Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
Printer

Better Test Pages for Color Printers? 30

Posted by Cliff from the printers-as-peacocks dept.
AigariusDebian asks: "I bought a new color laser printer (Lexmark C510), but I am not sure if my driver configuration gets all the best images out of it. I would also like to have some way to brag about it to other laser printer wielding geeks. That means that I'd need to make some kind of repeatable and measurable testing of the printers quality. Are there any good Postscript test files that will allow a non-expert to do a full test of their color laser printer? I am thinking about both resolution, dithering and color matching testing."
This discussion has been archived. No new comments can be posted.

Better Test Pages for Color Printers? More

Better Test Pages for Color Printers?

Comments Filter:

  • How about this (Score:3, Informative)

    by mattdev121 ( 727783 ) on Thursday June 23, 2005 @07:42PM (#12895801) Homepage
    It won't test actual images that normal people can understand but at least geeks will stare in awe (or disgust) at the CUPS standard test page.

    The page tests all of the colours, as well as a resolution test by drawing lines in a circle 1 degree apart. A standard on most *nix machines, the CUPS page should be well recognized and serve as a good benchmark.

    • Re:How about this (Score:2, Interesting)

      by markild ( 862998 )
      I've done a few tests of printers over the years, and I must say that the lines in circle is a damn good way to see the accuracy of the printer. ...at least when you look at the effort you have to go through...

      You wan't see much about color accuray etc, seeing that all colors when you see them in the context of the other colors pretty much looks perfect. To get an test that does that, I would guess you would need an hardware color-pick-thingy.
      • For a start, I'd suggest :
        Use the standard CUPS page with it's colour pinwheel, but also enough circles of 1 deg lines to do a pair of each primary colour. Make the line transition from one primary at the inner point to another primary at the circumference.

        A set of colour circles like that would give you an indication of:
        - the colour resolution (from the radials expanding, just like normal B/W)
        - mixing (from the transition from one primary to another)
        - and any problem with colour overlay. (Eg, the ones inv

  • Drivers? (Score:3, Interesting)

    by Stevyn ( 691306 ) on Thursday June 23, 2005 @07:50PM (#12895860)
    First of all, you have a lot of time on your hands. But then again, I'm commenting on this so that doesn't say much about me.

    Second of all, what OS is this for? Windows, OSX, Linux? The driver options may allow you to tweak some things that can make a big difference.

    I'd recommend you find a nice very high-res photograph. Nothing looks better than a nature seen with a lot of color. It will look better than some test pattern. You could also tweak it in Photoshop or the gimp to bring out the color and contrast. I'm sure there are many resources that professional photographers submit for public use that google could help you out with.

  • Now I know what I need!! (Score:4, Funny)

    by OneDeeTenTee ( 780300 ) on Thursday June 23, 2005 @07:57PM (#12895923)
    A great printer is sure to attract a girl.

    Thanks for the idea!

  • Try mine... (Score:5, Informative)

    by digitect ( 217483 ) <digitectNO@SPAMdancingpaper.com> on Thursday June 23, 2005 @08:23PM (#12896108)

    I wrote this one: ColorCard.pdf [sf.net] (1.4Mb).

    It tests all the characteristics of a printer that I cared about:

    • color accuracy -- C,M,Y,K,CM,CY,MY,CMY,CMYK in density increments of 10% from 10-100%
    • SWOP color accuracy -- Same as CMYK except in 25% increments
    • grayscale accuracy -- Increments of 5%
    • resolution -- CMYK tests of lines down to 0.25pt at 45 degree angles

    Of course, you'll need a reference card by Kodak, etc. to test against, but I've found this is the simplest way short of developing a whole complex spot color sample group of colors closer to the edges of gamut.

  • PDF test pages (Score:3, Insightful)

    by srmalloy ( 263556 ) on Thursday June 23, 2005 @08:44PM (#12896257) Homepage

    On this page [alotofthings.com] are download links for four PDF files you can use as a printer test:

    You should be able to get a decent picture of how well your printer renders colors with those files.

  • Printer/Monitor Calibration Images (Score:4, Informative)

    by elkyle ( 875715 ) <elkyle@comcast.net> on Thursday June 23, 2005 @08:54PM (#12896323)

    Here is a test image [comcast.net] that I believe was designed for printer/monitor color calibration (I dredged it up from the data recovered from a hard drive failure a while back). It is 2297x3600x24 at a resolution of 360 pixels per inch for a final image size of 6.381x10 inches. Included on the image are several Kodak color charts, along with a variety of household objects and several faces of varying skin colors.

