Like many HDTVs, this year the Mitsubishi LT-46144 has three HDMI inputs, all of which are located on the back panel. Unlike many sets, the LT-46144 is certified by SimplayHD, a more-rigorous HDMI compatibility standard than HDMI itself uses to ensure compatibility. Perhaps that explains why the TV seemed to take a lengthy 8 seconds or so to display a picture when we switched between HDMI sources and resolutions--component-video sources appeared in a more reasonable two seconds. Strangely, we didn't notice the same kind of delay switching sources with the WD-65734, which is also SimplayHD certified. For the record, we haven't experienced many HDMI compatibility issues during testing in the last year or so, and when we have, they're almost always the fault of the source device, not the display. Back-panel connectivity is rounded out by a pair of component-video inputs, two AV inputs with S-Video and composite-video, two RF-style inputs for antenna and cable, an optical digital audio output and an analog audio output. An analog VGA-style PC input, common to many HDTVs, goes missing on the LT-46144.
Along the panel's right side, below the column of control buttons, there's a third component-video input that can also accept composite-video sources. Mitsubishi also includes a USB port on the side, although this model is missing the side HDMI port found on many other HDTVs this year.
Mitsubishi's LT-46144 came through with a decent performance, although it was not near the top of the high-end LCD picture-quality heap. Its strengths included good black-level performance and relatively accurate primary colors, but its bluish color temperature is a significant drawback, and a few less-important ones combined to spoil its chances against the best LCDs we've reviewed.
During our standard setup of the Mitsubishi LT-46144 for optimal viewing in our completely darkened theater, we tamed its prodigious light output to 40 footlambert (FTL) and adjusted black levels accordingly. Among the three picture modes Natural was best, although unfortunately it doesn't engage the Low color temperature setting automatically. For its part, the Low setting, the most accurate of the two, did not come very close to the 6,500K standard, and as a result many areas of the picture looked noticeably bluer than we'd like to see. We were unable to adjust color temperature with user-menu controls, so we were stuck with that bluish tinge. We did tweak color balance a bit using the PerfectColor controls, and primary colors were a good deal more accurate out of the box than on previous Mitsubishi sets, so we didn't miss having PerfectTint controls that much. For our full user-menu picture settings, click here or check out the Tips and Tricks section above.
For comparison testing, we set the Mitsubishi up next to a few other HDTVs we had on hand, including the 46-inch Sony KDL-46XBR4 and the 47-inch JVC LT-47X898, both LCD TVs with 120Hz processing, as well as the Pioneer PDP-5080HD, the Pioneer PRO-FHD1, and the Samsung FP-T5084, all 50-inch plasmas. We began by watching Fantastic 4, played over our Samsung BD-P1200 Blu-ray player at 1080i resolution.
First up, as always, we fired up a dark scene, in this case when Dr. Doom takes aim at Johnny Storm with a shoulder-fired heat-seeking missile. The dark recesses of the room, the letterbox bars above and below the image, and the night sky itself looked nice and deep--darker than on the Samsung plasma and the JVC LCD, for example, although not quite at the level of the Sony LCD or the Pioneer PDP-5080HD plasma. Our measurements bore these impressions out. Details in shadows on the Mitsubishi were passable for an LCD; we could see the contours of Doom's mask and his shadowed eyes quite well, for example, although things weren't quite as well defined as with the Pioneer or the Sony.
Color on the Mitsubishi was a mixed bag. The bad news is that, as we mentioned, the grayscale was skewed quite a bit toward blue, so skin tones and other delicate areas looked less realistic than on more-accurate TVs. When Jessica Alba reappears on the Brooklyn Bridge in her underwear, for example, her face and arms appeared somewhat paler and more washed-out than the other sets, and other skin tones followed suit. That went for white or gray fields too, such as the overcast sky or the white of the Fresh Direct van, which all looked too blue. Again, this issue is exacerbated by the inability to adjust the set's grayscale via user controls. Primary colors, on the other hand, looked good, from the bright red of the fire truck to the green of the tree leaves along the entrance to the bridge.
Unlike many 120Hz HDTVs, the Mitsubishi is not designed to smooth out judder, whether during pans, camera movement, or the movement of objects on the screen. We watched as the camera panned up the side of a building, for example, or followed the moving space station, and the telltale stutter of film was clearly visible on the Mitsubishi, which looked the same judder-wise as the sets--naturally, only the Sony delivered that smoothness we've come to associate as the main benefit of 120Hz. We've mentioned before that the smooth action can be disconcerting in many scenes, but having it available--and defeatable--certainly increases the appeal of a 120Hz TV in our opinion. We tried feeding the Mitsubishi 1080p and 1080p/24 sources and again there was no appreciable smoothing effect.
