Tuesday, April 05, 2011

Technical sales bullshit

To measure is to know, said Lord Kelvin.
But as marketing departments get more and more creative with their published specifications, what we’re left measuring — and by extension, knowing — about our gear is increasingly worthless.

With the gadget-buying in season year round, most of us will sometime soon be turning to those ubiquitous columns of numbers, ratios, and percentages before making our final selections.

Frequency responses will be consulted, dynamic contrast ratios compared, and color gamuts critiqued — all in an effort to gauge performance, determine value, and quickly pit one product against one another.

The only problem?

In many cases, you’d better off consulting chicken bones and fingernail clippings.

Not only are a growing number of published specs misleading and/or overinflated, some have become downright meaningless.

And it’s getting worse.

Remember how impressive something like Blast Processing sounded when you were 15?

Made the Super Nintendo look downright wimpy, right?

Well, spec cooking operates on more or less the same principle.

Only instead of inventing empty marketing words manufacturers plop a bunch of faux math in our laps.

These lies and fabrications happen for a few reasons.

First, numbers have tremendous sway over the decisions we make.

A recent study in the Journal of Consumer Research suggests that quantitative specifications are so powerful that, even when given the ability to directly test the attributes of a given product ourselves, we still tend to choose the thing with the longer list and bigger numbers (ahem, megapixels).

Another reason for the proliferation of BS specs?


The gadget world is loaded with gimmicks and lies because it’s extremely competitive,” says Raymond Soneira, president of DisplayMate Technologies.

Soneira, who penned what many consider the debunking Bible for display specifications over at MaximumPC, says that as technological complexity increases in the gadget world, it gives manufacturers and marketers even more leeway to futz with the numbers.

And futz they do.

Most consumers don’t understand the technologies anyway so they are easily misled, fooled and even swindled,” he says.

More than anything though, there’s one simple reason behind the rise of dubious specs:

It’s become an industry necessity.

The temptation to exaggerate is now so overwhelming that attempting to stay out of the gimmick game is now seen as akin to product suicide.

Try to anchor your specifications in the real world (with meaningful numbers) and your product will look inferior.

Don’t publish them at all, and you’ll look like you’re trying to hide something.

It’s an insidious Catch-22 for anyone with an ounce of integrity, so manufacturers and marketers simply make the easy choice.

David Moulton, a veteran audio engineer, musician and producer characterizes the gadget spec situation like this: “

When engineers make a product they use specific tests to measure the performance.

But when sales departments gets a hold of those test measurements, they start using those numbers as describers of value.

They become, in essence, sales arguments.”

So which “sales arguments” should you avoid, dismiss, or at the very least raise a skeptical eyebrow at?

We’ve compiled a quick list of some of the more brazen spec gimmicks to be wary of

Color Gamut

What it is: This spec represents the range of colors a given display can produce, and is usually expressed as a percentage of a particular color standard, like Rec.709 (HDTVs) or sRGB (computers and digital cameras).

Why it’s bullshit: Manufacturers don’t tell you this, but the color gamut you actually want on all of your displays is the same one that was used when the content you’re viewing was created.

If it’s different, you’ll see different colors than you’re supposed to see.

Nevertheless, most companies are happy to exploit the common misconception that a wider color gamut is somehow indicative of a better display.

So what’s up with those 145 percent color gamuts?

Nothing special, really.

Here’s what a larger gamut will do: make everything look saturated.

Indeed displays claiming to have more than 100 percent of any standard color gamut aren’t able to show colors that aren’t in the original source image, says Soneira.

Contrast Ratio

What it is: Divide the brightness of peak white by the brightness of black on a display (after it’s been properly calibrated) and, voila, you’ll get what’s known as the contrast ratio.

Why it’s bullshit: In the real world, this measurement typically falls between 1,500:1 and 2,000:1.

And that’s for the best LCDs, says Soneira.

But those numbers are a thing of the past.

The allure of bigger ratios has prompted manufacturers to bake this specification into a full-fledged nonsense soufflé.

Today, we get what’s known “dynamic contrast ratio.”

That’s reached by measuring blacks when a display’s video signal is entirely, well, black (when it’s in a standby mode).

As you can imagine, that significantly reduces the light output of the unit and is obviously much darker than what’s actually used to determine the traditional contrast ratio with an actual picture present.

Using this trick you’ll get, in some cases, astronomical contrast ratios like 5,000,000:1 or, in Sony’s case, “infinite.”

While still technically true, this spec is utter nonsense and completely unhelpful in gauging real world performance.

The only information that dynamic contrast ratio can relay is how much brighter the whites can be than the blacks.

Response Time

What it is: Also referred to as latency or response rate, response time is a standard industry test that tries to quantify how much LCD motion blur you’ll see in fast moving scenes. (It doesn’t apply much to plasma displays).

It’s determined by measuring the time it takes for one pixel to go from black to peak white and then back to black (rise-and-fall).

And it’s not a particularly good indicator for real picture blur.

Why it’s bullshit: Consider this.

In the span of five short years, display response times have gone from 25ms (milliseconds) to, in some cases, 1ms.

How did this magic happen?

Well, it kinda didn’t.

The problem here, according to Soneira, is that most picture transitions involve much smaller, more subtle shades of gray-to-gray transitions, which usually take much longer (3-4 x) to complete.

Those response times are far more important to a display’s ability to handle motion blur.

But consumers often have no way of knowing which response time is being measured (gray-to-gray or rise-and-fall).

Because the published specifications can have a considerable impact on sales, it is often more important for a manufacturer to reduce the black–to–peak-white–to–black response time value rather than improving the visually more important gray-to-gray transitions.

The result?

The LCD display with the fastest response time specifications may not have the least visual blur.

Viewing Angle

What it is: Pretty simple stuff: the maximum angle at which a display can be viewed with acceptable visual performance.

Yes, there are generalities about viewing angle that everyone should know:

A plasma display, for instance, will yield a wider view angle.

But when it comes to the listed angles that manufactures include in spec sheets, you can pretty much ignore them.

Why it’s bullshit: Today, it’s not uncommon to see 180-degree + (total) viewing-angle specifications for many displays.

This has absolutely no bearing on the actual acceptable viewing angles, according to Soneira.

What most consumers don’t realize is that the angular spec is based on where the contrast ratio falls to a level of 10:1, hardly an acceptable (or visually pleasing) figure.

More realistically, an angle of ±45 degrees may reproduce an acceptable contrast ratio, but only with very bright and saturated colors.

Pictures that include a wide range of intensities, hues and saturations will appear “significantly degraded” at much smaller viewing angles.

Of course, no one tells you this.

by Bryan Gardiner, Gizmodo.com

Thanks I liked that article!

No comments: