The Better Mousetrap?

FEATURE

THE BETTER MOUSETRAP?
Texas Instrument's DLP technology will be challenged by JVC's D-ILA

by Peter H. Putman, CTS

It's remarkable how fast the face of the electronic display industry changes. Just five years ago, the first commercially-available single-chip Digital Light Processing (DLP) projector was introduced at INFOCOMM in Dallas. This 20+ pound, 848x600 resolution model represented the start of a major shakeup of display technology, one which has affected every category from small portable projectors to high-brightness, large venue projectors.

Since that time, DLP has become a dominant imaging process. It has matured and outlasted the company that introduced that first DMD-equipped projector (nView), forced one manufacturer of large venue projectors to move away from transmissive LCD panels and get with the program (Barco) and led another to retire its complex CRT/LCD hybrid projectors (Hughes-JVC).

Thanks to a well-coordinated media blitz and marketing campaign, the acronym "DLP" has become an industry buzzword, even if many who use it understand little about how DLP actually works. The coincident development of electronic cinema propelled DLP technology even further with screenings of numerous feature films on DLP Cinema chassis'. DLP projectors have been used extensively at film festivals, as rear projection devices for TV dynamic set pieces, and in every conceivable configuration or staging jobs.

Three companies - Christie Digital (nee' Electrohome), Digital Projection, and Barco - have signed exclusive agreements with Texas Instruments to develop cinema-grade projectors, based on TI's 1.3" 1280x1024 "dark" chip DMDs. A variant of that chip, masking to 1280x720 pixels, has been picked up by three more manufacturers (Mitsubishi, Hitachi, and Panasonic) for use in rear-projection TV sets.

Sony, Panasonic, and NEC are also manufacturing 3-chip large venue DLP projectors in a variety of sizes, resolutions, and brightness levels. Among the desktop crowd, In Focus, Proxima, Plus, Sharp, Mitsubishi, Panasonic, Davis, and NEC are all cranking out ultraportable and desktop single-chip DLP light boxes, while Dwin, Runco, Seleco, Vidikron, and Marantz are targeting the home theater and high-end consumer channels.

TI's avalanche of reflective imaging devices would appear to be unstoppable. But there is another player looming on the horizon.

JVC - the parent company of the now-defunct Hughes-JVC division - has decided to ramp up production of their Digital Image Light Amplifier (D-ILA), a reflective LCD imaging technology that is also described as Liquid Crystal On Silicon (LCOS). The D-ILA is a small (.9" diagonal), SXGA (1365x1024 pixel) imaging device that shows great promise, and is scalable to higher resolutions.

At this year's INFOCOMM show, JVC announced that second-generation 1.3" D-ILA panels with a native resolution of 2048x1536 pixels (yes, you read that right) will be available in 2001. This resolution (also known as Quad XGA, or QXGA) will allow 1080i HDTV images to be shown at full resolution, without the need for scaling or cropping. That's a trick that TI can't pull off presently, and their timetable for introduction of QXGA DMDs is still unknown.

But JVC didn't stop there, showing prototypes of a 1.3" QUXGA panel, which has a horizontal resolution well over 2500 pixels. Adding to the fun, JVC also announced a smaller version of their SXGA panel, shrinking it down to .7" diagonally. An .5" XGA (1024x768) version is also expected in 2001, as is a .5" 1280x720 pixel (16:9) panel. And there's more: A .9" 1920x1080 (16:9) panel is on the calendar for 2002.

Pretty ambitious plans, these. But you can bet that Texas Instruments is taking these announcements very seriously, as their smallest currently available devices (also .7" diagonal) are only capable of 1024x768 resolution. Smaller imaging panels are of paramount importance, as they will make it possible to design compact, high-brightness, high-resolution installation projectors using xenon-based RGB color imaging. That's a market every projector manufacturer is drooling after.

JVC topped off their announcements with a demonstration of the DLA-G3010Z, a three-panel, 1365x1024 resolution desktop/installation projector that weighs just over 14 pounds and is claimed to develop 1300 ANSI lumens. Right now, there isn't any such product available to the professional AV or staging and rental channels, so the DLA-G3010Z (if and when it starts shipping) is a perfect fit for the compact boardroom/conference room installation market.

While JVC's plans for the D-ILA are ambitious, they have a couple of hurdles to jump to be competitive with DLP. To start with, the current optical path in D-ILA projectors isn't too efficient, requiring lamps with 1.5 to 2 times the power level of comparably-equipped DLP projectors.

As an example, the MLA-4000U high-brightness D-ILA chassis needs a 1.6 kW xenon lamp to develop its claimed 4000 lumens, while a comparable, typical DLP product (Christie Digital's Roadster S4) comes close to that ANSI specification using a lamp rated at half the wattage (700 W xenon). And, the Roadster S4 is considerably lighter, tipping the scales at 92 pounds vs. the MLA-4000U's 157 pounds.

Another issue has been video signal quality. Most demos of JVC projectors have looked best with either 1080i HDTV, or an external, high quality video decoder and scalar doing the de-interlacing and color separation of NTSC sources. To compete in the large-venue arena, JVC will need to spend additional time, money, and research into improved video decoding and image scaling.

