Weird lines in presswork - any ideas as to cause?

[gordo]

A printer in the UK contacted me with a problem with lines appearing in flat tint screen builds. I can't figure out the cause. Any ideas?
Some details: 150lpi AM Elliptical dot shape 2400 dpi CtP. Heidelberg Web 8 Press. The banding marks which you can see in the attached photo appear laterally across the press – Gear Side to Operator side. They are approximately 2-3mm wide. Plates have been imaged in both Landscape and Portrait and the banding stays in the same position on press. Plates tested: Kodak Sword Excel and Kodak XD imaged on a Trendsetter 3244F, Quntum 800 and Magnus VLF – still the same problem.

Plates imaged on an Agfa CtP run on this press do not have the banding problem.

Most artifacts like this that I'm aware of occur in the press direction - not across the web. If it is CtP imaging related you'd think the banding would change direction when imaging was switched from portrait to landscape - but in this case it doesn't, which suggests a press problem. But the same plates imaged on an Agfa CtP don't have the problem - which suggests a CtP problem.

Any ideas?

thx, gordon p

 

[RGPW17100]

We recently switched to the new azura plate and had similar lines. Plate reader read plates as 99.7 on 100 patch and when put on press it had faint lines on the solids and the neutral dots looked bad. Laser was re focused and the problem went away. We never tried to turn the image or the plate 90 degrees. Focus was close but had a small adjustment and the problem went away.

 

[gordo]

We recently switched to the new azura plate and had similar lines. Plate reader read plates as 99.7 on 100 patch and when put on press it had faint lines on the solids and the neutral dots looked bad. Laser was re focused and the problem went away. We never tried to turn the image or the plate 90 degrees. Focus was close but had a small adjustment and the problem went away.

If it's plate banding, which is caused by the laser swath not having an even energy profile across its width then, in my experience, it appears in the same direction as the travel of the imaging laser. So if you switch from landscape to portrait exposure the banding will appear in presswork in line with and then across the direction of travel.

But that's not what 's happening in this case. Whether landscape or portrait exposure the banding in the presswork remains in the same direction. That's what confuses me.

gordon p

 

[vik la luna]

the laser head needs to be replaced
had the same problem years ago and this was the only solution that worked after exhausting all poossible causes , including press chemistry rollers etc
creo replaced the head and all was well again ,Mr. Pritchard

 

[seejay]

HI Gordo,

Were the different kodak plates imaged on the same unit?
I've seen something very similar to this on two occasions at two different sites, coincidentally with Kodak plates.
Both issues were caused by the 'grain direction' on the plate (either the grain direction of the aluminium or the direction the plate was coated - cannot remember which was the cause sorry). both issues were fixed by an adjustment/replacement laser. My suggestion would be for the printer to request a 'batch' of plates with the grain running in the opposite direction to see if it goes away. If this is the same problem, then the imperfection in the plate is highlighting the problem with the laser unit.

Incidentally, when I experienced the issue on the second occasion, the UK ctp specialist claimed that it was impossible that the grain direction of the plate would cause this issue - when it was bounced back to R&D in Germany they confirmed that grain direction would highlight the laser problem!

 

[rich apollo]

Gordo, just to clarify, they imaged the plates in both directions? Highly unusual to be able to do that.

Were they run through the processor in the same direction? Have they checked all the processor roller settings? My experience with the Sword Excel is that the processor performance/condition cannot be overstated.

Looking it up, it appears that this is a narrow web device. The cutoff is longer than the web is wide, so imaging problems could show up across the web.

 

[Lukas Engqvist]

I'm with rich on this. Not seen it on plate, and maybe you already ruled it out but looks like a roller that has a bad cogg, this could be in the plate processor, or could be blankets that need re tightening, sounds too basic to have been missed (also all rollers are parallel?)… but I'm willing to look the fool. The press is normally out of my territory and I ask my pressmen silly questions, some times totally off but it has happened when they try to explain to me why it can't be that they stumble across an answer in their own explanation.
One more silly question, that is at a certain raster right? did you try different angle / screening?

