Projector Reviews

The DVP1080 also offers the usual inputs—composite video, S-Video, component (or RGB), and HDCP-compatible DVI. Outputs are DVI and component or RGB on BNC connectors (plus a VGA monitor out). There are also two RS-232 connectors and two 12V triggers.

Since the DVP1080 is also sold separately, it has 12 selectable resolutions for matching nearly any display (see "Specifications"). But here we are interested only in 1920x1080p—the correct match for the Faroudja projector.

Setup
Apart from the caveats concerning the lack of lens shift and the extra setup care that requires, getting a watchable image from the Faroudja projector and processor is as straightforward as with any digital projector.

The most important controls on the projector include Gamma, Color Temperature, Test Patterns (10 provided), Menu Position, Mask, Image Shift, and Lamp Time. There are three selectable Gamma curves, but the projector has been specifically setup to use just two of them: Normal for standard definition sources, and "A" for high definition. The latter has been calibrated to provide better shadow detail on HD material (though with some HD material, Normal worked better for me). The only Color Temperature setting you will need to use is D65—the one that has been calibrated by William Phelps.

The projector's Mask adjustment controls overscan. The minimum overscan depends to a small extent on the source resolution and input, but with a setting of "0" it never measured greater than 1.5%. The other settings—2.5% and 5%--are useful with DVI sources when garbage (such as white lines) is visible at the edge of the frame.

The projector's lamp life is rated for a maximum of 2000 hours. You'll want to keep tabs on the Lamp Time indicator. When you get to, oh, 1800 hours or so, start saving for a new lamp—assuming you haven't already put in a new one to compensate for the inevitable dimming that all projection lamps are prone to over time. At 2010 hours, the projector automatically shuts down until you replace the lamp. A careful internet search will turn up a replacement price of about $400.

The DVP1080 has two menus, Setup and User. Both of them are operated from the remote; the only control on the processor is the Power/Standby switch. The User menu can be viewed on-screen, the Setup menu only on the processor's front panel display.

The controls in the Setup menu you're most likely to use include NTSC Scan Rate, Active Output Type (Analog and/or Digital), Screen Shape, Black Level, HD-H Adjust (a horizontal adjustment to compensate for the sometimes slightly different horizontal image position of SD and HD sources), DVI Input (normally set to 16-235 for video sources), (analog) Output format (RGB or YPrPb), and 1080i-to 1080p Hi-Bandwidth (more about this below). The significance of these controls is fully explained in the DVP1080's excellent owner's manual. I won't go into them in further detail here except to point out that there are also a number of arcane controls in the Setup menu—like RS232 Echo and Baud Rate—that are probably best left in their factory settings and ignored. Or left to an installer to figure out! If you get into trouble during setup, though, there's a control that restores all settings to their factory defaults.

You'll enter the User menu more often. There are no conventional picture adjustments on the projector itself—no Brightness, Contrast, Color, Tint, or Sharpness. These reside in the DVP1080 processor (something to keep in mind should you elect to use a different processor with the projector).

Not all of the adjustments in the User menu operate with all sources; both component and DVI, for example, lock out the Tint control. Picture adjustments are stored independently for different inputs (such as DVI and component) but are global for source resolution. That is, you cannot have different stored settings that automatically engage for 480i, 480p, 720p, or 1080i material as it arrives at the same input. But there is a Profile feature that lets you save specific settings, in memory, for later recall.

According to the manual, the Horizontal and Vertical Position settings in the menu are supposed to move the image, but the Vertical did nothing in my sample, while the Horizontal moved the menu, not the image itself. The Bypass position of the Advanced Color System is said to turn off Faroudja's chroma edge processing circuits, but this had little effect in my tests.

Other menu options adjust the gray level of the side curtains in a 4:3 image on a 16:9 screen (for the DILA1080pHD projector, where burn-in is not a concern, these can be full black), and activate additional test patterns. But as the manual points out, the best test patterns to use for setup are those in the source, not the processor (or the projector).

The small projector remote (dimly backlit) has fewer buttons than most remotes, but they're more than adequate for the job. The separate processor remote, though not illuminated, provides direct access to each input, the video adjustments, and the aspect ratio controls.

Curmudgeoning
Before I get to the important stuff—the picture—a few criticisms need to be cleared out of the way. I've already mentioned the lack of lens shift controls.

In addition, the projector will respond only to 1080p/60fps sources, provided by either a video processor or directly from a source at this native rate. In the event we ever get 1080p/24fps source material the Faroudja projector will not display it. (A rate of 1080p/24fps projected directly would be unwatchable because of flicker, but 1080p/48fps might be tolerable. 1080p/72fps—accomplished by projecting each frame three times—would be even better as it would eliminate flicker and produce smooth motion without the slightly jittery effect of standard 3:2 pulldown).

The projector's fan is quite loud—a feature it shares in common with other JVC-sourced projectors I've experienced. The fan appears to have two operating levels, moderately loud and, when the heat buildup in the projector requires it, louder. When Faroudja designed the projector's case, they did not allow room for extra sound baffling. While the DILA1080pHD is certainly no louder than the Marantz VP-10S1 3-chip DLP (review in progress), it's a tornado (particularly in its high speed mode) when compared to the Sony Qualia 004, the Fujitsu LPF-D711, or, in particular, Sony's little LCD projector, the Cineza VPL-HS51. The latter was as close to silent as I ever expect to (not) hear in my home theater.

To be fair, it must be said that I am referring to all of these other projectors in their low lamp power modes (which are plenty bright on my relatively small screen). And movie sound nearly always masks the sound of the Faroudja; when I watched parts of Shakespeare in Love, with its relatively sedate soundtrack, for example, the fan noise was not bothersome—and the projector was less than three feet from me.

The back panel of the projector, with its exhaust ports, becomes very hot in use. Not hot enough to burn, but hot enough to make you think twice before caressing it, or sitting directly behind it. Small children, however, may find touching it very unpleasant. The best solution to this issue is a ceiling installation.

Two Projectors, Two Processors
I began my time with the DILA1080pHD as driven by Faroudja's DVP-4000 DVD player/video processor. Midway through the test period Faroudja needed to borrow the projector back for an overseas demo. When they returned it something had gone wrong, possibly in shipment. The colors were weird. They exchanged it for another sample. Since the DVP-4000 was no longer being made, we received the DVP1080 processor described above along with the second sample. The comments that follow apply to the new projector and DVP1080, but apart from the DVP1080's superior 1080i-to-1080p de-interlacing the performance of the two combinations were virtually identical. I used a Pioneer Elite DV-59AVi for DVD playback. The Pioneer is one of the few DVD players that will output 480i on its HDMI connector, allowing the Faroudja DVP1080 to do all of the deinterlacing and scaling in the digital domain during testing.