Rating the Tech Behind the Specs

Buying computing products is a numbers game--but our Test Center can help.

Call them buzznumbers: the tech specs that vendors pelt you with every time you make a purchasing decision. In theory buzznumbers help you pick the best products. But too often they confound rather than clarify--no matter how experienced a shopper you happen to be.

Will you be any more productive, for instance, if you spring for a 2.8-GHz Pentium 4 PC instead of a 2.66-GHz one? How many ANSI lumens does a projector need to wow an audience? Do your most cherished 35mm photographs deserve a scanner with a full 48 bits of color, or will 42 bits do the job?

The short answer to these buying dilemmas--and countless others--is that they aren't mathematical propositions. Some buzznumbers bear no relation to how real people use tech gear. Others are outweighed by intangibles like the sophistication of a product's software. And specs can't be compared unless vendors use the same yardstick--which they often don't.

Consider the hot commodities known as LCD monitors. One of the principal buzznumbers used in marketing them is contrast ratio--the ratio of a screen's whitest white to its blackest black. All things being equal, the image from a display with a 400:1 contrast ratio should outclass one with a 300:1 ratio. One with a 500:1 ratio should look better still. But do the specs help you buy a better monitor?

One of the privileges and pleasures of working at PC World is that we can answer such questions. At the suggestion of Senior Associate Editor Seán Captain, we hauled 15 LCDs into our Test Center, where Senior Performance Analyst Jeffrey Kuta measured their contrast ratios. We worked with an independent expert, Raymond Soneira; his DisplayMate utility for optimizing image quality has long been a favorite of savvy PC users, including those in our lab.

You'll find the results in "LCD Specs: Useless?" written by Senior Associate Editor Tom Mainelli. The news wasn't all bad: Some LCDs delivered a better ratio than their vendors claimed. But others fell short of their official rating by as much as 50 percent; certain units that touted a 500:1 ratio underperformed one that advertised a mere 350:1 spec.

If those findings startle you, join the club: "I emerged with even more respect for our lab," says Senior Editor Yardena Arar, the article's editor, "and even more skepticism for vendor numbers."

Our findings also raised eyebrows at the Video Electronics Standards Association, whose procedures are widely used to rate contrast ratios. VESA plans to revise its tests for more consistent ratings, and will ask manufacturers to label ratings that adhere to its methodology.

In other words, our experiment helped prompt an industry to tighten its standards. That's yet another reason I'm glad to spend my days at PC World--and a real victory for everyone who thinks specs should actually mean something.

Two Great Chances at Some Great Prizes

Want a shot at winning a home theater and TV system? Take our subscriber survey on readers' experiences with PCs and other technology products they own or use. Subscribers can rate this issue for a chance to win an Epson Stylus Photo 960 ink jet printer; you can enter from May 5 through June 5, 2003. You'll need your subscriber number (from the mailing label on the front of your magazine) to take these surveys online. Each site includes the official rules and details on how to participate in the drawings if you're not a subscriber.

Harry McCracken is editor of PC World.