 

[gordo]

Here is all the info I have received. It's a poser.:


Description of Web 8 Trials done so far:

1st Trial - Test run on the - 18/05/09

> 1st print run was using Agfa plates imaged at outside supplier.

No Banding on Printed copy

> 2nd Print run was run using Sword Excel plates – imaged as normal, on the manual plate setter. Plate imaged landscape, ran using the following Prinergy settings, Linear, AM Screening, Creosettes CMYK 75-45-90-15, EllipicalP, 2400dpi.

Banding appeared on Printed copy across the web

> 3rd Print run was run using Agfa plates on the Upper unit and running Kodak Sword Plates (Imaged as above) on the lower unit.

Banding appeared on Printed copy only on the Lower unit across the web (Using Sword Plates)

Conclusion of the results from these press trails was that by running a different plate corrected the error on press therefore the issue must reside with either the plate itself or the imaging device – please note no other change occurred on Press except changing the plates from Agfa to Kodak etc.

2nd - Trial on the 19.05.09

> 1st Print run was using the Manual plate setter with the plate imaged Portriat this time (Opposite direction to how we normally image plates) and using the same settings, Linear, AM Screening, Creosettes CMYK 75-45-90-15, EllipicalP, 2400dpi.

Banding appeared on Printed copy across the web opposite direction to imaging.

> 2nd Print run was using the Quantum 800 plate setter with the plate imaged Portriat this time (Opposite direction to how we normally image plates) and using the same settings, Linear, AM Screening, Creosettes CMYK 75-45-90-15, EllipicalP, 2400dpi.

Banding appeared on Printed copy across the web opposite direction to imaging.

Conclusion of these results was that the only constant here was the plate itself as we had banding on press but in the opposite direction to how we image therefore the issue must reside with the plate itself and so we asked for different plates to be tested on the Web 8 from Kodak.



3rd – Trial on the 21.05.09

> 1st Print run was using Sword plates – imaged as normal, on the manual plate setter. Plate imaged landscape, ran using the following setting, Linear, AM Screening, Creosettes CMYK 75-45-90-15, EllipicalP, 2400dpi.

Banding appeared on Printed copy across the web

> 2nd Print was run using Kodak XD plates which will were imaged at another printshop using our settings from Prinergy which were plate run Linear, AM Screening, Creosettes CMYK 75-45-90-15, EllipicalP, 2400dpi.

No Banding on Printed copy

Conclusion of results was that by changing the plates we resolved the problem, as we had run Agfa Plates and now Kodak XD plates and the problem was solved. So we came to the natural conclusion that the issue was the plates.


4th - Trial on the 03.06.09

>1st print run was using Kodak XD plates imaged on the Magnus VLF Plate setter – workflow slightly different as we write a 1 bit Tiff for the VLF but all Prinergy settings are the same plate run Linear, AM Screening, Creosettes CMYK 75-45-90-15, EllipicalP, 2400dpi.

No Banding on Printed copy

> 2nd Print run was using Kodak XD plates imaged on the manual Plate setter – the Plate was imaged landscape, ran using the following setting from Prinergy , Linear, AM Screening, Creosettes CMYK 75-45-90-15, EllipicalP, 2400dpi.

Banding appeared on Printed copy across the web

> 3rd Print run was using Sword Excel Plates imaged on the Magnus VLF Plate setter – workflow slightly different as we write a 1 bit Tiff for the VLF but all Prinergy settings are the same plate run Linear, AM Screening, Creosettes CMYK 75-45-90-15, EllipicalP, 2400dpi.

No Banding on Printed copy

Our conclusion was that the issue must reside with the plate setter itself, because we take the same plate and data and imaged it on the Magnus using either XD or Sword plates the issue is resolved on Press.
Please note the only variable we changed during these trials was the plates all other press consumables were consistent.


best, gordon p

 

[arkay_desai]

Hi gordon,

Just for clarification, can you please provide me processed plate thinkness and the process (Manual plates or CTP)

Regards
Arkay

 

[apepp]

...i've seen something similar to this before back in the days of the creo trendsetter late 90's, the solution then was the lens needed cleaning by an engineer due to dust. I have also seen this with faulty plate batches from Agfa imaged on a Palladio system...

