1080i static test patterns to “pixel perfect” 1080p images. However, despite the perfect mapping of each 1920 x 1080 source pixel to the projector’s 1080p native D-ILA panels, single pixel black and white lines in an AccuPel 1080p digital multiburst pattern were rendered with very slightly reduced contrast that was apparently a result of the lens sharpness. However, the red and cyan color tinting that was prominently visible on the RS1, when displaying closely spaced lines (e.g. single-pixel bursts, moving resolution wedges, and zone plates), was substantially reduced in the RS2. I didn’t see any of those “fine line color artifacts” while watching movies, nor on the offi- cial’s striped shirts during NFL games.
Scaling And Overscan
The projector’s scaling produced about 2 (1080p) pixels of outlin- ing around 720p horizontal lines, and 2 to 3 pixels of outlining around vertical lines. The 480i/p scaling produced about 3 to 4 pixels of out- lining along horizontal lines, and 3 to 5 pixels of outlining and slight ringing along vertical lines. However, outlining around digital 480i vertical lines was extremely faint and barely noticeable. The outlining around vertical lines in any format can be removed by reducing the Sharpness control from its center default setting, but that also softens the image. The Sharpness control has no effect on the scaling arti- facts (outlining) along horizontal lines.
I was impressed that there was zero overscan on all formats with digital signals. There was also zero overscan for analog signals, except for about 0.5 percent overscan on the sides of analog 480i signals, and 1 percent on the sides of 576i signals. JVC added an Overscan function for standard-definition signals that produced 2.5 percent overscan on each edge of the image.
The Masking control (0, 2.5, 5 percent) provides 0, 1.25, or 2.5 percent blanking at each edge for high-definition formats. The mask- ing is electronic blanking rather than scaling, so “pixel perfect” map- ping without scaling artifacts is still maintained for 1080i and 1080p formats, but the visible area of the active video is reduced. The zoom lens could be used to restore the size of the image to completely fill the screen if masking is used. However, since electronic masking is not provided for standard-definition signals, those signals would then be projected larger than the screen.
DVD Viewing
Although I now primarily write about high-definition sources in my projector reviews, I also watch standard-definition DVDs in my evalu- ation process. In this case, I was particularly interested in how the RS2 compared to the RS1 on some of the specific DVDs that I used in that earlier review.
The most informative DVD that I revisited was Star Wars: Episode IV—A New Hope (1977). That film looked very impressive on the RS1, but it looked simply stunning on the RS2. I would attribute the primary difference to the RS2’s higher full-field contrast ratio and the increased gamma value that I was able to select. The scenes on Tatooine looked as spectacular as they have on a CRT projector, with just the right intensity of bright highlights and dark shadows. The higher gamma restored depth to the craggy cliffs, which were filled with superb texture and detail. There wasn’t a hint of haze as R2-D2 was sucked into the Jawa Sandcrawler, and the interior scenes pro- duced a revelation of shadow detail that usually fades into a murky darkness. The clarity of the dark features in the background is extraordinary even with bright highlights in the foreground. It is best described as a CRT-like experience. The fades to black weren’t quite as deep as a CRT projector, but in most cases my eyes didn’t have enough time to adapt to fully recognize a difference, and there were no bright screen corners to disturb those moments.
6/7
THEROGERSREPORT .
JVC® DLA-RS2U 1080p D-ILA® Front Projector
Image depth is the result of a complex combination of factors including resolution, sharpness, and intra-image contrast. The latter is also determined by a complex set of factors that include the full- field contrast ratio, the m-ANSI contrast ratio, and gamma. Depending on specific scene content, image depth may be influ- enced more by the visibility of fine detail or the delineation of bright- ness differences. The best DLP projectors have a slight advantage in sharpness and a higher m-ANSI contrast ratio, while the RS2 excels in full-field contrast ratio, and therefore, its ability to perform success- fully with a higher gamma. The latter is seldom an acceptable option with the lower full-field contrast ratio of DLP projectors.
The better full-field contrast ratio and higher gamma of the RS2 improved the image depth throughout Star Wars, including the high- contrast scenes aboard the Death Star, and the scene aboard the Millennium Falcon as Luke practices with his light saber. In the past, I’ve seen a clear image depth advantage for the DLP projectors in those scenes, but the RS2 is now close to parity using the higher gamma. There was also noticeably less visible glare adjacent to bright objects in the high-contrast scenes than there was in the RS-1.
The trash compactor scene is a brutally difficult scene to repro- duce well, and again, the RS2 reminded me of a CRT projector. The scene comes alive with exceptional detail, excellent depth, and not a hint of haze or veiling. The higher gamma produces the right balance of dark detail and bright highlights that are appropriate for this scene.
The RS2’s biggest weakness is that its primary colors are consid- erably more saturated than the standard SMPTE-C color gamut. Fortunately, the hues remain quite accurate, even for skin tones, but the saturation of flesh colors is often too deep. Overall the color palette is deeper and more vivid, and many viewers will prefer that to the standard SMPTE-C colors. But users who want the more accurate SMPTE-C primaries will need to use an external video processor with a color-management system (CMS).
Pierce Brosnan’s skin tones in The Thomas Crown Affair (1999), which is a deeply saturated film transfer, were too deep to appear natural. If an external CMS processor isn’t available, the Color control can be reduced to improve the fleshtones, but in this film that diluted the intensity of other colors more than I would like. In other films, such as Star Wars, I reduced the Color control to tame the flesh col- ors without overly compromising the brilliance or vividness of the overall color palette. Oversaturated primaries affect each film differ- ently, depending on color content. For instance, Leeloo’s hair in The Fifth Element was a bit redder and certainly more saturated than nor- mal, but her skin still looked natural. The green fields of the Shire in The Lord Of The Rings: The Fellowship Of The Ring (2001) were too vivid, and beyond what I consider realistic. Consequently, if color accuracy is a critical priority, I would recommend using an external CMS processor.
The RS2 produces images that are free of dithering and spurious pixelization. Film grain appears very natural, and mosquito noise and other compression artifacts aren’t exacerbated. The AccuPel 1 to 10 IRE gray scale pattern demonstrated excellent near-black linearity and was free of noise with digital signals. However, there was minor random noise in the dark gray scale with analog YPbPr signals. There was no fixed-pattern banding, which was a significant problem for LCoS projectors in the past. The three-panel D-ILA system is completely free of sequential color artifacts, which can still be a problem with some single-chip DLP projectors.
High-Definition Viewing
The Gennum VXP 1080i motion-adaptive deinterlacing has proved to be an excellent performer with 1080i broadcast video. It is excep- tional at eliminating jaggies and line twitter, while producing only minor resolution pumping with movement. Resolution pumping
Widescreen Review • Issue 129 • March 2008 | 7 |