Of course, DLP projector design is at least a year or two ahead of the D-ILA, and there's no reason to expect that JVC won't be able to get over these hurdles. Plus, they've got an ace up their sleeve, and that is to make D-ILA chips available to any projector manufacturer who wants to use 'em.

There's precedent for this move. Epson - a major LCD projector manufacturer - has done very well selling transmissive LCD panels and complete LCD optical engines to numerous OEM partners. Sony has done the same on a more limited basis with both Sharp and Sanyo. This process helps speed up engineering cost recovery and return on investment for each new D-ILA device that JVC brings to market. Having several OEM partners also means that many more brains are available to work on issues such as brightness, weight, size, and image quality.

Of course, the DMD remains the only 100% purely digital flat-matrix imaging device available, and will probably keep that crown for some time. Both transmissive and reflective liquid crystal panels can be addressed digitally, but the individual LC pixels have an analog response. The advantage of reflective over transmissive liquid crystal is its efficiency - more light arrives at the lens. The switching transistors are part of the LCD panel's backplane, not formed on an imaging surface, so their thickness and transparency isn't as much of an issue..

This is part of the reason that the D-ILA can be reduced in size while increasing in resolution. There are no restrictions on image brightness, either - as long as infrared light (heat) from the lamp is controlled, it's entirely feasible to build a 10,000 lumens D-ILA chassis. The development of a small polarized beam splitter (PBS) will help reduce the size of the optical path.

The tricky part will be improving contrast. Reflective imaging devices usually have a hard time with black levels, partly because of scattered light around the surface and structure of the DMD or LCD panel. Even though the light is polarized (necessary to achieve a shuttering effect and vary brightness levels), some light will always misbehave and go outside the optical path.

Texas Instruments reduced this problem somewhat in their DMDs by coating all surfaces inside the DMD (except the actual mirrors) with a light-absorbing substance, hence the name "dark chip". Such a trick may also be necessary for the D-ILA, as most of the models I've tested to date have fairly high black levels (10% to 15% gray). Dropping black levels is a must if JVC hopes to get the D-ILA into any cinema applications.

What does this reflective technology tussle mean to you, the systems integrator? It's good news, particularly if you have clients that are looking for higher resolution in a flat-matrix projection system. It's a sure bet that Texas Instruments won't have QXGA DMDs available for at least a year. Assuming JVC can finesse the image quality and efficiency issues, the D-ILA could turn out to be a worthy match for the DMD.

Copyright ©2000 Peter H. Putman / Primedia Intertec
This article appears in the November 2000 issue of Sound and Video Contractor.


						FEATURE
					THE BETTER MOUSETRAP? Texas Instrument's DLP technology will be challenged by JVC's D-ILA


