THE FRONT LINE: NOVEMBER 4, 2004
HAS LCOS BECOME A FOUR-LETTER WORD?
In recent weeks, both Philips and Intel have announced they will stop manufacturing Liquid Crystal on Silicon (LCoS) imaging panels for consumer televisions. To paraphrase the late, great NY Yankees and Mets manager Casey Stengel, “can’t anyone here make this technology work?”
Philips’ announcement was a bit of a surprise since they had introduced a pair of rear-projection TVs incorporating the technology in late 2003. Both models were shown at InfoComm 2004 and Cedia Expo 2004 as viable products, although the image quality (even with HDTV sources) left a lot to be desired.
Of the two ‘exit’ announcements, Intel’s was far more significant, thanks to the bluster and hoopla that accompanied their ‘official’ entrance into the LCoS business back at CES 2004. To hear Intel executives talk then, you’d have thought making LCoS panels was no more difficult than baking brownies.
But even the best bakers burn a few brownies here and there. And things get really sticky when most of the brownies are burned on each sheet you pull from the oven. With LCoS panels, their high fill factor (over 90%) is due to super-thin cell gaps that unfortunately can break down during fabrication.
The company had spent tens of millions on LCoS research and fab development over the past few years before concluding there just wasn’t enough market potential to justify additional expenditures. In other words, they didn’t feel they could sell enough LCoS-equipped rear-projection TVs to make a viable business.
Figure 1: You can forget about seeing this Philips 1280x720 LCoS RPTV
any time soon.
Many of us in the trade press have become so skeptical about this technology
that we often joke, “LCoS is simply another word for bankruptcy”
and that “It’s called LCoS because all of the other four-letter
words were taken.”
Given the demise of LCoS manufacturers DisplayTech and Zight, plus Three-Five System’s spinoff of its LCoS subsidiary Brillian last year and Toshiba’s and RCA’s abandonment of LCoS in favor of Texas Instruments’ DLP technology for their RPTVs, you’d wonder if anyone can make it work.
The news isn’t all that bad. JVC has been hanging in there for several years with many different models of LCoS panels, including 2K (1920x1080) and 4K versions. Sony brands its current LCoS product as an SXRD (Silicon Reflective Crystal Device), and also has 2K and 4K panels in production. Even Hitachi is still making SXGA (1365x1024) panels as well as a complete projector.
At the present time, LCoS (by whatever name you choose to call it) is the only viable alternative for an efficient, compact high-resolution microdisplay to DLP. (Transmissive LCD panels with high pixel density were, until recently, too large to allow compact light engines.)
Figure 2. You can, however, expect to see this Sony 1920x1080 LCoS RPTV
real soon!
When it works, LCoS can make some really nice pictures, albeit somewhat
limited in contrast and black levels. Both JVC and Sony launched new LCoS
RPTVs this year to intense scrutiny, and the JVC offering is already winning
over some reviewers. There are also a few Chinese companies manufacturing
LCoS RPTVs, but the resulting pictures I’ve seen from these products
haven’t been all that impressive.
The expanding consumer and professional demand for true 1080p imaging (whether warranted or not) continues to draw more moths to the LCoS flame. After all, if a RPTV manufacturer wants 1080p imaging and doesn’t want to beat a path to Plano, Texas, LCoS is pretty much the only way to go.
Indeed, this past summer Intel tried to switch gears from designing and manufacturing 1280x720 panels to 1920x1080 panels in an attempt to stay ahead of that perceived market demand for 1080p.
Instead, Intel wound up merely grinding their gearbox. So did Philips. Anyone else want to take a stab at it?
(Editor’s note: For more details on the Philips and Intel announcements plus additional analysis, go to www.insightmedia.info.)