    Here [comcast.net] is another test photo of the same style as the first one. According to the CNET Labs printer page [cnet.com], this is "the industry-accepted PhotoDisc Target document." Not sure about that (especially with the website logo in the corner of the image), but whatever. It works nonetheless.

    Following with the previous recommendation of finding a vibrant nature photo, I located a rather beautiful photograph [sxc.hu] (free registration required for download) of the Grand Canyon with a large variety of colors at 1200x1600x24.

    I have been thinking about this on and off for a while now (still stuck in lowly inkjet-land). Thanks for finally motivating me to do some research. Hope this aids you in your obsessive-compulsive quest to achieve satisfaction from knowing every little defect in your printer.

    • Yep, that image has it all, the MacBeth and IT8 charts in the middle, the Kodak step wedges underneath, and several "Shirleys" on the bottom. Everything else is pretty useless.

      I'd personally recommend using just the MacBeth and IT8 charts without all the other bullshit.

  • "Shirley" (Score:5, Informative)

    by sakusha ( 441986 ) on Thursday June 23, 2005 @11:45PM (#12897372)
    For many years, Photographers have used test shots of women, they nicknamed her Shirley for no reason I've ever been able to figure out. Oftentimes the woman is a pale redhead, as those particular skin and hair tones are hard to render in print, but easy to see if they are not rendered correctly. Skin tones tend to be pale and even a slight miscalibration will cause dramatic color shifts. After working in prepress a few years, I found that I can tell if a color separation is OK just by looking at the individual C M Y and K negatives, if I look at skin tones in a face.
    In that regard, one of my favorite test images is "Ole No Moire" that used to come in versions of Photoshop (I don't think it comes in current versions, I haven't seen it in years). Ole no Moire also has step wedges on the image. The one thing it won't do is test out the fancier PostScript features, it's just a bitmap image. I like to create my own step wedge charts in Adobe Illustrator, so I can test exactly the features I want to see.
    There are other sorts of high-tech test images, like the IT8 test pattern. Of course, a printout of this is absolutely unintelligible except to a colorimeter, but if you have a full color calibration system, these are essential. A more general color guide used by photographers is the MacBeth ColorChart, which is a set of 24 basic colors and greys, they're all chosen carefully so that if the color balance is wrong, at least one color will stand out as mismatched when you compare it to the original chart.

    • Re:"Shirley" (Score:2, Interesting)

      by Marcion ( 876801 )
      >For many years, Photographers have used test shots
      >of women, they nicknamed her Shirley for no reason
      >I've ever been able to figure out.

      There is also the Lenna picture [wikimedia.org], see the wikipedia [wikipedia.org] for the background story.
      • I always thought the Lena image was a particularly bad test image. It was scanned from a halftone screened CMYK magazine page, which means it had a restricted gamut before it ever went into a scanner, and this was an early, primitive scanner.
        It is preferable to use an original continuous-tone slide or photographic print scanned by modern equipment for a test image.

  • Brag (Score:1, Funny)

    by Anonymous Coward
    • If you want to brag then your first mistake was buying a Lexmark. LOL
    • Good luck with that!

  • Go with a classic.. (Score:4, Interesting)

    by technos ( 73414 ) on Friday June 24, 2005 @06:40AM (#12898877) Homepage Journal
    Lena! [cmu.edu]

    No matter what your final test image, you'll have to work that one in for the inevitable conversation with your wife.

    "Who the hell is this hussy you keep printing nude pictures of and throwing them away?"
  • If you create your own test page and not use the genuine one Lexmark may sue you under the DMCA!
  • druckerchannel.de, part 1 [druckerchannel.de]
    druckerchannel.de, part 2 [druckerchannel.de]

    heise.de [heise.de] (those are meant primarily for scanner tests, but they also come in handy for printer comparisons)
    heise.de, older version [heise.de] (unfortunately, they offer only a 300 dpi version, not the original)

    And while you're at heise.de, check out those [heise.de] cool backgrounds, each available in several resolutions. OK for individual use, no redistribution allowed.
  • Norman Koren links to several test images on his site [normankoren.com].

Slashdot Top Deals

One man's constant is another man's variable. -- A.J. Perlis

Close