Xitron Sierra Screen Angles

[phoenixrje]

Hello. We recently switched our workflow from Rampage to Sierra. We are getting screen patterns on the printed sheet when printing 4-color, noticeable in highlight and midtone areas. Faces are especially noticeable. I was told by Xitron that the screen angles cannot be changed other than toggling between colors. I'm surprised that angles cannot be manually changed. Has anyone else run into this issue? Thanks.

 

[Chasfinch]

missing rampage

 

[gordo]

Hello. We recently switched our workflow from Rampage to Sierra. We are getting screen patterns on the printed sheet when printing 4-color, noticeable in highlight and midtone areas. Faces are especially noticeable. I was told by Xitron that the screen angles cannot be changed other than toggling between colors. I'm surprised that angles cannot be manually changed. Has anyone else run into this issue? Thanks.

What specifically are the sceen angles (and lpi/dpi)?

 

[phoenixrje]

Cyan 165; Magenta 45; Yellow 90 Black 105
175 line screen, 2540 dpi
I just realized that when Sierra was set up the black and magenta were switched. The default is Magenta 105 and Black 45. I am going to give that a try on a job and see what happens.

 

[gordo]

Cyan 165; Magenta 45; Yellow 90 Black 105
175 line screen, 2540 dpi
I just realized that when Sierra was set up the black and magenta were switched. The default is Magenta 105 and Black 45. I am going to give that a try on a job and see what happens.

OK, you say that the moiré is happening in light areas of skin colors. You need to determine which colors are clashing and causing the moiré.

Keep in mind that when two colors are less than 30º apart there is a risk of moiré. In your case that would be K and Y which are just 15° apart. Also, normally the Y lpi is run at an lpi about 108% to minimize the visibility of the moiré caused by being just 15° away from its neighbour color (typically C or M - but K in your example). I.e. if your using 175 lpi your Y is running at 189 lpi. Y being very light means the moiré it causes should be very light and hence less visible. However, if the Y gets contaminated it goes darker and the moiré becomes more visible. With K just 15° away from Y it's quite possible that your Y ink is being contaminated by the K printer causing the moiré to become visible. Check the Lab values of your Y ink to see if that's happening.

The least visible color, Yellow, is usually placed at the most visible angle 0° (90°). Then the most visible color, Black, is placed at 45° since its screen pattern is very noticeable when positioned at 0° and is least visible when rotated 45° (that's a human vision issue). The cyan and magenta are then placed between those two. Cyan at 15° (105°) and magenta at 75°. These angles represent a best all around compromise for most pictures and represent the standard, most commonly used screen angles. They also form the least objectionable moiré – the rosette pattern.

The standard CMYK U.S. screen angle set is:
15, 75 (same angle as 165) , 0, 45 (that causes a possible moiré in greens since C and Y are only 15º apart)

Standard CMYK European screen angle set is:
15, 45, 0, 75 (same angle as 165) (possible moiré in greens since C and Y are only 15º apart)

If you run M at 105° you will get an M/Y moiré as they are only 15° apart. That could be a problem with skin colors which typically have a lot of Y and M. I would suggest running the C at 105° and the M at 75°. That will mean that C and Y are just 15° apart and therefore cause a moiré in the greens - however that's not usually a problem because greens are typical of subjects like grass and tree leaves which, unlike skin tones, don't have an area large enough to reveal the C/Y moiré.

Another possible (but not common) cause of moiré is called single channel moiré. It's highly unlikely to appear with a modern screen - but it is possible. Single channel moiré is a moiré within one color. It can happen when the LPI (e.g. 175 lpi) is not an even divisor of the dpi of the RIP (e.g. 2540 dpi). In that situation each halftone dot can have an extra pixel, or not, as the RIP tries to reconcile not being able to image a half pixel. I.e. 175/2540 = 14.5 the RIP can't image a .5 pixel so it averages out that .5 by having one halftone dot have an extra pixel while the next halftone dot doesn't. If those extra pixels repeat with a fixed frequency then they can harmonically beat with the screen and you'll get a moiré pattern in that one color - usually Black because it's the most visible color. The RIP vendor's fix is to offset the screens by 7.5° which introduces noise into the screen and eliminates the repeating pattern and hence the moiré.



I hope all this makes some kind of sense.

 

[semi-tech]

If the original file has embedded screen angles (perhaps set by the artist in Photoshop), couldn't that create the same issue once the Sierra gets the job?

 

[gordo]

If the original file has embedded screen angles (perhaps set by the artist in Photoshop), couldn't that create the same issue once the Sierra gets the job?

