Sharp XV-Z12000 Mark II DLP Projector Tests and Calibration
Using an Accupel signal generator with both DVI and component signals, the Sharp XV-Z12000 Mark II showed negligible overscan at any resolution, and the kind of near-perfect geometry we expect from digital. While I saw no serious color fringing issues on a white crosshatch, I did see a slight softening on some of the intersecting lines near the center of the screen. Some lines were clean and perfectly strong and two pixels in width. The soft lines showed a gray, ghostly outline into a third row of pixels. This is the only evidence I found to corroborate my impression that the image was occasionally not as sharp as I've seen. Given that these deviations were never more than a pixel in width, this is more of an observation than a criticism.
Looking at the Accupel's DVI output with luma and chroma multibursts at 720p and 1080i revealed generally excellent performance. At 1080i resolution was very strong at 18.5MHz, which is two pixels in width, but noticeably softer at the 37.1MHz, single-pixel width area of the bursts. This essentially means that the highest frequencies of 1080i signals will be rolled off slightly resulting in a slight softness compared to 720p signals, which retained strong resolution even at the single-pixel width areas of the bursts. I'd note though that these are static patterns, and that 1080i program material is obviously in motion, and that Sharp's excellent deinterlacing makes 1080i material look very crisp and dimensional, evening the score some.
Switching up to the same battery of tests with component video at 480i and 480p, it quickly became apparent why component 480p looked so good. The bursts confirmed that the Sharp's 480p inputs are capable of resolving all of the detail in the signal right to the format's limits with no visible noise to speak of. Very clean. 480i was nearly as good, but with some noticeable softening at the very highest frequencies.
As mentioned in the review, the Sharp Z12000 Mark II showed very good "out of the box" grayscale tracking, which I measured in the Medium iris mode, Economy Mode off, Standard Gamma Position, the CMS set to standard, and the Color Temperature set to 6500. The grayscale tracking was exceptionally linear around 6900K. After calibration, the color temperature never deviated more than 250K from the 6500K standard, and the x/y coordinates were within 0.003 from 20IRE to 100 IRE of the D6500 standard of 313/.329. In short, I've never seen grayscale tracking this good elsewhere, let alone better.
Because this is a follow-up—though a very thorough one—to the original Z12000, and because the color temperature was so linear and accurate, particularly after calibration, we aren't including our usual graph here.
After Shane finished his report, I measured the peak contrast ratio of the Sharp with our Minolta LS-100 light meter. There were about 100 hours on the projector at this point, and all measurements were taken on my 78-inch wide, Stewart Studiotek 130 screen (1.3-gain). Keep in mind that a larger and/or lower gain screen will drop the peak white level significantly (which will be true for any projector).
With the Economy mode turned off (lamp on high) and the iris set to Medium, the Sharp's peak contrast measured about 2400:1 (12 foot-Lamberts 100IRE full white field/0.005fL video black). With the lamp in its low setting (Economy) it still measured 2400:1, though both the peak white level and video black levels dropped significantly (9.6fL/0.004fL).
Things got interesting when I switched the iris to its lowest (High Contrast) setting, with the lamp on high (Economy mode Off). The peak white level only dropped slightly, to about 11fL, but the video black dropped to an astonishing 0.003fL, for a peak contrast ratio of 3660:1. And with the Economy mode On (lamp low) this increased to 4272:1 (8.543fL/0.002fL). These are the lowest black readings I have ever measured from a digital projector on my Studiotek screen. The actual peak contrast might actually be a bit less jaw-dropping (though still very good), as the Minolta meter is only specified down to 0.001fL, and I do not trust its accuracy to more than about ±0.001fL at such low light levels. Nevertheless, the Sharp has the best black levels I have yet measured on a digital projector, combined with a comfortable peak white output (on my screen).
What happens when you open up the iris all the way, with the projection lamp on high (Economy mode Off)? The light output on my screen increased to a peak of 27.13fL, while the black level soared to 0.022fL, for a contrast ratio of 1233:1. But contrast ratio and black level are less important in a room with ambient light—which is the only situation in which you are likely to use this maximum output. In my opinion, Sharp needs to provide another iris setting, somewhere between this peak output and the current Medium setting. This would allow more flexibility for lower gain and/or larger screens, plus provide a margin for the inevitable fading of the projection lamp (which in my experience starts to become significant with most consumer projectors at under 500 hours of operation).