Standard Line Screen

[Cory Smith]

What line screen do you run? do you run a different line screen based on what paper the job is being printed on?

we run a standard 200 here, and just recently started running 150 on some jobs, depending on paper.

 

[umesh]

Re: Standard Line Screen

Hi Smith,
In my opinion its very important to see the type of paper and its surface before printing, to decide the LPI to be used.
As a thumb rule we can say that, "For a rough paper like NEWS PRINT one should use coarser screen, While running a smooth surface paper like ART PAPER one should go for finer screen rulling".
I would suggest
75 to 85 LPI for NEWS PRINT PAPER
133 to 150 LPI for MAPLITHO
175 to 200 LPI for ART PAPER

Reason is, if we use finer LPI like 150 on news print paper, the dots of lower percentage like 5 to 10 will not come on the paper, on the other hand if we use coarser screen like 75 LPI for ART paper, the white gap between the dots will become visible due to the smooth surface of the paper.
I hope this is right....

Edited by: umesh on Sep 21, 2007 2:08 PM

 

[[email protected]]

Re: Standard Line Screen

We run 175 lpi for uncoated stock and 240 lpi for coated stock.

Abe Hayhurst
Director of Color and Technology
We Do Graphics, Inc.

 

[tony_mclachlan]

Re: Standard Line Screen

Hi, we currently run 175 for most of our work, 150 for some jobs also depending on stock. We just recently installed an Agfa Acento so we are looking to experiment with 200lpi Sublima screening, hopefully it will produce good results and we can move up to that for our higher end colour work...

Cheers, Tony

 

[Fatboysmart]

Re: Standard Line Screen

We run 175lpi on uncoated stock and 200lpi on coated. We now use Hybrid Screening (Heidelberg) which is great!

 

[Gerhard]

Re: Standard Line Screen

HI,
it's nice seeing that you all use different line screens to improve you printing quality. Please with all this keep in mind that lpi (Lines per Inch) and ppi (Pixel per Inch) goes hand in hand.

For 150 lpi you need your pictures to be saved in 300 ppi. For 200 lpi you need your pictures in 400 ppi. If you use 200 lpi with 300 ppi pictures you just have a finer screen but not a better picture quality because you simply do not have enough pixels available for this kind of fine screen.
You can use this small chart:
ppi lpi
150 75
160 80
240 120
300 150
400 200
480 240
The problem is that a picture saved in 400 ppi is much much bigger than a 300 ppi one.
If you want really to improve your printing having only 300ppi pictures you need using FM screening with about 20 µ.

Best regards,
Gerhard

 

[pcmodem]

Re: Standard Line Screen

The Pixels Per Inch (ppi) to the Lines Per Inch (lpi) used to be a 2:1 ratio like what Gerhard said. But you can get the same quality with a 1.5:1 ratio.
So the chat would be something like this:
ppi.......lpi
112.5..75
120.....80
180.....120
199.5..133
225.....150
262.5..175
300.....200
337.5..225
360.....240
375.....250

 

[zoran]

Re: Standard Line Screen

Since Gerhardt mentioned 20micron stochastic, does anybody use 10 micron?
Do you have different setups for coated and uncoated?
What would be good "default" for different papers, like 10 micron coated and 25 micron uncoated?

Does it make sense going smaller than 25 micron on uncoated paper?

We are commercial shop and we use 10 micron extensively. So far we did not have special rules for coated vs uncoated but now we are looking into it so I figured I'll ask for opinions.

Thanks

Zoran

 

[disbellj]

Re: Standard Line Screen

From the old days (I figure it still rings true), the smallest resolution you can get without getting pixelazation is 1.2 pixels per printing dot. So if your linescreen is 175, you'll need at least 210 dpi (175 x 1.2 = 210) to not get pixelization (blurry images where you can see the stair step).

Don

 

[disbellj]

Re: Standard Line Screen

Of course, 300 dpi is what we recommend, but just saying the least dpi we can get away with without seeing the "stair-step".

Don

 

[umesh]

Re: Standard Line Screen

Hi everyone,
this thread is really becoming interesting,starting with LPI we discussed about the type of paper and then image resolution(pixels), now one more important thing should be considerd, ie. Imaging resolution or dpi used by the CTP for exposing.
Not much attention is given to this part, because most of the RIP's ask you to enter only the required LPI and then the corresponding RESOLUTION or dpi is automatically choosen by the RIP.Basic rule says that there should be atleast 256 GRAY LEVELS for the image to looks smooth, ie. without any BANDING or STEPS.
The number of steps for a particular combination of LPI and DPI can be easily found out using

*Gray levels= (DPI/LPI)^2^*

Why i'm telling this ???
because just shooting the file for higher LPI, will not give good results, if the imaging resolution of the exposing device is not high.

i dont know how it works when FM screening is used, if any one knows please share it ..

