with a Ultradottie. (FYI, just started with this company and it's the first "live" gig with a customer)

 

[kansasquaker]

I then used colorflow to create a compensation

Since you have access to colorflow, have you considered aiming for the appropriate industry specification for the work and creating a DLP to make the press match? Assuming the press is in decent condition, you should be able to come close to Gracol, SWOP or whatever is deemed appropriate. That would also solve your proof issue - just have whoever makes the proof output to the same specification.

 

[gordo]

Just trowing a wrench into the monkey works...
Dot gain/TVI is not a target, it is a process control metric.
Tone reproduction/tone curve is a target.

 

[Slammer]

In the end this is where I want to go with this customer, they have a unused CTP and all the goodies from Kodak in house but it is gathering dust in a garage at the moment.
Problem is that the plates are made in one company and the proofs are made in another and unless the customer gets everything together under one roof it is going to be impossible to set any kind of standard.
But I am working on it, seems they have a open ear to change.
At Gordo, sorry if I made the impression that I regard a TVI is as a target, it is of course a process control method.

 

[Deleted member 16349]

Dot gain/TVI is not a target, it is a process control metric.

Tone reproduction/tone curve is a target.

Hi Gordon,

I find this comment a little confusing. Can you clarify what you mean?

Thanks.

 

[kansasquaker]

Problem is that the plates are made in one company and the proofs are made in another . . .

I'm sorry but I don't see why this is an impossible problem. Inconvenient? Yes - Impossible? - no. It's not materially different than profiling any press. Just align the press with a target, and align the proof with a target.

I assume the printer is sending 1 bit tiffs to the plate company for output. Assuming the 3rd party platesetter is a stable device, you have complete control over the dot on the plate. It doesn't even matter if their device is linear, just be sure that they use the same configuration when they output the calibration plates as they will when they make production plates. I assume whoever is making the inkjet proofs can hit an industry standard with that inkjet (if not, they're proofs are probably meaningless anyway).

So pick Gracol, or whatever, and tell the proof people to make that their target, then create a profile using color flow with Gracol as the target to align the press.

 

[gordo]

Hi Gordon,

I find this comment a little confusing. Can you clarify what you mean?

Thanks.

If dot gain - say 18% - was a target then switching from 100 lpi to 175 lpi to FM screening or one press to another would result in printing that had different tonal reproduction. It would look darker or lighter, more ore less contrasty according to the parameter that changed.

However, if tone reproduction is the target, then switching from 100 lpi to 175 lpi to FM screening or one press to another will result in printing that has similar tonal reproduction. It would look basically the same as the parameters change.

Historically dot gain is used as a process control metric. Whatever dot gain you have it is used to monitor how it is affected/changed during the manufacturing process from file to press.

 

[Deleted member 16349]

If dot gain - say 18% - was a target then switching from 100 lpi to 175 lpi to FM screening or one press to another would result in printing that had different tonal reproduction. It would look darker or lighter, more ore less contrasty according to the parameter that changed.

However, if tone reproduction is the target, then switching from 100 lpi to 175 lpi to FM screening or one press to another will result in printing that has similar tonal reproduction. It would look basically the same as the parameters change.

Historically dot gain is used as a process control metric. Whatever dot gain you have it is used to monitor how it is affected/changed during the manufacturing process from file to press.

I must be missing something here. I thought one tried to obtain a tonal match by adjusting the files so the final TVI's were the same for each printing condition such as you mention above. They would have the same target TVI values.

How would one define a tone if not by the TVI value?

Maybe I am missing the context of your comments. Thanks for any further explanation.

Edit

OK, dot gain is not TVI. TVI is the apparent dot size, which is what I think is used to determine tone. So the same TVI could be the target for all printing conditions but not the dot gain, which would be different in order to get the same tone.

I hope that is basically what you mean.

 

[Slammer]

I'm sorry but I don't see why this is an impossible problem. Inconvenient? Yes - Impossible? - no. It's not materially different than profiling any press. Just align the press with a target, and align the proof with a target.

I assume the printer is sending 1 bit tiffs to the plate company for output. Assuming the 3rd party platesetter is a stable device, you have complete control over the dot on the plate. It doesn't even matter if their device is linear, just be sure that they use the same configuration when they output the calibration plates as they will when they make production plates. I assume whoever is making the inkjet proofs can hit an industry standard with that inkjet (if not, they're proofs are probably meaningless anyway).

So pick Gracol, or whatever, and tell the proof people to make that their target, then create a profile using color flow with Gracol as the target to align the press.

You are correct, it is not impossible but in this case highly improbable as the different companies do not talk to one another, the proofs are as you say meaningless as the print shop has no control over them and the proof shop simply prints colorful pictures, they are pretty though, hit a industry standard? They could not hit a barn door this way.
The studio sends PDF's to the plate maker, who makes the plates (duh) and sends the plates to the printer, in the meantime the "proofs" (note the """ marks) land at the printer who may or may not open the package and use them, of course all print setting standards done this way are by Eyeball version 1.1

 

[gordo]

[SNIP] of course all print setting standards done this way are by Eyeball version 1.1

Dang! Version 1.1? I must have missed the update, I'm still on Eyeball v 0.9.

 

[gordo]

I must be missing something here. I thought one tried to obtain a tonal match by adjusting the files so the final TVI's were the same for each printing condition such as you mention above. They would have the same target TVI values.

Wouldn't you need a different profile for each print condition and end up with different files? I think this would be unmanageable.

How would one define a tone if not by the TVI value?

Source file request/specified on press response. I.e. when I ask for 50% in my file I expect to measure 68% on the press sheet irrespective of halftone screening or press. I.e. there'll be variable dot gains but the end tone will be constant.

OK, dot gain is not TVI. TVI is the apparent dot size, which is what I think is used to determine tone. So the same TVI could be the target for all printing conditions but not the dot gain, which would be different in order to get the same tone.

Dot gain is TVI (Tone Value Increase). It's just that, with different processes, there aren't always dots. So the term was changed from "dot gain" to "TVI" to more clearly indicate what is happening.

 

[Deleted member 16349]

Wouldn't you need a different profile for each print condition and end up with different files? I think this would be unmanageable.[/SIZE]



Source file request/specified on press response. I.e. when I ask for 50% in my file I expect to measure 68% on the press sheet irrespective of halftone screening or press. I.e. there'll be variable dot gains but the end tone will be constant.



Dot gain is TVI (Tone Value Increase). It's just that, with different processes, there aren't always dots. So the term was changed from "dot gain" to "TVI" to more clearly indicate what is happening.

Oh boy, I am mixing up in my head the "apparent dot size" (murray davies eq.) with TVI. Of course now I see that TVI is dot gain but the tone that you want to get the same is the "apparent dot size" (such as that 68%) of the different print conditions.

I haven't been thinking much about these issues lately and it shows. Thanks for helping to clear this up. At least I hope so.

 

[aqazi81]

gordo. If a 50% patch on plate prints let's say 72% on press and we want to bring that TVI to 18%, do we decrease that 50% patch to 46%?

 

[gordo]

gordo. If a 50% patch on plate prints let's say 72% on press and we want to bring that TVI to 18%, do we decrease that 50% patch to 46%?

No. The correct method is described here: http://the-print-guide.blogspot.ca/2010/02/principle-of-dot-gain-compensation.html