The Sony KDL-46Z4100 is certainly a capable performer, especially in terms of color accuracy and video processing, but we were a bit disappointed in its black-level performance for such an expensive HDTV. The Z4100 didn't match the superb depth of black delivered by the W4100. Also, like its less-expensive cousin, it failed to maintain a consistent black level in certain very dark scenes.
Our standard user-menu calibration improved on the Theater preset, bringing color temperature in line thanks to the user menu controls. There was still some inconsistency in grayscale tracking--the Sony tended toward blue in dark areas, and varied a bit more then we'd like to see in brighter areas--but it was still consistent to most LCDs. We chose the Standard color space to achieve those accurate color points noted in the Geek Box. Click here for our full picture settings.
For comparison we had more than a few sets on hand to place next to the Sony, including from the LCD camp the company's own KDL-46W4100, the Samsung LN52A650, and LN46A550, and the Vizio SV470XVT, along with a pair of plasmas, the Panasonic TH-50PZ800U and the Pioneer PDP-5020FD. We played the Blu-ray version of The Scorpion King on our trusty PlayStation 3 for our main image quality tests.
Black level: We lauded the KDL-46W4100 for its depth of black, so we were a bit surprised to see that our Z4100 review sample didn't get quite as deep as the W4100 in side-by-side comparisons. When Balthazar speaks before the tribe in the firelight, for example, the letterbox bars, along with shadows in the background and on the lee side of the actors' faces and bodies, appeared brighter on the Z4100 than on the W4100 and both of the plasmas; the Samsungs were about the same as the Z4100 and the Vizio appeared lighter. We're not sure why the Z4100 would have worse black levels than its less-expensive cousin, and both have the same specifications, but that's what we observed after carefully calibrating both (along with the rest of the sets in our comparison, as always).
We also noticed that the Z4100 showed the same fluctuating black levels as the W4100, which is the biggest weakness of the display in our opinion. During dark scenes that faded completely to black, the display's backlight would darken abruptly and then brighten again, an effect that was clearly visible and distracting in certain transitions. We didn't notice it during the numerous dark scenes in Scorpion King, but in I Am Legend, at the same place we complained about in the W4100 review (about 12:08 into the film), the fade from darkness up to semidarkness as the camera pushes into the bathroom caused the backlight to brighten noticeably. It happened later than on the W4100, however, and was less noticeable overall.
This issue is entirely dependent on content--darker material triggers it more often--but it's still something that's not an issue at all on most other displays we've tested. We tried a variety of picture settings and modes, including switching to the Custom picture mode as one reader suggested worked on Sony's V4100 series, but nothing removed the fluctuation.
Color accuracy: Nearly perfect primary and secondary colors, along a pretty good grayscale after calibration, led to commendable marks in this category. The Z4100 displayed the olive light brown skin tone of Cassandra emerging from the bath nearly as accurately as the reference Samsung A650 and Panasonic plasma, and without the slight ruddiness of the Pioneer and the Vizio. However, we did notice a slightly bluish tinge in midbright areas, especially with more delicate skin tones. The greens of the plants and the orange and red fruit on the tables, for their part, also looked as accurate as the references, as did the cyan of the pool Cassandra and Mathayus fall into.
On the downside, color decoding was off a bit, albeit in a different direction from the W4100 so we had to increase the color control to get full saturation. Afterward the colors were fairly rich, but they didn't have quite the punch of those shades seen on the displays with better black levels. We also noticed that very dark areas tended toward blue, although not as badly as many displays we've seen
Video processing: Like similar modes available on other displays, Sony's MotionFlow processing is designed to smooth out motion--including the judder or faint stuttering inherent in 24-frame-per-second film material. Judder can be perceived most easily in pans and camera movement, but once you notice it, it seems to pop up everywhere. Some viewers find the smoothing effect desirable, while some think it looks too video-like and even cartoonish in some instances, particularly Hollywood films. We're of the latter camp, but we feel dejudder processing can be effective in some scenes.
The KDL-46Z4100's MotionFlow behaved the same as that of the W4100 as far as we could see. Of the two MotionFlow modes available, Standard introduces less smoothing while High, naturally, introduces more. We preferred the slight judder left by Standard to the ultrasmoothness of High, which often introduced a "halo" distortion that appeared around an object moving against a stationary background. Speaking of artifacts, the Sony evinced no trace of the "triple puck effect" that was still a bit visible on the Samsung A650.