 

[disbellj]

Re: Standard Line Screen

umesh,

Guess I should be using 150 ls for both coated and uncoated then? I have images at 300 dpi (most comes in that dpi) and platesetter resolution 2400 dpi (set to 2400 instead of 2540 because of problems encountered that were solved by making the LW and CT evenly divisible), and have been using 175 ls for coated and 150 ls for uncoated. So using the math you provided, 2400/150=16x16=256 levels of gray, so 150 ls is fine for uncoated. But using the same math, 2400/175=13.714285x13.714285=188.08161 levels of gray, so looks like I need to make both coated and uncoated 150 (because I'll still get mostly 300 dpi images and need the ratio of LW/CT res to be an even whole number, which would necessitate 2400). What do you think I should do?

Don

 

[gordo]

Re: Standard Line Screen

Hi,

The formula you quoted: Gray levels= (DPI/LPI)2 only applies to very basic halftone AM screening as might be used in B&W low resolution laser printers. Most modern halftone screening (since 1990 at least) use a form of "supercell" screening. This gets around the limitations implied by the gray levels formula. In simplest terms, it does not matter what gray level a specific halftone dot represents - what matters is the grey level of an area of dots. I.e. If you cannot create an individual 11% dot you can create the effect of an 11% dot area by alternating 10% and 12% dots. Bottom line, modern AM screens that use supercell algorithims do not have the grey scale limitations that the formula suggests they should have.

best, gordo

 

[Paul Sherfield]

Re: Standard Line Screen

Hi

Using differing screen rulings for differing papers is just an very basic method of process control for TVI. Running 150 screen for uncoated and 200 for coated will reduce the TVI on the uncoated to levels close to a 200 line screen from common files. But if this is not measured and controlled....

This is not the best, or the correct way. I have clients running 300 screen and FM on uncoated papers. By using process control to measure TVI and using ISO 12647/2 based profiles, the results are fantasic.

Regards

Paul Sherfield
The Missing Horse Consultancy

 

[umesh]

Re: Standard Line Screen

Hi Mr.Pritchard,

Agreed that this is a very basic formula, and now super cell technology is used.But i guess supercells means grouping of number of cells for obtaining accurate screen angles and screen frequency.
And another thing in offset, variable densities cannot be printed, so number of _gray levels per pixel_ has to be maximum of 2 while imaging the plate, as against the case of digital printers, where number of gray levels per pixel are more than 2,so even low imaging resolution will do.
I may be wrong sir but if possible please put some light on this part... this may also solve Don's query..

Regards
Umesh

Edited by: umesh on Sep 25, 2007 2:31 PM

 

[gordo]

Re: Standard Line Screen

Hi Umesh,

RE:
"Agreed that this is a very basic formula, and now super cell technology is used.But i guess supercells means grouping of number of cells for obtaining accurate screen angles and screen frequency."
Basically correct.

RE: "And another thing in offset, variable densities cannot be printed, so number of gray levels per pixel has to be maximum of 2 while imaging the plate, as against the case of digital printers, where number of gray levels per pixel are more than 2,so even low imaging resolution will do.
I may be wrong sir but if possible please put some light on this part... this may also solve Don's query.."

I think Don was referring to the resolution of input images.
Modern supercell screening used for CTP and computer to film eliminate the gray level limitations of old style Rational Tangent screens.
There is an old resolution formula used for scans that will be used for output. Typically it is expressed as the resolution should be 2x the halftone frequency. I.e. you should have a 300 dpi scan for output at 150 lpi. It's a nice formula that in reality is meaningless. In reality, you can actually go as low as 1 to 1. I .e. you can use a 150 dpi image and output it at 150 lpi without any loss of image intergity - "staircasing". Best thing to do is run the same image, at different resolutions, out to film or plate and run on press and see for yourself if you can see a difference. I.e. use an image at 75, 100, 150, 200, 300, and 400 dpi output at 175 lpi and run on press. See for yourself what the impact of image resolution has, or does not have, on reproduction quality.

gordo

 

[disbellj]

Re: Standard Line Screen

Gordo,

I was first talking about resolution and the stair-step effect (pixelization), and said that back in the day we went by 1.2 pixels per printing dot and anything lower would show stair-step effect. So if my l.s. is 175, then the lowest image resolution I need is 175 x 1.2 = 210 dpi. Anything lower and I can expect pixelization.

