DTP


 
Lively discussions on the graphic arts and publishing — in print or on the web


Go Back   Desktop Publishing Forum > General Discussions > Print Production & Automation

Reply
 
Thread Tools Display Modes
Old 11-05-2007, 10:46 PM   #1
flite
Member
 
Join Date: Nov 2007
Posts: 3
Default Line resolution problem

Hi!

I produce reports to be printed on laser printers. In order to use smallest possible images, I try to figure out what line resolution the printers (Kyocera) use. It's 600 dpi printers so it should be 50-90 lpi. I suppose it can be found in some spec. or in the driver, but I haven't managed to do that.

So I did a simple test. I printed a continious gradient from black to white (4 inches wide on paper) at 200 ppi and the printers made 17 fields of half tones of the image. According to this formula:
number of half tones = (printer resolution[dpi] / line resolution[lpi])^2 + 1
17 = (600/lpi)^2 + 1
the printer uses a line resolution of 150 lpi, and the grid size is 4x4 dots (600/150).

Let's assume this is correct. Having a line resolution of 150 lpi would require 225-300 ppi for the image. I printed with images from 40 to 600 ppi, and they all produced almost identical results (17 half tones).

What is wrong with the theory above, and how can I find the line resolution?

If I print the 40 ppi on a HP laser printer, the gradient is continious, so one conclusion is that Kyocera printers are not made for half tone images.
flite is offline   Reply With Quote
Old 11-06-2007, 04:10 AM   #2
George
Member
 
George's Avatar
 
Join Date: Feb 2005
Posts: 1,036
Default

Quote:
Originally Posted by flite View Post
I printed with images from 40 to 600 ppi, and they all produced almost identical results (17 half tones).

What is wrong with the theory above, and how can I find the line resolution?
I don't know. But aren't your empirical findings sufficient for your objective??

George
George is offline   Reply With Quote
Old 11-06-2007, 11:46 AM   #3
Steve Rindsberg
Staff
 
Join Date: Nov 2004
Posts: 6,711
Default

>> Having a line resolution of 150 lpi would require 225-300 ppi for the image. I printed with images from 40 to 600 ppi, and they all produced almost identical results (17 half tones)

Knowing the lpi, you can calculate the optimale resolution for the images you use, yes. But changing the resolution of the images won't change the lpi of the printer; you can't change the number of halftone values by changing the image resolution. You have to change the linescreen.

You do that in some cases by specifying the value in the applications you're working with or by choosing different printer modes (or if you're a bit odd, by diddling the postscript file in an editor).

By the way, you don't mention whether you have a PostScript printer or not; the formulae you've used only apply to PS printers.

Are you using Windows or Mac?

   
__________________
Steve Rindsberg
====================
www.pptfaq.com
www.pptools.com
and stuff
Steve Rindsberg is offline   Reply With Quote
Old 11-10-2007, 07:12 PM   #4
Howard Allen
Member
 
Howard Allen's Avatar
 
Join Date: Oct 2007
Location: Calgary, Alberta, Canada
Posts: 824
Default

Are you a geek?

I am. If I can't find the LPI specified, I count 'em myself. I have a microscope, but a decent loupe or hand lens would work--count the lines per quarter or eighth inch, then multiply by 4 or 8.
Howard Allen is offline   Reply With Quote
Old 11-11-2007, 03:48 AM   #5
flite
Member
 
Join Date: Nov 2007
Posts: 3
Default

Hi Steve, and thanks for your answer.

>>Knowing the lpi, you can calculate the optimale resolution for the images you use, yes. But changing the resolution of the images won't change the lpi of the printer; you can't change the number of halftone values by changing the image resolution. You have to change the linescreen.

I know, the idea was to find where the image started to look bad, but every print was (equally) bad, so even 40 ppi was above the threshold.

>>By the way, you don't mention whether you have a PostScript printer or not; the formulae you've used only apply to PS printers.

It's PCL printers. I thought the formula applies to the RIP-process. Does it matter wheather it's PS or PCL going into this process?

>>Are you using Windows or Mac?

