Blinded by the Light

What brightness means—and doesn't mean—to projecting video.

It's probably not news, but over the last year there has been an aggressive display industry trend toward more affordable projector models. Most strikingly, there are now several SVGA models available for less than $1,000, as well as XGA units for less than $2,000. And in many ways, these bargain units do an admirable job putting a big-screen image where there once was none and allowing smaller video organizations to display content in group settings. Quality is good, especially with the barometer of history, and that's a testament to a progressively higher base level of technology manufacturers have been able to achieve.

Among today's crop of value-oriented projectors, smaller size and higher brightness often comes at the expense of brightness uniformity.

As projectors have become smaller and more affordable, and as manufacturers try to squeeze more light out of smaller lamps and light engines, it's almost inevitable that something would have to give. More affordable projectors generally aren't as bright as their premium siblings, but because brightness has shot up for all projectors in recent years, a little less brightness might not seem like a major compromise. But brightness isn't being used quite the way it used to be, and what's more, it isn't generally being reported as consistently.

Specifically, brightness uniformity, especially among those more affordable projectors, is falling, almost as a rule. In the interest of those smaller, brighter, and more affordable form factors, manufacturers are squeezing out high brightness, but often at the expense of displaying it evenly. While you might expect to see numbers like 80%, 85%, or even 90% uniformity listed in the brochures for premium products, that statistic is increasingly ignored on marketing literature, especially on affordable products, and for a good reason.

My experience with today's value products — while this trend is not exclusive to lower-price projectors, it is particularly common among that class (NEC's VT46, which I reviewed last fall, is a nice exception) — is that brightness uniformity often falls below 80% and even 70%. The result is, of course, a more prominent bright spot in the image's center. Whether that bright spot can be seen by a layman's naked eye depends upon how bad the disparity is and, in most cases, the nature of the material on the screen. Nevertheless, there's no question that the farther away from a uniform brightness a projector gets, the less versatile that unit will be in a variety of situations.

Minority Reports

In developing standard testing and reporting procedures for front projectors, ANSI (American National Standards Institute) established that brightness measurements taken over a nine-point tic-tac-toe-like grid across an entire projected image and averaged together would yield a uniform, industry standard for ANSI lumens. That makes good sense. After all, any bright light or lamp pointing toward an object — be it a flash light, a flash bulb, or a projector lamp — makes a bright spot at the center of its path. By reshaping projector lenses and light engines, the display industry had gone a long way toward reducing, and in premium products almost eliminating, those unwanted center bright spots from projected images.

Yet, prior to the formalizations of ANSI's testing methodology, there was a veritable hodgepodge of testing and reporting procedures from different projector makers, often in the interest of marketing advantage. Recent projector specifications seem to be moving back toward those days of measurement anarchy, particularly in cases where brightness uniformity is not a product's strongest feature.

Although ANSI lumens have been commonly reported for a few years now, brightness measurements in projector brochures are increasingly based on what is somewhat awkwardly referred to as “peak brightness.” It isn't that “peak brightness” is so incorrect a term for a projector's highest possible brightness, but it's a term derived from older CRT brightness measurements.

“Peak lumens” or “peak white lumens” used to refer to a brightness measurement taken from a small center rectangle of white surrounded by a black background. CRTs, which produce white from blending each of the three full-bright RGB light guns, can produce a brighter white if all the light is concentrated toward a single section (covering, for example, only 10 to 20 percent of the screen) rather than if the light is dispersed to cover the entire screen. Thus, rather than measure brightness across the entire image, this 10 percent white area became the source of “peak brightness.”

LCD and DLP projectors don't experience the same 10 percent brightness effect, so the “peak” tends to reference to the bright spot in the center. That area is usually brighter than the rest of an LCD or DLP projector's image, at least in more affordable projectors, but it doesn't matter what color it is or whether the rest of the image is white or black. As a matter of fact, you'll get the same brightness measurements for white whether you're measuring all white or the white squares in the ANSI black-white checkerboard contrast chart.

The bigger problem, however, with the trend toward reporting “peak brightness” is that it isn't a particularly accurate representation of the projector's capabilities, especially given increasingly poor brightness uniformity. What's more, the projector industry tends to rationalize brightness claims around the highest theoretical brightness for a given model. In other words, if 1000 projectors come off the same production line with a brightness range between them typically of some 10 to 20 percent, it's the top one or two percent that will set the mark for the brochure. And sometimes that's even rounded up a little to hit a good marketing number like 1000 or 2000. Ultimately, that means the average projector will have a reported “peak brightness” that's regularly 10 to 20 percent higher than reality, and because of decreasing brightness uniformity, an ANSI lumens average brightness that is even farther away from spec.

Hewlett Packard is, commendably, one manufacturer that tries to strike something of a balance between accurate reporting and industry competition. At the risk of greater consumer confusion, HP generally lists both an “ANSI lumen” and a “peak lumens” specification in its marketing material for those who understand the distinction. At least it puts the distinction out there for the public. NEC, in my experience, is an example of a company that does even better because it usually doesn't get caught in the fray. Instead, NEC, in the units I have tested, has done a good job staying accurate with uniformity and with brightness claims, with few exceptions.

For most manufacturers, however, industry competition ultimately means companies need to fight the spec war. Giving away marketing lumens is a risky proposition, especially in an industry dominated by OEM components used in many different brands. “Honest” manufacturers would do themselves a disservice if they “undersold” brightness as compared to competitive products with the same light engine.

And, the buying public is an often unwitting accomplice. It's just too easy to look at specs, at least as a way of initially narrowing the buying options or as a minimum requirement on an educational or governmental bid. There was a time, not too long ago, when projecting a strong, accurate image meant dimming room lights and drawing window blinds to ensure good viewing. That awkward segue left a strong, lingering impression that continues to drive prospective buyers toward higher brightness numbers.

Ironically, these days brightness is far from the most important specification, especially for watching good video. Brightness is almost always turned down by default in projectors' video/movie/cinema modes in the interest of richer color and deeper blacks. For example, in DLP-based models, that means “turning off” the white section of the color wheel in video mode. Many projectors also now offer a low-light or economy mode to automatically turn down brightness when light is controllable, as is often the case when watching video.

These days, it's easy to find a single-lamp installation, “portable” projector that can handily put out 2000 to 3000 or more lumens of brightness. Even tiny travel projectors regularly pass the 1000-lumen mark. That is, of course, unless you're looking for a video-oriented model. Most premium home theater and video projectors top out at 1000 lumens.

Why? Peak brightness is not only unimportant, but it can also detract from ideal viewing conditions. Contrast ratio is, for example, more important, but look at ANSI checkerboard contrast and not the wild 3000:1 type “full on/full off” measurements that DLP makers push. That's really not a representative number. Even more important is the range and uniform color temperature of grayscales that keeps red looking red and blue like blue, regardless of luminance or shadows. Black levels are critical for video and the antithesis to (although not necessarily exclusive to) high brightness. Also, many manufacturers are now listing fan noise (although be careful, because there is no true measurement standard here either).

Still, low fan noise, good optics, good image processing, de-interlacing, and scaling are good examples of projector makers paying better attention to detail, and those features ultimately matter more than the individual specifications. Rather than squeeze brightness out of relatively inefficient form factors, it's far more important for video projector makers to focus on good color, a full range of grayscales, and accurate black levels, while optimizing for uniform brightness.