ISF-certified video expert Kevin Miller performed the color temperature calibration on Michael Fremer's review sample. While even the Warm setting was too blue, the set did calibrate nicely (see chart).

Since Michael lives in New Jersey, 3000 miles away from our west coast facilities, I sought out another sample to perform some additional tests. Home Theater magazine (another Primedia publication) had samples of both the 50-inch and 60-inch models (though not at the same time), so I took advantage of that to subject both to further examination. I had intended to look only at the 60-incher, but observed some operational quirks that I was able to study at greater length on the 50-inch model. Apart from screen size, which should produce a brighter image on the smaller screen (and did), both models appear to be identical in both specs and features.

Neither set performed particularly well on the deinterlacing and scaling tests on both the Faroudja and Silicon Optix HQV Benchmark test DVDs. They displayed visible artifacts on the two HQV jaggies tests (rotating and oscillating bars), were tolerable on the waving flag (though some jagged edges were visible there, also), and failed the difficult test that involves a shot of empty bleachers. Both sets passed the unflagged 3:2 pulldown test on the Faroudja disc, and earned a fair score on mixed content (a text scroll with both video over film and film over video).

MF did not comment on any similar deinterlacing or scaling artifacts during the time he spent watching the Sony (on normal video programming, in contrast to special test material), so they are unlikely to be a distraction to most viewers. But buyers who are bothered by the occasional deinterlacing glitch may want to consider combining the Sony with a good progressive scan DVD player or outboard video processor.

Sony's flagship Qualia 004 SXRD front projector, reviewed in the May 2004 issue of the Stereophile Guide to Home Theater (which became Ultimate AV the following month), offered two user-selectable options for that projector's color gamut. (The color gamut is a triangle on the CIE color chart defined by the red, green, and blue color points.) The Normal setting was very close to the NTSC standard, but Wide provided a significantly expanded gamut. That sounds like a good thing, and in fact did produce richer, more vibrant colors. But since most programming is produced to the NTSC standard, reproducing it with an expanded gamut will simply result in inaccurate colors, however pleasing they may be (since all colors are produced by various combinations of those standard red, green, and blue primaries).

The DTV/HDTV color gamut is slightly different than the older NTSC standard, but the color in most production facilities is still analyzed and corrected using video monitors with NTSC phosphors. It's unlikely this will change anytime soon. So, paradoxically, you are more likely to get accurate color—on both standard definition NTSC and high definition (ATSC) program material—from a set that adheres to the NTSC color gamut for all programming!

Sony's new SXRD rear projection sets (both models) adhere roughly to the same wide color gamut that is offered on the Qualia 004. But unlike the Qualia projector, they do not give you any option—you cannot select a more accurate setting. This means that reds will always be a richer, deeper red than the standard calls for, green a more glowing, lime green, and blue a deeper blue. The eye will be more sensitive to the green deviation than any of the others (MF did remark on the lime greens—a design choice very common in today's digital displays).

While the enhancements still result in believable if slightly hyped colors—colors that will admittedly please many viewers, as they did Michael—I would prefer a more accurate color palette, or at least the addition of a Normal option, as in the Qualia.

I expected that the two iris controls (fixed and Advanced) would provide a good contrast ratio when employed together, and they did. But the way they do this was a little surprising. The Advanced Iris is configured so that when it is Off, the peak white level is considerably lower than when the iris is used in any of its three active settings. At the same time, none of the four Advanced Iris settings (including Off) dramatically change the black level. The net result is that the Advanced Iris, when engaged, does increase the contrast ratio, but primarily by increasing the peak whites, not by significantly enriching the blacks.

This was disappointing to me. Sony's own Cineza VPL-HS51 proved to me just how much the image can be improved when an iris that operates dynamically in response to the average picture level is employed to lower the black level. (To learn more about the Advanced Iris, which operates in just such an automatic or dynamic fashion, check out that Cineza review. These irises are showing up in more and more digital projection sets.)

With the 60-inch set, I obtained a peak contrast of 3075:1 with the fixed iris set to minimum (which closes it down to reduce the brightness as much as possible) and the Advanced Iris on Maximum (70.7 foot-Lamberts peak white at 100IRE, video black at 0.023fL) Since Home Theater had obtained a considerably higher contrast number on this same sample (at a much lower black level), I decided to repeat the tests on the 50-inch model, which arrived later, after the 60-inch set had been returned.

The 50-inch model did indeed produce a higher contrast, but largely because of its higher peak output (due to the smaller screen), not significantly better blacks. I measured a peak contrast on that set of 6142:1 (129fL peak white/0.021fL video black), again with the fixed iris at Minimum, the auto iris on High. With the auto iris Off, the peak contrast was 2995:1 (87fL white/0.029fL black).

Surprisingly, the various positions of the fixed iris made very little difference in the set's peak output. When it was opened all the way to Maximum, the light output increased from 129fL (at Min) to 136fL (at Max). The iris was definitely operating—I could see the change in light output as I raised and lowered the setting—but wasn't doing very much to change the measurable light output.

I don't want to leave the impression that the blacks in the Sony SXRD sets aren't impressive. MF certainly raved about them in his review. This, combined with the better black levels reported elsewhere on these sets, suggests that something may have been amiss in the operation of the iris (fixed and/or auto) on both the samples I tested at the time I checked them. If that is the case, we will ask for a follow-up look at one of the sets at a future date. (All of the samples we saw—and likely all of those reviewed elsewhere to date—have been early- or pre-production units.)

I also checked the resolution of the 50-inch set with my AccuPel test pattern generator. The only significant anomaly I saw was in HDMI 480i and 480p, where the luminance (black-and-white) response on the highest frequency multiburst (6.75MHz at 480i, 13.5MHz at 480p) was essentially gone; the bursts were a solid, undifferentiated gray. Oddly, the 480i and 480p component response looked fine up to their respective maximums. The HDMI luminance response also extended up to the maximum burst frequency in both 720p and 1080i (37.1MHz).

Finally, the set's overscan, measured with an HDMI input, ranged between 2.5% and 4% in the Normal overscan setting depending on the resolution of the source. It peaked at 3% at 720p and 1080i, and 4% at 480i and 480p.—TJN