It would even be better if he could be more precise on how to print the target.
As stated previously, the target should be printed using the exact same printer and color management (i.e., profile) setup that you plan to use for printing normal images (i.e., your own images).
Two examples:
1. Suppose you plan to print on an Epson 3800 on Epson Premium Luster paper using the Epson-supplied profile (Pro38 PLPP). You open the Outback Photo test images in PS and print using a color-managed workflow, such as the one suggested here:
http://people.csail.mit.edu/ericchan/dp/Ep...flow.html#colorPerhaps you make two sets of test prints, one using Perceptual and one using Relative Colorimetric. You make sure to choose the correct driver settings (e.g., Media Type = Premium Luster Photo Paper, 2880 dpi, etc.). You print, you let the prints dry (e.g., overnight), and then you evaluate and compare using -- for instance -- the criteria that Jack described in his article.
That's all there is to it.
2. Suppose you've gotten a custom profile made for your printer and you want to see how good it is, say, compared to the stock/canned profiles. You do the same process as described above, except that instead of using the stock profile you use your custom profile. Again, you make the print, let it try, and compare side by side. This should make improvements and deficiencies clear.
For instance, I recall a custom profile vendor suggesting to print a target *without* using profiles, just to test the intrinsic capability of covering color and tonal gamut by a printer and a particular set of media, i.e. paper and ink. Can the Outback target be used for this purpose, and how?
No. As stated previously, the Outback images are designed to evaluate profiles, not the raw state of a printer. While you certainly can print the Outback images with color management completely turned off, it's not particularly meaningful. You're better off in this case using the Atkinson Profile Test Images here:
http://homepage.mac.com/billatkinson/FileSharing2.htmland using the Hues, Blends, Twenty-Eight Balls, etc. images for testing a driver setup. One can also print a profile target to get an idea of the expected gamut for a driver setup.
In the Outback target, there are seveal gray scale and color patches. According to the following PhotoDisc evaluation, not all these are created equally. An user needs to know which ones to use for evaluation.
Sure, but for evaluation purposes it doesn't really matter how they're created. You have a reference -- how the patches appear on your color-managed display. You also know the exact RGB values -- e.g., by using the color picker in Photoshop. Now you have a test print made, and you can compare to the soft proof on your display how the results look. If the hues of the patches are way off, there's a problem. If the grays look distinct on your display but can't easily be differentiated in the print, there's a problem. If the grays have a green/magenta tint, there's a problem.
Again, many of these potential issues are already spelled out in Jack's essay ... just print and compare ...