DLP vs. LCoS

When HP's 58-inch md5880n DLP and Sony's 60-inch KDS-R60XBR1 SXRD rear projectors converged simultaneously on our lab, we knew we had the prime candidates for another installment in our series of display technology face-offs. We began by calibrating and reviewing each set independently (see pages 32 and 34) before putting the TVs alongside each other in our darkened lab and tuning their pictures to match the ideal settings for brightness and color. The sets were fed the same test patterns and program material through a Gefen HDMI splitter; test signals were native 1080i HDTV or standard-def DVD content upconverted to 1080i by our high-end Denon DVD player. Al Griffin and David Katzmaier were joined by our TV reviewer Phil Ryan to fill out our panel of experts.

DETAIL One debate we hoped to squelch with this test was whether DLP could match LCoS on image detail. Both sets are spec'd at the 1,920 x 1,080-pixel resolution that defines a 1080p display, but they deliver the image differently. LCoS TVs use microdisplay chips that have a discrete 1,920 x 1,080-pixel grid; all the pixels on the chip appear simultaneously on the screen. DLP currently achieves 1080p resolution using a technique called "wobulation": the pixels are split between two fields flashed up in rapid succession, each with half the total pixels.

As it turned out, the Sony was marginally better at resolving test patterns, but the HP actually scored a bit higher in the subjective apparent sharpness of the picture, demonstrated by its ability to resolve the fine brush strokes in the close-up of an old oil painting in an HDNet travel show on Madrid. We called it a draw: these stellar sets proved that wobulated DLP and LCoS can both deliver all the goods in today's 1080i signals.

COLOR The first thing we look at in evaluating a TV's color is its ability to accurately maintain an industry-standard tone of gray as the brightness of the image increases from black to full-on white. We call this "grayscale tracking," and the HP did it very well, delivering a perfect neutral gray image on black-and-white test patterns of varying brightness. The Sony, on the other hand, also closely tracked the same hue at different brightness levels - but instead of delivering perfect gray, it lent a slight pink cast to the picture. This was mostly evident in bright scenes, such as one from the DVD of The Lord of the Rings: The Fellowship of the Ring that takes place on a snowcapped mountain. Both snow and sky looked a bit rosy or purple next to the more natural white and blue reproduced by the HP, although the effect wasn't obvious on most scenes and difficult to detect absent the HP right beside it.

DETAIL ••••• •••••
COLOR ••••½ ••••
UNIFORMITY •••• ••••½
CONTRAST •••• •••••

Color accuracy was very good with both TVs as measured by test patterns, but the Sony's colors were noticeably more saturated than the HP's. It was satisfying to see such rich colors, but they occasionally looked unnatural and hyped next to the HP, especially red. That helped explain why the HP had more accurate and natural skin tones in one of the brighter scenes we viewed from the DVD of The Matrix Reloaded.

Are these differences attributable to the display technologies? For what it's worth, the rosy cast on the SXRD was reminiscent of a similar effect David has observed in earlier LCoS TVs, while the Sony's better saturation seems tied to its ability to produce a near-perfect black - something that may or may not be technology-dependent (more about that later). Either way, all things considered, our panel scored the HP slightly better overall on color.