Test Bench: Philips Ambilight 42PF9731D 42-inch LCD HDTV
Unless indicated otherwise, all tests were performed via the HDMI input from an upconverting DVD player set to 720p output.
Color temperature (Warm color temperature, Personal mode before/after calibration): Low window (20 IRE): 6,407/6,657 K High window (80 IRE): 6,814/6,529 K Brightness (100-IRE window before/after calibration): 186/45 ftL
The Philips 42PF9731D has five non-adjustable picture presets, but none of them applies the most accurate color temperature setting, Warm, so I measured and calibrated exclusively in the Personal picture mode, the only one that can be adjusted. Prior to grayscale calibration the set varied by an average of 229 kelvin from the 6,500K standard, a good but not great result. The grayscale was overly red in dark areas and became too blue as it got brighter. After calibration it was more accurate, at 114K average variation. In the default maximum contrast setting I measured a blinding 186 footlamberts peak brightness, but I later reduced it to around 45 for home theater viewing.
The 42PF9731D lacks an adjustable backlight control (not to be confused with Ambilight), which could vary the intensity of the light behind the screen and help produce darker black when set low. Using a checkerboard pattern I measured a contrast ratio of 331 and an average black level of 0.11 ftL after calibration, which is less impressive than from some of the best LCDs I've seen but still respectable. The Philips maintained a stable black level regardless of picture content.
Multiburst resolution patterns from my Sencore HDTV test generator revealed that the 42PF9731D's component-video inputs yielded softer pictures than its HDMI inputs - a normal result - and that 1080i sources were not as sharp as 720p. Edge enhancement was negligible with Sharpness set to zero. Uniformity was worse than on some LCDs I've seen; some screen areas were brighter than others in black and near-black fields at levels of 25 IRE or below, and the bottom right of the screen was visibly brighter in letterbox bars. Off-angle viewing was also poorer than average; the image washed out more than it should have when seen from the sides, top, or bottom, although, as usual, this was more noticeable in dark areas than light ones. Overscan was average at about 4%. Primary color accuracy and color decoding for red and blue were excellent, but green was significantly undersaturated and appeared too yellowish.
I also measured the Ambilight backlight as reflected off a Stewart Grayhawk screen placed directly behind the television. With a full-raster white test pattern of about 6,500K filling the 42PF9731D's screen, the backlight measured between about 5,000K with a 20-IRE screen and 4,200K with a full-brightness 100-IRE screen in the four active Ambilight modes. In other words, the backlight was a good deal warmer (redder) than the onscreen white fields. In the Color mode, which does not vary according onscreen content, I measured about 7,574K in Cool White - commendably close to the 6,500K standard, if still blue and not as accurate as a separate home theater-quality backlight. I could not improve this number by adjusting the controls. The other setting, Warm White, measured an unacceptably red 2,608K.