Motion blur, which generally only occurs on flat-panel LCDs, has never been a very noticeable issue for us, but 120Hz is supposed to nearly eliminate it. One area where we have noticed some blur on several LCDs is in a quick-moving ticker on ESPNHD, and sure enough, the Mitsubishi's ticker was as blur-free as that of the plasmas. If blur during fast motion on other LCDs bothers you, then the Mitsubishi's 120Hz processing might be worthwhile.
As we mentioned above, the LT-46144 failed to resolve every detail of 1080i and 1080p sources via HDMI or component-video. This failure in the form of detail loss was nearly impossible to spot in normal program material, however, as the LT-46144 looked every bit as sharp as the Sony next to it and indeed as sharp as any other TV in the room. According to test patterns from the HQV disc, the LT-46144 also failed the test for 1080i deinterlacing, so we recommend using a 1080p source if a good one is available. Of course, that failure was again nearly impossible to spot in normal material and even our standard test from Ghost Rider, where the camera pans over the highway at the end of Chapter 6, looked quite clean.
Uniformity across the Mitsubishi's screen was below average for an LCD. We did notice that the sides of the screen appeared a bit lighter than the middle, and the upper-right and -left corners evinced brighter-still streaks that edged in diagonally a couple of inches. Those streaks were visible in the letterbox bars of the film, but we didn't find that other flat fields, such as the starry space around the station, appeared especially nonuniform. From off-angle, the Mitsubishi was about average; the picture did get a slightly reddish tinge when seen from extreme angles, and like all LCDs, the black areas washed out somewhat from less-extreme angles, but overall the picture didn't suffer as much as we've seen from many LCDs.
False contouring was minimal on the LT46144, but we did detect some edge enhancement that couldn't be removed despite turning the sharpness down to zero. It was nearly impossible to spot in normal material, but graphics, such as the big white-on-blue Marvel logo at the beginning of the film, did betray those enhanced edges on the Mitsubishi and not on the other sets in the room.
With standard-def sources, tested using the HQV DVD disc at 480i resolution via component-video, the Mitsubishi did not particularly impress. Although the TV resolved every detail on the disc, we saw some unwelcome flicker in the color bar pattern. The details in the stone bridge and grass from the Details test, however, appeared relatively sharp. We did notice quite a few jagged edges in the waving American flag and the moving diagonal lines from the test patterns. The Mitsubishi offers three levels of noise reduction, but we couldn't see much difference between them--which wasn't a big deal, since simply turning on the NR to any level worked well to clean up the noisy shots on HQV. The set engaged 2:3 pull-down effectively, although it wasn't quite as quick as with many other sets we've tested.
As a PC monitor, the Mitsubishi LT-46144 cannot be counted among the best LCD TVs. It lacks an analog VGA input, which usually isn't a big deal because HDMI inputs, which can connect to the DVI outputs of so-equipped PCs, usually provide for better PC image quality anyway. When we connected our test PC to the LT-46144, however, we were again disappointed by the lack of a dot-by-dot scaling mode, which caused the set to not fully resolve the 1,920x1,080 source, and introduced overscan along the edges, obscuring the taskbar and leftmost row of icons, for example, on a Windows desktop. Text looked blocky, softer, and even a bit edge-enhanced (with faint white borders on white-on-black and black-gray text) compared to the relative perfection we've seen on 1080p LCDs that have true dot-by-dot modes. Strangely, we also heard a faint, high-pitched hum when the set displayed the full-screen vertical resolution pattern, as well as a quieter hum on some related resolution patterns. We didn't hear the hum in other circumstances, including full-screen white, so it's not a big deal in our book.
|Before color temp (20/80)||7,499/7,556||Good|
|After color temp||n/a|
|Before grayscale variation||+/- 1,059K||Poor|
|After grayscale variation||n/a|
|Color of red (x/y)||0.645/0.319||Average|
|Color of green||0.257/0.624||Average|
|Color of blue||0.151/0.059||Good|
|Black-level retention||All patterns stable||Good|
|Defeatable edge enhancement||No||Poor|
|480i 2:3 pull-down, 24 fps||Yes||Good|
|1080i video resolution||Pass||Good|
|1080i film resolution||Fail||Poor|
|Mitsubishi LT-46144||Picture settings|
|Picture on (watts)||309.58||203.31||N/A|
|Picture on (watts/sq. inch)||0.34||0.22||N/A|
|Cost per year||$103.59||$64.29||N/A|
|Score (considering size)||Poor|