					by Peter H. Putman, CTS It's remarkable how fast the face of the electronic display industry changes. Just five years ago, the first commercially-available single-chip Digital Light Processing (DLP) projector was introduced at INFOCOMM in Dallas. This 20+ pound, 848x600 resolution model represented the start of a major shakeup of display technology, one which has affected every category from small portable projectors to high-brightness, large venue projectors. Since that time, DLP has become a dominant imaging process. It has matured and outlasted the company that introduced that first DMD-equipped projector (nView), forced one manufacturer of large venue projectors to move away from transmissive LCD panels and get with the program (Barco) and led another to retire its complex CRT/LCD hybrid projectors (Hughes-JVC). Thanks to a well-coordinated media blitz and marketing campaign, the acronym "DLP" has become an industry buzzword, even if many who use it understand little about how DLP actually works. The coincident development of electronic cinema propelled DLP technology even further with screenings of numerous feature films on DLP Cinema chassis'. DLP projectors have been used extensively at film festivals, as rear projection devices for TV dynamic set pieces, and in every conceivable configuration or staging jobs. Three companies - Christie Digital (nee' Electrohome), Digital Projection, and Barco - have signed exclusive agreements with Texas Instruments to develop cinema-grade projectors, based on TI's 1.3" 1280x1024 "dark" chip DMDs. A variant of that chip, masking to 1280x720 pixels, has been picked up by three more manufacturers (Mitsubishi, Hitachi, and Panasonic) for use in rear-projection TV sets. Sony, Panasonic, and NEC are also manufacturing 3-chip large venue DLP projectors in a variety of sizes, resolutions, and brightness levels. Among the desktop crowd, In Focus, Proxima, Plus, Sharp, Mitsubishi, Panasonic, Davis, and NEC are all cranking out ultraportable and desktop single-chip DLP light boxes, while Dwin, Runco, Seleco, Vidikron, and Marantz are targeting the home theater and high-end consumer channels. TI's avalanche of reflective imaging devices would appear to be unstoppable. But there is another player looming on the horizon. JVC - the parent company of the now-defunct Hughes-JVC division - has decided to ramp up production of their Digital Image Light Amplifier (D-ILA), a reflective LCD imaging technology that is also described as Liquid Crystal On Silicon (LCOS). The D-ILA is a small (.9" diagonal), SXGA (1365x1024 pixel) imaging device that shows great promise, and is scalable to higher resolutions. At this year's INFOCOMM show, JVC announced that second-generation 1.3" D-ILA panels with a native resolution of 2048x1536 pixels (yes, you read that right) will be available in 2001. This resolution (also known as Quad XGA, or QXGA) will allow 1080i HDTV images to be shown at full resolution, without the need for scaling or cropping. That's a trick that TI can't pull off presently, and their timetable for introduction of QXGA DMDs is still unknown. But JVC didn't stop there, showing prototypes of a 1.3" QUXGA panel, which has a horizontal resolution well over 2500 pixels. Adding to the fun, JVC also announced a smaller version of their SXGA panel, shrinking it down to .7" diagonally. An .5" XGA (1024x768) version is also expected in 2001, as is a .5" 1280x720 pixel (16:9) panel. And there's more: A .9" 1920x1080 (16:9) panel is on the calendar for 2002. Pretty ambitious plans, these. But you can bet that Texas Instruments is taking these announcements very seriously, as their smallest currently available devices (also .7" diagonal) are only capable of 1024x768 resolution. Smaller imaging panels are of paramount importance, as they will make it possible to design compact, high-brightness, high-resolution installation projectors using xenon-based RGB color imaging. That's a market every projector manufacturer is drooling after. JVC topped off their announcements with a demonstration of the DLA-G3010Z, a three-panel, 1365x1024 resolution desktop/installation projector that weighs just over 14 pounds and is claimed to develop 1300 ANSI lumens. Right now, there isn't any such product available to the professional AV or staging and rental channels, so the DLA-G3010Z (if and when it starts shipping) is a perfect fit for the compact boardroom/conference room installation market. While JVC's plans for the D-ILA are ambitious, they have a couple of hurdles to jump to be competitive with DLP. To start with, the current optical path in D-ILA projectors isn't too efficient, requiring lamps with 1.5 to 2 times the power level of comparably-equipped DLP projectors. As an example, the MLA-4000U high-brightness D-ILA chassis needs a 1.6 kW xenon lamp to develop its claimed 4000 lumens, while a comparable, typical DLP product (Christie Digital's Roadster S4) comes close to that ANSI specification using a lamp rated at half the wattage (700 W xenon). And, the Roadster S4 is considerably lighter, tipping the scales at 92 pounds vs. the MLA-4000U's 157 pounds. Another issue has been video signal quality. Most demos of JVC projectors have looked best with either 1080i HDTV, or an external, high quality video decoder and scalar doing the de-interlacing and color separation of NTSC sources. To compete in the large-venue arena, JVC will need to spend additional time, money, and research into improved video decoding and image scaling. Of course, DLP projector design is at least a year or two ahead of the D-ILA, and there's no reason to expect that JVC won't be able to get over these hurdles. Plus, they've got an ace up their sleeve, and that is to make D-ILA chips available to any projector manufacturer who wants to use 'em. There's precedent for this move. Epson - a major LCD projector manufacturer - has done very well selling transmissive LCD panels and complete LCD optical engines to numerous OEM partners. Sony has done the same on a more limited basis with both Sharp and Sanyo. This process helps speed up engineering cost recovery and return on investment for each new D-ILA device that JVC brings to market. Having several OEM partners also means that many more brains are available to work on issues such as brightness, weight, size, and image quality. Of course, the DMD remains the only 100% purely digital flat-matrix imaging device available, and will probably keep that crown for some time. Both transmissive and reflective liquid crystal panels can be addressed digitally, but the individual LC pixels have an analog response. The advantage of reflective over transmissive liquid crystal is its efficiency - more light arrives at the lens. The switching transistors are part of the LCD panel's backplane, not formed on an imaging surface, so their thickness and transparency isn't as much of an issue.. This is part of the reason that the D-ILA can be reduced in size while increasing in resolution. There are no restrictions on image brightness, either - as long as infrared light (heat) from the lamp is controlled, it's entirely feasible to build a 10,000 lumens D-ILA chassis. The development of a small polarized beam splitter (PBS) will help reduce the size of the optical path. The tricky part will be improving contrast. Reflective imaging devices usually have a hard time with black levels, partly because of scattered light around the surface and structure of the DMD or LCD panel. Even though the light is polarized (necessary to achieve a shuttering effect and vary brightness levels), some light will always misbehave and go outside the optical path. Texas Instruments reduced this problem somewhat in their DMDs by coating all surfaces inside the DMD (except the actual mirrors) with a light-absorbing substance, hence the name "dark chip". Such a trick may also be necessary for the D-ILA, as most of the models I've tested to date have fairly high black levels (10% to 15% gray). Dropping black levels is a must if JVC hopes to get the D-ILA into any cinema applications. What does this reflective technology tussle mean to you, the systems integrator? It's good news, particularly if you have clients that are looking for higher resolution in a flat-matrix projection system. It's a sure bet that Texas Instruments won't have QXGA DMDs available for at least a year. Assuming JVC can finesse the image quality and efficiency issues, the D-ILA could turn out to be a worthy match for the DMD. Copyright ©2000 Peter H. Putman / Primedia Intertec This article appears in the November 2000 issue of Sound and Video Contractor.