It's Windows.
flite is offline   Reply With Quote
Old 11-11-2007, 07:22 AM   #6
Steve Rindsberg
Staff
 
Join Date: Nov 2004
Posts: 6,711
Default

>>I know, the idea was to find where the image started to look bad, but every print was (equally) bad, so even 40 ppi was above the threshold.

But to reiterate (that's French for "beat a dead horse" isn't it?) changing the image resolution will have no effect on the printer's linescreen setting. Your tests will give you valueable insight into how much image resolution is needed at a minimum and how much is too much (ie, over a certain amount, you won't see any image improvements), but won't teach you much about linescreen settings and their effects.

>>It's PCL printers. I thought the formula applies to the RIP-process. Does it matter wheather it's PS or PCL going into this process?

The formulas apply to the way PS handles halftones. All of my experience is with PS and way back in the past, a wee bit of PCL, which used a dithering process rather than a dot-based halftone process. IOW, back then, the PS formulas didn't apply to PCL but what little I knew about PCL is probably all wrong now.

   
__________________
Steve Rindsberg
====================
www.pptfaq.com
www.pptools.com
and stuff
Steve Rindsberg is offline   Reply With Quote
Old 11-16-2007, 08:19 AM   #7
curveto
Member
 
curveto's Avatar
 
Join Date: Feb 2005
Location: some room with white padded walls ... now surrounded by Saguaro Cacti
Posts: 351
Default

Your methodology is correct. As Steve noted your thoughts/testing relate to achieving gradients via screening vs. dithering. You can short circuit some of your testing by looking at the device's PPD in a text editor. That said, the testing you're doing is always an eye opener. Don't forget to calibrate (and, if color, profile/normalize) the device as that will yield improved performance (for reasons that you'll find obvious after doing so).

On a related note... unlike what routinely passes for correct advice (e.g., "300dpi enough image data to support a 10 lpi screen" or something similarly wife's tailish) if you run objective tests you'll find that the amount of image data you push at a device in support of a given screening frequency is MUCH lower than most people believe/espouse. Mathematics dictates that accuracy of halftone dot composition diminishes above 1.5x lpi. Every time I've ever done objective testing (i.e., please tell me which of these prints looks the best and please tell me where in this series of increasingly dense images (at a constant lpi) you stop seeing a quality improvement) no one has ever cited a sample above 1.8x as "improving." The vast majority of people I've tested don't notice any difference above 1.6/1.65x. The only ones that do generally fit squarely into the "design audiophile" sort of camp (i.e., my $250,000 speakers are better than your $100,000 ones ... they tend to like phonograph-era tech too).

Dithering (i.e., FM screening) requires less data (1 to 1.5x the "halftone" frequency), btw. ...at the slight expense of loss of subtle detail (for equivalent device settings).

...and on an unrelated note:

JPEGs for screen use can be heavily reduced in quality (and thus size). I routinely build 72dpi (at screen/absolute dimensions) JPEGs with Quality: 3 (sometimes 4) with exceptional results. ...and the files wind up being tiny.
curveto is offline   Reply With Quote
Old 11-16-2007, 10:09 AM   #8
flite
Member
 
Join Date: Nov 2007
Posts: 3
Default

Hi, and thanks for your replies!

The reason of the 17 shades of gray was the (default) use of "ordered Bayer" dithering method. Changing this to floyd-steinberg made the gradient "continious" and other images way better. A quick test with different images and resolutions showed that 120 ppi was enough.
flite is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Screen resolution? Bo Aakerstrom Hardware & Gadgets 7 08-30-2007 08:51 PM
Printer Ready PDF: 10 ft x 4ft (resolution?) FlipSide Print Production & Automation 7 10-06-2006 10:30 AM
Quark 6.5 pdf - resolution/file name? limacguy Print Production & Automation 6 09-17-2005 03:30 PM
Scanning resolution Adrian Images 9 07-24-2005 08:25 AM
photo resolution plus CYMK or RGB dehavenphoto Images 5 03-24-2005 06:27 AM


All times are GMT -8. The time now is 08:50 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
Contents copyright 2004–2014 Desktop Publishing Forum and its members.