Yamaha DPX-1200 DLP projector Testing and Calibration

Testing and Calibration

At 720p, the Yamaha's luminance response with an HDMI source held up strongly to 37.1MHz, the highest frequency luminance burst available from my new Accupel HDG-3000 test-pattern generator. At 1080i, however, the 37.1MHz response via HDMI was sharply rolled off, though the 18.5MHz response was still crisp.

With a component input, the luminance multibursts showed that the projector's response holds up well to 18.5MHz at both 1080i and 720p. But at 37.1MHz, both 1080i and 720p were noticeably rolled off.

The Yamaha's video processing worked flawlessly, as we have come to expect from displays equipped with Faroudja's DCDi de-interlacing. It passed all the tests on the Faroudja test DVD without a hitch. On the Silicon Optix HQV Benchmark DVD (Version 1.4), it performed as well as any processor I have seen on this disc's two tests for jagged edges. And on the HQV's racetrack test, after a fraction of a second's delay to lock on to the source, the bleachers were as stable as any on other processor I've tested to date—and far better than most.

With a 480p input over HDMI, the Yamaha's overscan in the Full position was 1% on all sides. In Standard, it increased to 4% to 4.5% on all sides

You can't use the Color Balance System properly without test gear, but they're effective controls for those equipped to use them. After tweaking, I was able to move the color points in the WRGB mode to the following:

Red x=0.634 (0.635) y= 0.340 (0.340)
Green x=0.304 (0.305) y= 0.594 (0.595)
Blue x=0.152 (0.155) y= 0.070 (0.070)

The numbers in parenthesis indicate the standard coordinates for NTSC. These are the points I aimed for. At present, virtually all HD programming is being produced on HD monitors equipped with standard definition (SMPTE C) phosphors. The separate WRGBYCM mode could be used to store a separate setting for the HDTV coordinates, or to tweak the complementary color points (yellow, cyan, and magenta) along with the RGB.

The calibrated color points above are very close to the standard. The most obvious benefit: the Yamaha produced more realistic greens than we usually see on a digital display.

At the 6500K (Standard) setting, the as-delivered color temperature was very respectable. Calibration brought it down slightly. The Before and After color temperatures are very close, but the actual x,y coordinates of the white points were slightly closer to the D6500 standard after the calibration.

With such a wide variety of brightness-adjustment controls to choose from (Lamp Power, White Peaking, Iris), I ended up making 30 peak-contrast measurements with the iris in its minimum position alone (all on my 78-inch wide, 1.3 gain, Stewart Studiotek 130 screen). The maximum peak contrast I measured was 3987 (11.96fL peak white, 0.003fL video black), with the White Peaking on 3 and the lamp at 90%. The minimum measurement of 2360 (11.8fL peak white, 0.005fL video black) was with the White Peaking at 0 and the lamp at 100%. Opening the iris to its midpoint only increased the peak output by about 1fL in most cases.

With the iris fully open, the White Peaking at 5, and the lamp at 100%, I measured 28.7fL peak white on the Studiotek screen, with a peak contrast ratio of 2393. But the video black deteriorated to a mediocre 0.012fL. A fully open iris would, however, provide sufficient output for a larger screen.

I obtained a pseudo ANSI contrast of 259, measuring the inner two columns of squares on the 4-column (16-square) ANSI checkerboard pattern, with the White Peaking on 5, the iris at minimum, and the lamp at 90%. (For logistical reasons, I took this ANSI reading on the Screen Research CP2 screen instead of the Studiotek used for the other measurements.)