Like all CRT-based projection televisions, the WS-55315 requires some maintenance to produce its best picture. Aside from the color decoder and the grayscale calibration, we also had to fix the convergence (alignment of the three CRTs to eliminate blue and red fringing) and tweak the geometry to get straight lines. Even afterward, the picture didn't seem as sharp as that of the LCD-based Sony KDF-50WE655 we compared it to. It also had a much more noticeable hot spot (brighter area in the middle of the screen), less-consistent color and brightness across the screen, and a narrower viewing angle; sitting more than a few degrees to the screen's left or the right edge caused brightness to decrease dramatically. Again, these issues are common to all CRT big screens, and in the case of viewing angle, the Mitsubishi seemed a bit better than some models we've seen.
It also shared with other tube-based sets the ability to reproduce dark scenes just as well as bright ones. We dimmed the lights and put in Alien, and the deep, inky black of space was quite satisfying. In our darkened room, the KDF-50WE655 appeared significantly lighter by comparison, which lent the pans across the darkened spaceship halls a more washed-out quality. Showing dark scenes with the lights off, the Mitsubishi was in its element. One nitpick: the set's mediocre DC restoration--how well the television is able to maintain a consistent color of black regardless of the brightness of the rest of the image--prevented us from getting an even inkier black.
Returning to high-def, we watched the end of a preseason NBA matchup on TNT-HD. Here the Mitsubishi appeared distinctly less impressive than the Sony. The edges of text on players' jerseys looked a bit softer, the graphics were a bit less stable (especially on paused screens), and we saw a bit more video noise--which looked like a faint veil of fine snow--on the backs of the seats. The difference in black level was also much less noticeable inside the well-lit arena. Tests using the Accupel HDTV signal generator showed that the Mitsubishi's 1080i image was indeed less detailed than that of either the KDF-50WE655 or the KD-34XBR960.
| TEST | RESULT | SCORE |
| Before color temp (30/80) | 8,052/5,998K | Poor |
| After color temp (30/80) | 6,502/6,495K | Good |
| Before grayscale variation | +/- 681K | Average |
| After grayscale variation | +/- 248K | Average |
| Overscan | 5% | Average |
| Color decoder error: red | 20% (0%) | Poor |
| Color decoder error: green | -5% (0%) | Good |
| DC restoration | Gray pattern visible | Average |
| 2:3 pull-down, 24fps | Y | Good |
| Defeatable edge enhancement | Y | Good |
What You'll Pay
- Set Price Alert