TEST BENCH

Unless indicated otherwise, all tests were conducted with 1080i-format signals via the TV's HDMI input. COLOR TEMPERATURE (Warm2 color temperature and Movie mode before/after calibration) Low window (20-IRE): 6,972/6,553 K High window (80-IRE): 6,941/6,556 K BRIGHTNESS(100-IRE window before/after calibration): 72/38 ftL

The Samsung HL-R6168W gave mixed results on the test bench. Its out-of-the-box color temperature in Warm2 mode was still blue overall, and while mid-tones were more accurate, very dark areas were noticeably bluer. After calibration, the grayscale came much closer to the NTSC standard of 6,500 K overall, but it still showed variations in mid-tones, appearing somewhat greener from 40 to 70 IRE. Overall, the grayscale fluctuated more than normal for DLPs I've tested. (Calibration needs to be performed by a qualified technician, so discuss it with your dealer before purchase, or contact the Imaging Science Foundation at imagingscience.com or 561-997-9073.)

The television was capable of extreme brightness, maxing out in Cool2 mode with 100% contrast at 159 ftL. It had a difficult time maintaining a consistent level of black regardless of picture content, however, and achieved much better blacks and a better overall contrast ratio when contrast was reduced to around 35%.

According to multiburst resolution patterns from my Sencore VP403 signal generator, the Samsung couldn't resolve all 1,920 horizontal lines of a 1080i-format signal via HDMI, component-video, or VGA. It also couldn't resolve all 1,280 lines of a 720p signal.

Overall, component-video looked less clean than HDMI - a typical result. Color decoding was excellent for both high-def and standard-def sources, with no noticeable accentuation or desaturation of red or green vs. blue. Significant edge enhancement was still visible with sharpness reduced to zero and the DNIe circuit turned on, but disengaging DNIe eliminated edge enhancement. The ramp pattern revealed some striation (contouring), especially toward the darker end, but no worse than most DLPs. Geometry was slightly off, with a minor pincushion effect to the side and slight bowing along the bottom, but focus at the corners was excellent.

Overscan measured an average of 4%, but the entire image was shifted 2% toward the bottom, where overscan measured 6% - more than average. Hot-spotting was a bit worse than usual on most full-raster gray patterns, and a very slight shadow was visible along the edge of the image, but overall uniformity was very good.