Test Bench: Brillian 6580iFB03 65-inch LCoS HDTV
I made the following measurements via the DVI input from an upconverting DVD player:
Color temperature (Warm color temperature, Cinema gamma preset, Night mode before calibration): Low window (20 IRE): 6,353/6,464 K High window (80 IRE): 6,879/6,531 K Brightness (100-IRE window before/after calibration): 119/41.3 ftL
The Brillian 6580iFB03 65-inch LCoS HDTV offers a multitude of picture controls, and prior to adjusting anything I achieved the best results for a completely dark home theater from the Night mode, the Warm color temperature, and the Cinema gamma preset. With these settings the TV achieved a relatively accurate grayscale, varying by an average of just 50 degrees Kelvin, although it did fluctuate more than I would prefer, especially in dark areas. After-calibration results were excellent, thanks to the comprehensive controls - varying by an average of just 21K from 20 to 100 IRE. I'm confident than a qualified calibrator could achieve the same near-perfect grayscale with this TV.
Before calibration, light output even in Night mode was exceedingly high at 119 foot-lamberts. I reduced it to just over 40 ftL for my darkened home theater. The Brillian lacks a mechanical shutter or iris in the lens path that's included on some competing high-end sets to enhance contrast and reproduction of black. But I didn't miss it - the 6580iFB03 achieved an excellent 242.5:1 contrast ratio after calibration, with average black level an inky 0.1225 ftL. And it maintained a stable black level regardless of picture content.
Multiburst resolution patterns showed that the Brillian 6580iFB03 failed to display every line of a 1080i signal via DVI or component video, although the set scaled 720p sources better than most 1080p HDTVs I've tested. The set can accept 1080p/60-fps (frame-per-second) signals via DVI, but it couldn't handle the 1080p/24-fps output from my Sencore signal generator. Edge enhancement was negligible with the Filter in the Low setting and Sharpness set to zero. Geometry was excellent for a rear-projection set, with almost no pincushion effect, although there was a bit of fringing around lines near the corners. Corner-to-corner focus was very good. Fill factor was also superb, and at 4 feet from the screen I couldn't discern individual pixels. Uniformity was better than on other LCoS displays I've tested, with little discoloration in white and gray fields, although I did notice the edges were darker than usual in comparison to the middle of the screen.
The Brillian exhibited the usual amount of stationary screen grain, a subtle pattern evident to some degree in all rear projection screens. Overscan was minimal via all inputs at the 2% setting, although my attempts to further reduce overscan caused the picture underscan slightly, not quite filling the screen. Color decoding was excellent for both high-def and standard-def sources.
I found notable differences between two different samples of the 6580iFB03. The first exhibited a very faint starburst pattern, detectable on uniform gray and color fields, that proved to be the result of a defective mirror. This pattern wasn't apparent in a second sample sped to our lab, but it also showed less depth of black, less accurate color, and inferior uniformity compared to the first, which were said to have resulted from rushed calibration at the factory. My observations are based on the first sample, which Brillian assures me is typical (minus the mirror defect) of 6580iFB03s in the field.