Marantz VP-12S3 DLP projector Calibration
For logistical reasons, the sample of the VP-12S3 measured here was not the same sample RD reviewed. Our measurement sample had the standard-throw lens.
Measured with our Leader HD generator at its 1280x720 setting with an RGB output, the VP-12S3 resolved more than 700 horizontal lines per picture height, which means it's easily capable of full 1280x720 resolution. Using the component input, however, limited the resolution quite severely; judged visually (the only way you can judge such patterns on a video display), it would be generous to say that I saw no more than 400 lines per picture height. Clearly, you should use only the RGB or DVI inputs for high-definition programming. This limitation is carried over from the Marantz VP-12S1 and VP-12S2, and is apparently caused by routing any component input, standard or hi-def, through the projector's Faroudja processor. With the proliferation of DVI or HDMI (and sometimes RGB) outputs on HD sources, this limitation is less serious than it was two years ago, when we tested the VP-12S1. Still, potential buyers should keep it in mind when considering this projector.
Overscan averaged about 2.5% with a 720p input (RGB), 4% with a component input (480p). The as-delivered color temperature with the user control set to "3" is shown in the graph. It was actually quite close to correct, with just a small lack of red keeping the result from being anchored almost on top of the desired D65 point. This factory setting would produce a fine picture even without additional calibration—as RD found with his sample—but I decided to check the action of the VP-12S3's auto-calibration lens cap. I ran the calibration procedure twice and got nearly the same result both times: good, but actually a little worse than the out-of-box setup—there was now a little too much blue. Finally, I did some minor manual tweaking using our Photo Research spectroradiometer. The result, shown in the graph, was now very close to D65 across the board. (For the more technically inclined, the deviation was no more than 0.003 in either the x or y coordinates on the CIE color chart, at any point from 20 to 100 IRE.)
Checking the location of the RGB color points, green was quite close to the correct value; Marantz's efforts to filter out excess yellow from the greens has clearly paid off. Blue was marginally shifted toward greenish-blue, and red was pushed just a little too deep into the red region of the CIE color chart. None of these deviations had a subjectively negative effect on the image. In fact, a little extra depth in the reds, while not accurate, is rarely unpleasant, and it wasn't here. Keep in mind, however, that errors in red, green, and blue affect the accuracy not only of those colors, but of all colors—all hues are produced by a display's combinations of the red, green, and blue primaries.
I measured the Marantz's contrast with our Minolta LS-100 light meter, reading reflections off my 80-inch-wide, 16:9 Stewart FireHawk screen, with the VP-12S3's Lamp Mode set to High and the High Brightness setting turned off. In the F3.0 iris setting, the peak contrast measured 1935 (11.61 footlamberts full-field white, 0.006fL video black). At the F5.0 iris setting it measured 2707 (8.12fL full-field white, 0.003fL video black). Note that while the Minolta meter is rated to read values down to 0.001fL, even small errors at these low levels can considerably affect the result. For the 16-square contrast reading (a variation of the ANSI measurement procedure, which normally assumes the use of incident, not reflective, readings), I obtained essentially identical values of 154 (rounded to the nearest whole number) for both iris settings.
A check of the VP-12S3's internal scaler on the Faroudja test DVD unearthed no surprises: jagged edges were conspicuous by their absence, 3:2 pulldown (with an unflagged signal) was quickly captured, bad edits were well handled, and mixed content (film over video and video over film) was cleanly reproduced—Thomas J. Norton
I preferred the F3.0 iris setting; I found the F5.0 position just a little too dim, even on my modestly sized FireHawk screen. The measured peak contrast in F3.0 was impressive on an absolute basis, but it was not as high as I expected from a projector with the new HD2+ chip. The peak contrast on the Sharp projector I reviewed in the May SGHT measured about 50% higher because its blacks were just slightly darker (in its Economy mode, Mid iris setting). Nevertheless, without a side-by-side comparison, I found the subjective contrast of the Marantz just as impressive. (The ANSI contrast readings on the two projectors were nearly the same, which might explain why.)
Regarding the performance of the Marantz VP-12S3, I agree with RD point for point. It is a superb projector in all respects. I found its image to be remarkably consistent over a wide range of good program material. It was exceptionally sharp and detailed, yet never looked edgy or over-etched. The color was first-rate, with greens a little less overtly and unnaturally vivid than they are on most of the digital displays I've seen (though still not quite as subtle and refined as from CRT or projected film). Last but not least, rainbows were nearly nonexistent. I saw one or two during an evening's relaxed viewing, but attempts to generate them voluntarily by moving my eyes rapidly over material that has displayed rainbows in the past were notably unsuccessful. —Thomas J. Norton