• On The Insider: Hotties from Across the Pond
Click Here
advertisement

  inkjet printers
CNET Labs tests three aspects of printer performance: print speed, print quality, and economy (cost of consumables).


Test environment

We test inkjet printers on a Dell Dimension XPS B 733r running Windows 98 SE with a 733MHz PIII processor, 128MB of RAM, and an Nvidia TNT2 M64 graphics card.

Before testing each printer, we install a fresh drive image containing the Windows OS and the applications required for testing. All the printers are installed according to the manufacturer's instructions for USB installation.


Print-speed tests

Manufacturers and independent testing labs use a variety of different methods to time printer performance. CNET Labs times its tests using the click to clunk approach. Click refers to the last command that is executed to initiate a print job, and clunk refers to the moment when the last printed page of the job falls into the output tray. We believe this method portrays most accurately the user's own experience.

CNET Labs' inkjet-quality test document
Click to enlarge
We perform all our speed testing in default mode on plain paper, except for photographs, where we use the printer's own photographic-paper default setting. We perform at least three trials of each test, ensuring consistent, repeatable results. The printer is powered off between each trial.

We test performance in several categories, depending on the type of test document printed. Because inkjet printers can print a wide variety of documents, we test each model using text documents, graphics documents, and a color photograph. Our text document is 10-page, text-only, 62K Microsoft Word file that employs a variety of font sizes and fonts, including serif, sans serif, italic, boldface, and plain. Our graphics document, modeled on a typical business report or sales chart, is a 1-page, 62K Microsoft Word file that features various fonts and colors as well as a pie chart. Our test photograph is a 2.07MB, 1-page JPEG file of two smiling children in front of a chalkboard.



Print-quality tests

We test inkjet printer output quality by printing typical documents on typical media: text printed on both plain and coated papers; graphics printed on both plain and coated papers; and photographs printed on photographic paper.

We conduct plain-paper quality testing using letter-sized, office-style, 20-pound bond copy paper. The coated inkjet papers and various photographic papers used for quality testing are those designed by the manufacturer for use with the printer being tested.

We print all of the test documents using the printer's default settings, with the exception of the photographic-quality test. In our photograph-print test, we print at the best quality each printer is capable of generating.

Each printer's output samples are evaluated on their own merits, then compared to that of other, similar inkjets CNET Labs has tested recently.

Text, graphics, and photo samples
The text sample we print is a one-page Microsoft Word document containing a variety of serif and sans serif fonts, ranging in size from 2 points to 10 points, in italic, bold, and regular typeface.

The graphics sample is a one-page Adobe Portable Document Format (PDF) file that contains several monochrome and color photographs, as well as a variety of graphical elements, including numerous gradients, seven blocks of solid color, a multicolored overprinting test pattern, and several line-art representations of a spectrum of objects.

The test photograph is an Adobe Photoshop sample photo in JPEG format. This is the same photo we use for performance testing: two smiling children standing in front of a chalkboard.

Jury evaluations
CNET Labs conducts jury evaluations to determine quality ratings for each printer's output samples. Our team of imaging specialists examines the output in a well-lit room. The text samples are evaluated for overall legibility; the clarity, density, and weight of the letters at different sizes and in different forms; and the absence or presence of a wide range of printer-induced flaws. We also consider the quality of the black ink: Is it nicely dark or a muddier brownish or bluish color? We compare graphics samples to an original/archetypal version printed on an IRIS commercial offset-press proof and fine-arts-reproduction inkjet printer, paying close attention to the following factors: the range of tones it can produce in color and monochrome gradients; color qualities such as warmth, coolness, or depth; color accuracy; photo quality; details; and the absence or presence of other printer-induced flaws. We rate the photograph output sample on many of the same criteria: its crispness, clarity, color accuracy (especially rendering skin tones), and overall quality.


Printer-economy tests

To determine an inkjet printer's cost per page, we install a new ink cartridge in the printer and print special drain documents created in Adobe Photoshop for the CMY (color) or K (black) cartridges. The K drain document comprises 50 percent pure black bars and 50 percent blank space. The CMY drain document is composed of alternating groups of cyan, magenta, and yellow bars with blank spaces between the bars. The printer outputs the documents until the image deteriorates due to lack of ink or until the printer refuses to continue. This test yields two cost-per-page numbers, one for the black ink cartridge and one for the color ink cartridge.

Calculating cost per page
To determine the cost per page for black ink, we divide the current street price of a replacement black cartridge by the number of pages printed in the K drain. We then divide this result by 10 to determine the cost per page based on an actual ink coverage average of 5 percent. CNET Labs uses a similar calculation for the CMY result. The CMY cost-per-page calculation uses a 25 percent color-coverage figure and assumes some black ink is also used in most color documents.