Some RIPs (wisely) ignore embedded screen angles. I don't know about the Sierra. One would have to do a simple test - e.g. create a greyscale patch of tone and embed an arbitrary screen angle of, say 20°, but set the RIP for K to screen at 45°. If the file outputs at 45° then the RIP has ignored the embedded screen angle. If the file outputs the K at 20° then the RIP has honored the embedded screen angle. Then, unless there is a very good (unusual) reason to honor embedded screen angles you need to resave the PShop images without embedding angles. Or if the RIP can be set to ignore embedded screen angles then that is how the RIP should be set.

Very complex algorithms are used by RIPs to calculate screen angles and frequencies. It is no easy task. The RIP will deliver its best compromise approximation to what is requested, however, to get the best result it is seldom exactly what is requested as far as lpi and angle is concerned. For example, a 133 lpi request for Agfa Balanced screening will result in a 138.158 lpi for the Black printer at 2540 dpi and 138.142 lpi for the M and C screens. I.e. you ask for 133 lpi but that's not what you get because actually giving you 133 lpi would result in a less than optimal final result. Each RIP vendor (e.g. Heidelberg, Kodak, etc) will have their own "secret sauce" to deal with the issues related to digital screening so they will all give you a slightly different result from the same request.

 

[buckeye]

We too moved from Rampage to Sierra. We are using the "Flexo angle family" screen angles of Black@113, Cyan@173, Magenta@53 and Yellow@98 and have not in a year's time of using Sierra had any moire issues.

 

[phoenixrje]

Thanks for your help, gordo and semi-tech. There are no screen angles embedded in the art, so I know that is not the issue. We seemed to get better results moving the magenta to 105 and black to 45, the default settings in Sierra. My bigger concern with Sierra is that there is no way to manually set screen angles, which I am wondering if any other Sierra users were surprised by this. With Rampage, you could set the colors to any angle you pleased. I'm sure I am not the first Sierra user to question this.

 

[phoenixrje]

Buckeye, what line screen are you running, and what DPI? Are you running sheetfed offset presses? Thanks in advance.

 

[gordo]

We too moved from Rampage to Sierra. We are using the "Flexo angle family" screen angles of Black@113, Cyan@173, Magenta@53 and Yellow@98 and have not in a year's time of using Sierra had any moire issues.

Halftone screens are quadratically symmetrical. What that means is that, for example, 0°, 90°, 180°, and 270° are all the same screen angle.

Your "flexo" angles would be about a 7° rotation from the base screen angles. That rotation is needed in flexo because of the anilox roller (and is used by some RIPs to eliminate, as mentioned before, single channel moiré)

So, if you factor out that 7° in the flexo screen set - your base screen angles are:

C=75°
M=45°
Y=0°
K=15°

Although the relationship of the angles is correct (30° offsets for 3 colors and 15° for 1) and is basically the same angle set the OP uses, that would not, IMHO be an optimal screen angle set.

 

[buckeye]

rjeggs,

We are an offset shop running 175 line at 2400 dpi.

 

[phoenixrje]

Any particular reason you moved to the flexo angles from the angle family (2) set?

 

[semi-tech]

Some RIPs (wisely) ignore embedded screen angles. I don't know about the Sierra. One would have to do a simple test - e.g. create a greyscale patch of tone and embed an arbitrary screen angle of, say 20°, but set the RIP for K to screen at 45°. If the file outputs at 45° then the RIP has ignored the embedded screen angle. If the file outputs the K at 20° then the RIP has honored the embedded screen angle. Then, unless there is a very good (unusual) reason to honor embedded screen angles you need to resave the PShop images without embedding angles. Or if the RIP can be set to ignore embedded screen angles then that is how the RIP should be set.

Very complex algorithms are used by RIPs to calculate screen angles and frequencies. It is no easy task. The RIP will deliver its best compromise approximation to what is requested, however, to get the best result it is seldom exactly what is requested as far as lpi and angle is concerned. For example, a 133 lpi request for Agfa Balanced screening will result in a 138.158 lpi for the Black printer at 2540 dpi and 138.142 lpi for the M and C screens. I.e. you ask for 133 lpi but that's not what you get because actually giving you 133 lpi would result in a less than optimal final result. Each RIP vendor (e.g. Heidelberg, Kodak, etc) will have their own "secret sauce" to deal with the issues related to digital screening so they will all give you a slightly different result from the same request.

We were getting screen angle issues with one client in particular, and only when using our digital press; nothing in-house. The issue seems to have disappeared, but that doesn't mean it won't re-appear. Thanks for the testing idea.