In Scorpion King, we saw examples of the halos everywhere in High, such as when the camera pulls back from Mathayus' face as he's buried in the sand near the fire ants, or when Arpid moves to the rescue against the stationery rocks. Standard was cleaner, and removed the judder from the camera's advance, for example, but the look still appeared too video-like and vaguely fake for our tastes, especially with the unnatural CGI look of the ants themselves. As with the W4100, we preferred the look of the Sony's smoothness to that of the Samsung, even when the latter was in Low mode, but the difference is by no means drastic. The Sony seemed to maintain its rate of smoothing better than the Samsung, where the smoothing seemed to kick in more abruptly.
We preferred MotionFlow's Off position for films, of course, and we also preferred the look of 1080p/24 as opposed to 1080p/60 on the Z4100. Engaging 24-frame output on the PS3 caused the image to judder just the right amount, as it were, which is part of the reasoning behind 120Hz in the first place. During a sweeping pan over some marching minions, for example, the camera movement had a nice, consistent cadence that looked filmlike and not unnaturally smooth, but better than the non-120Hz Samsung A550, for example.
Unlike the W4100 we reviewed, the Z4100 delivered all of the detail of the film resolution loss test from the HQV disc. With regular program material, however, it was impossible for us to see any difference in detail between the two displays.
Update 08-19-2008: We also noticed one other artifact that didn't show up on the W4100 or any of the other displays in our comparison. Darker objects moving against lighter backgrounds, such as the dark hair of the kid whom the King lifts from the basket or of the King himself as he strides across a workshop, showed a brief trailing edge of red. We confirmed the red trails with test patterns, and they appeared regardless of which processing mode we chose. They certainly weren't visible with most moving objects, but we can see how they might become bothersome with certain material. Thanks to Al Griffin at Sound & Vision for pointing this issue out.
Uniformity: Brightness was relatively consistent across the Sony's screen. The left and right sides appeared just a bit darker than the middle in test patterns, but the difference was less obvious than on the Samsung LCDs and invisible in program material. Off-angle viewing was a bit worse, however, with dark areas washing out a bit more and discoloration setting in a bit faster, although it was still better than the Vizio and most other LCDs we've seen.
Bright lighting: The Sony KDL-46Z4100 has the same screen as the W4100, and its effective antireflective properties were apparent in our bright room. Compared with the plasmas and the Samsung A650, it did a better job of attenuating in-room reflections from the windows and light sources.
Standard-definition: With standard-definition sources the Sony fell a bit below average. While it resolved every line of the DVD format, details in the grass and stone bridge appeared softer than we'd like to see. It removed jagged edges from moving diagonal lines and a waving American flag somewhat, although there were still more jaggies than on many sets we've tested, such as the Samsung LN52A650. Sony's noise reduction is still excellent, cleaning up the noisiest areas of low-quality material almost completely in its strongest NR mode, and offering a great selection of NR settings between to deal with higher-quality material. Finally, like the W4100, the Z4100 did engage film mode to remove the moire from the bleachers behind the speeding car on the HQV test disc, it fell out and then back in to film mode quickly, thus failing our 2:3 pulldown test. The results for this test were the same in both Auto 1 and Auto 2 CineMotion settings.
PC: With analog PC sources connected via the VGA input, the Sony performed very well, resolving every pixel of a 1,920x1,080 signal with no overscan and delivering crisp text, although we did see a bit of edge enhancement, even in the special "Text" TV preset, that we couldn't eliminate. Strangely, despite the TV taking the 1080p source well, an error message appeared telling us that the signal was unsupported. Via a digital HDMI connection, PC performance was as perfect as any 1080p TV we've seen, with every detail resolved, no edge enhancement or overscan.
|Before color temp (20/80)||6228/6393||Good|
|After color temp||6454/6531||Good|
|Before grayscale variation||+/- 199||Good|
|After grayscale variation||+/- 147||Average|
|Color of red (x/y)||0.646/0.335||Good|
|Color of green||0.288/0.613||Good|
|Color of blue||0.151/0.053||Good|
|Defeatable edge enhancement||Y||Good|
|480i 2:3 pull-down, 24 fps||N||Poor|
|1080i video resolution||Pass||Good|
|1080i film resolution||Pass||Good|
|Sony KDL-46Z4100||Picture settings|
|Picture on (watts)||268.57||124.71||100.97|
|Picture on (watts/sq. inch)||0.3||0.14||0.11|
|Cost per year||$83.36||$38.83||$31.48|
|Score (considering size)||Average|