 

[DavidMa]

Were there any test plates made to verify that the exposure was correct for each plate manufacturer?
Manual and CTP?

 

[colinobrien]

Hi Gordon,

I have seen this with Screen CTP units, it looks like you have either a power issue getting to your laser head, or your manual CTP unit is out of focus.

Colin

 

[gordo]

If exposure or focus or energy swath are incorrect then I would expect to see the lines imaged in the direction of the CtP drum rotation which would result in the lines running in the same direction as the web through the press rather than across the web which is what's happening. Also if the plate is rotated and imaged then the lines would then be rotated 90 degrees and then run across the web. Exposed portrait or landscape the lines only run across the wed.

gordon p

 

[TechDoc]

If I understand it right : The lines are parallel to the punch holes on the plate ? Means horizontal lines ?
Reason for Vertical Lines (normally FastScan direction of the CTP) can be dust on the Optic, Focus, Zoom ... etc..
But horizontal lines (SlowScan in the engine) is nearly always PixelData missing in FastScan..
Imaging a plate in port-or-landscape doesnt change FastScan and SlowScan

 

[rich apollo]

If it was a problem with the laser, then the lines would change direction depending on the orientation that the plate was imaged.

It does look A LOT like to little power getting to the plate - as you would see with dirty optics. However, rotating the plate would change the direction.

I'm still very skeptical that the plate was imaged in portrait orientation. I'm from the Show-Me state - I'd have to see it done.

 

[gordo]

Thanks everyone for your ideas. I've been passing them on to the printer in the UK, but have not gotten a response from them.

Your help is much appreciated.

gordon p

 

[Lukas Engqvist]

1) One question aboout the rotation. Do you rotat the plates in the developer? Could that be tested?
2) Is the defect visible on the plate? (may be more visable inked in) Would help to know if it is before press or in press?
3) When viewing the print under a loupe. is there smearing of the screen? (trying to determine if it is keinetic/mechanical deformation, chemical inconsistency in the process or digital aproximation)

I still feel thet ther is not a clarity of where in the process the fault happens.

 

[maas]

Is alcohol used on their press?, does the ink look emulsified in between the dark bands?
I have seen this before on a Roland sheetfed press and corrected it by increasing the alcohol (by only a fraction)

Check blanket condition for build up or piling when the press is pulled up, any marks on the rollers that are in contact with the plate when under impression?

Maybe i missed this but are the lines visible on the developed plate? if so do they continue around the plate bend? are any lines in non image areas on the plate?


Rgeards
Maas

 

[me.awash]

I have same problem. In my case, wierd thin vertical lines appear on the left side of the plates as you can see in the pictures I uploaded. We have been overdeveloping the plates to get rid of those lines. When developing, those lines are not visible with naked eyes but soon we start printing, those lines appear right in ourfaces. I tried to underdevelop the plates to observe the pattern of those lines and got to see them clearly as I have uploaded the picture of the plate too. Left pic is the plate and right pic is the result after printing. Didnt have any issue earlier but recently we have been bothered with this problem. Any solution from you guys would be very greatful.

 

[Marc.Ram]

Hi me.awash,

What CTP imager are you using?

Does it support calibrating the laser diodes individually?

What is the width of the space between the lines?

Do you see this problem on smaller plates?

These types of vertical lines look like the width of the swath is not perfectly aligned. It could also be that the laser diodes are not all firing at the same intensity.

Because you only get them on one side of the plate may mean that your 'Head' (Carriage Optic if its a fiber system) may be slightly further away from the drum on one side than the other.

You may be able to solve/prevent the symptom by redoing a focus calibration that brings the imaging to within the tolerance of the depth of field of the optics for both the left and right sides of the plate.

Marc