Then, umesh gave a formula to calculate if one is getting 256 levels of gray when printing. The formula was:
levels of gray = platesetter resolution divided by linescreen and the result squared

So I used the formula, and 2400/150 = 16, and 16 squared is 256, so for my uncoated (where I use 150 l.s.) has 256 levels of gray according to the formula provided.

But when I use the same formula for my coated (which uses 175 l.s.), I get 2400/175 = 13.714285, and that squared is 188.08161, so for my coated I have 188 levels of gray according to the formula provided. This means I can expect banding because I don't have 256 levels of gray. I'm asking is this correct? I use round dot screens, Nexus' CQS Nominal screening 165,45,90,105 (aka 75,45,0,15). Does the formula still apply, and should expect banding? And if so, what could I do to fix it?

Don

 

[gordo]

Re: Standard Line Screen

Don,

RE:
"So I used the formula, and 2400/150 = 16, and 16 squared is 256, so for my uncoated (where I use 150 l.s.) has 256 levels of gray according to the formula provided.
But when I use the same formula for my coated (which uses 175 l.s.), I get 2400/175 = 13.714285, and that squared is 188.08161, so for my coated I have 188 levels of gray according to the formula provided. This means I can expect banding because I don't have 256 levels of gray. I'm asking is this correct? I use round dot screens, Nexus' CQS Nominal screening 165,45,90,105 (aka 75,45,0,15). Does the formula still apply, and should expect banding? And if so, what could I do to fix it?"

Don, that gray level limitation formula is outdated for modern AM/XM screens. I'm not familiar enough with Nexus screening - you will need to check with them directly to find out whether this is an issue for you. Vendors such as Agfa (ABS), Heidelberg (HQS & IS), Kodak (Maxtone) use a technique called dithering to increase the number of tones without the need to increase the resolution and compromising performance. For example, let's say that a combination of resolution and screen frequency (according to the formula) gives us only 100 tones and we want 200. By alternating 70% halftone dots with 71% halftone dots, we can simulate a tone area of 70.5% - an average of the values of the adjacent halftone dots. Doing this for all values from 0% to 100% effectively doubles the number of tones that can be represented. Some screening implementations allow you to set the parameters of how and how many gray levels will be maintained as you increase LPI (to avoid shadestepping/banding). In general, FM screens eliminate shadestepping.
Just a few more thoughts regarding points raised in this thread.
1) You do not necessarily need to go to a lower screen ruling when printing on uncoated paper.
2) Changing screen ruling is not an optimal method of managing dot gain
3) AM/XM screening dot shape will not impact number of possible gray levels
4) The actual screen frequency you get at any specific request will depend on the the particular screening algorithm being used (varies by vendor) and device resolution, and sometimes color). for example, on a 2400 dpi device - a request for 175 lpi may actually deliver a 169.706 lpi ruling

best, gordo

 

[disbellj]

Re: Standard Line Screen

Gordo,

Unfortunately if I call our vendor, since we are no longer under support contract with them, they will want to charge us thousands of dollars to renew the contract before answering the simplest of questions. So I will not be talking to them if I can help it. We already paid too much for them causing and not fixing issues for me to want to give them another dime at this point. Although I like their software that I use, we were sold what I consider additional (beta) software and I haven't put that stuff into production.

Heck, just in the last couple days I checked my screens and although it's set in Nexus to be CQS Quality Screening (CQSNominal), and screen angles being C 165 (aka 75), M 45, Y 90 (aka 0), K 105 (aka 15), when I checked my screen angles on the screened file in DotSpy, the angles were C 15, M 45, Y 0, K 75. I know it was correct at one time, but Nexus has gotten corrupted or something and now I'm getting switched angles for cyan and black! I was fixing to start worrying, but I checked the dot percentages and the bump curve is getting applied correctly, so that's why no one has noticed it (otherwise it would be hard to match proofs because the bump on cyan is 50 in = 54 out and black is 50 in = 49 out, so switched would see a 5% difference in color than expected). Oh well, just another example to me of stuff changing that shouldn't be, but glad I didn't "get bit" by it since it's at least using the correct bump curves still.

Don

 

[John W]

Re: Standard Line Screen

Don,
That's 5% dot not 5% effect. At the 50% area, any change has the largest effect because the dots have the maximum surface area and perimeter to gain. The 49 vs 50 is probably no big deal but the 50 to 54 is huge IMO and may translate into much more than a 5% visual effect depending on what the subject matter is, n'est-ce-pas?
I get upset if I am outside a plus or minus 1% dot at the 50 and for stamp work that tolerance is more like plus or minus half a percent.
I hope you can resolve the variance your set up seems to make. Do you have a swap c for k angles at the last minute as a quick workaround?
Cheers,
John W