42IN WS LCD HDTV 1080P 1920X1080 BLK 3HDMI SVID SPKR TUNER review

Our calibration of the Philips consisted mainly of adjusting light output, tweaking the color control, selecting the Warm color temperature preset, and turning off the Philips' many automatic picture controls. The Warm picture preset actually came fairly close to the standard color temperature, but it was still too bluish and we wished it was adjustable. The lack of a backlight control was also a hindrance, and without it we were unable to achieve the best balance between black level and shadow detail.

We disabled Active Control because its automatic adjustments of picture parameters were too heavy-handed and ended up being less accurate than our normal calibration. In some scenes it increased contrast too high, clipping detail in bright areas, and in other scenes that changed in intensity from bright to dark or vice-versa, the fluctuation of the backlight was noticeable and distracting. Disabling the function did hurt black levels in dark scenes, but that was better than the alternative.

For comparison's sake, we lined the Philips up next to a few other entry-level LCDs, including the Sony KDL-52V5100, the JVC LT-46P300, and the Panasonic TC-32LX1. We also included our Pioneer PRO-111FD for reference. For this review we used the Blu-ray of "I Am Legend" for most of our image quality tests.

Black level: In dark scenes it was obvious that the Philips produced the brightest and least realistic shade of black among the LCDs in our lineup. As Will Smith dreams of the race to get his family out of the city, for example, the shadows inside the car, the black of his jacket, and the letterbox bars, for example, all appeared brighter and lower-contrast, with less pop, than they did on the other displays. Compared with the Sony and the JVC, the difference was visible even in brighter scenes.

Shadow detail was as good as can be expected on a set with lighter black levels, and we could make out about as many details in the darkened lab, for example, on the Philips as we could on the Sony. Details appeared more obscured on the JVC, but of course the latter's superior blacks made dark areas it displayed much more realistic overall.

Color accuracy: In brighter scenes the PFL6704D acquitted itself relatively well. During Smith's initial drive through Manhattan, for example, the red of the Dodge and the green of the trees in Madison Square Park looked quite realistic, if lacking the saturation and punch of the other displays (aside from the Panasonic). Smith's skin tone didn't appear quite as natural as on our reference display, but it was still OK--about the same as the JVC and better than the Panasonic, but not as good as the Sony. In darker scenes we couldn't help but notice that the Philips tinged shadows and black areas bluer than any of the other sets in our comparison.

Video processing: Digital Natural Motion is the company's name for its dejudder processing, and overall it didn't perform quite as well as the Sony we tested. As we've mentioned before, we prefer to leave all such processing turned off when watching film-based material since they tend to make film look more like video, and both strengths of DNM on the Philips were no exception. If forced to choose between the two we'd take Minimum as opposed to, naturally, Maximum, because the latter is particularly artificial-looking and much more prone to artifacts. Sony's equivalent to Minimum, dubbed Standard, did a superior job of maintaining a more film-like look and reducing artifacts.

We checked out Smith's race through Manhattan to compare between the two, as on the Philips we noticed a slight halo around the speeding car at times, particularly when it passed in front of a complex backdrop like a building, whereas the Sony's halo was much less noticeable and frequent. Bits of fast-moving objects, like the car's rear end, would occasionally detach slightly and reattach on the Philips, but that artifact didn't occur as often as on some sets we've seen.

We appreciated that, when we deactivated Digital Natural Motion, the Philips handled 1080p/24 sources properly. During the flyover of the Intrepid, for example, the planes on the deck were rendered smoothly and film-like, without the stuttering motion characteristic of 2:3 pull-down, as we saw on the 60Hz JVC.

During the motion resolution test the Philips performed about as well as most 120Hz LCDs we've evaluated. It delivered between 500 and 600 lines of resolution with Digital Natural Motion engaged, and between 300 and 400 with the processing turned off. The set correctly de-interlaced 1080i film and video sources, and as expected resolved every line of 1080i and 1080p still resolutions when in the Unscaled aspect ratio. As usual, it was difficult to discern any of these resolution characteristics in program material as opposed to test patterns.

Uniformity: When seen from straight on, the Philips maintained relatively even brightness and color across the screen. The left and right edges were just a bit dimmer than the middle, but not to a significant degree in program material. From off-angle, however, it performed worse than the other LCDs in our lineup. Dark areas washed out and lost contrast more quickly and colors shifted more noticeably, toward red and blue, to an extent we haven't seen on many LCDs we've tested recently.

Bright lighting: The matte screen of the PFL6704D performed well in bright light. It didn't reflect as much ambient light as our glass-screened reference plasma display, and fared as well as the other matte LCDs in our lineup.

Standard-definition: The Philips processed standard-def sources better than many TVs we've tested. It resolved every line of the DVD format and the grass and stone bridge looked well-detailed. It also removed jaggies from moving diagonal lines and the edges of a waving American flag. The Philips' noise reduction, on the other hand, wasn't very effective at all at removing moving motes and noise from lower quality images. Finally, it successfully engaged 2:3 pull-down detection.

PC: Via HDMI from a PC, the PFL6704D performed well, delivering every line of a 1,920x1,080 source with no overscan when we selected the PC mode and Unscaled aspect ratio. We did see some slight edge enhancement, however, which prevented it from matching the picture quality of most 1080p LCDs with digital PC sources.

Power consumption: In its default Standard picture mode, the Philips 42PFL6704D relies on a room lighting sensor to help set backlight level, which has a direct effect on power use. Since you can't disable the light sensor without switching out of the Standard mode, we performed our testing, as dictated by Energy Star, with a light level of at least 300lux striking the sensor. In those conditions the Philips still performed well enough to merit an Energy Saver badge, thanks, in part, to its Active Control circuit's backlight fluctuations. For our calibration we turned off Active Control to stabilize the backlight and, as a result, measured significantly worse power consumption--on par with Philips 42PFL5603D "Eco TV" from last year. Compared with newer LCDs like the LG and Toshiba below, the Philips is definitely a power hog unless you engage Active Control.

Juice box
Philips 42PFL6704D	Picture settings
	Default	Calibrated	Power Save
Picture on (watts)	136.8	183.16	91.05
Picture on (watts/sq. inch)	0.18	0.24	0.12
Standby (watts)	0.25	0.25	0.25
Cost per year	$29.66	$39.65	$19.80
Score (considering size)	Good
Score (overall)	Good

How we test TVs.