Google
 

 

 

HDTV Wars

Microsoft discovers Howdy Doody, and gets Goofy

Francis Vale


"The computer industry must think we are plain stupid." -- Robert Graves, the head of the Advanced Television Systems Committee, speaking at the NAB (National Association of Broadcasters) convention in Las Vegas, April, 1997.

"We are striking up a good relationship with them." --Robert W. Stearns, Senior Vice President, Compaq Computer Corporation, speaking about the TV industry, at the same NAB convention.


Such contradictory statements from this NAB convention prove that if the computer industry has accomplished anything, it can readily turn reality on its head. This questionable capability probably makes the industry a natural for creating TV sitcoms. But factor in the $150 billion that was up for grabs at the Las Vegas NAB, and you get an over budget war movie This huge sum reflects the replacement costs for scuttling and replacing every analog TV in America--all 230 million or so--with new digital sets within the next ten years.

Enforcing this analog TV Bataan death march is the FCC, which is mandating that all analog terrestrial broadcasting in America must forever be off the air by 2006. The FCC also wants HDTV coming over the airwaves and into the nation's top ten TV markets no later than 1998. And by November, 1999, more than half of all viewers should be able to get the new digital broadcasts. Thus, the $150 billion dollar NAB conference question: From what company and industry will you buy your new digital TV set? And what will it look like?

The giant consumer electronics companies, like Sony, Toshiba, RCA, and Mitsubishi, have firmly staked out their new HDTV turf. They firmly believe consumers will demand the excellent picture quality of a true "high definition" 1,080 line x 1,920 pixel digital TV set. They certainly have good grounds for thinking this way. The tremendous, rapid success of digital, direct satellite broadcasting (DSB) is mostly due to its offering significantly better video and audio quality than regular TV and cable broadcasts. But 1,080 line HDTV more than doubles DSB's already impressive picture quality.

To put these line numbers into context: your current NTSC analog TV can support up to 525 lines of resolution, but that's pure spec'manship. A S-VHS tape achieves 400 lines; a laser disc, to which DSB is roughly comparable, can get 425 lines on the screen; and even a DVD can only put out 483 lines (not 500 lines, as the press usually states.) Current terrestrial analog broadcasts, which max out at 330 lines, as well as cable transmissions, produce significantly worse video and audio than any of these foregoing systems. As a consequence, digital high definition TV can be a mind blowing experience when you see it for the first time on a big screen.

Although the TV and consumer electronics industries are totally committed to preserving the current interlaced broadcast format, the computer's progressive scan format will also be available on their new HDTV sets. And in order to service the set's likely standard Internet connection, this device will offer some measure of PC-style interactivity. The question is, how much? Recent Internet-enabling set-top systems, like WebTV, are probably what the consumer electronics makers consider to be a reasonable compromise.

In complete contrast, the computer industry has decided that what you really want in your living room is a MS Windows-based, "PC Theater" box hooked up to a big screen HDTV. These new consumer-targeted units typically will have MS Windows hiding behind a more user-friendly interface shell. This user interface will look somewhat like the extra cost, highly informative, on screen TV program guides that are now broadcast during the TV's vertical blanking interval (e.g., Star Sight). But as the PC Theaters come Internet-ready, these Windows-layered shells seamlessly display interactive and Internet/Web "TV programming" selections, along with letting you know what time Seinfeld and The Simpsons are on.

These new PC Theaters will offer about 80% of the capabilities of a full function PC. Another PC Theater compromise is its screen display when in "PC mode" (e.g., when viewing Web pages), which will offer no more than 640 x 480 VGA resolution. This is because owners must be able to read text on their PC Theater's screen at a distance of more than 10 feet. Any higher resolution, and the text becomes so small they can longer read it. So, these consumers will only be getting the resolution of an antique PC.

But if getting 80% of a PC's functionality and restricted display quality don't tickle your fancy, then you can also expect to see a new generation of HDTV-capable, "Broadcast PCs." When connected into a big (and very expensive) VGA and above monitor, these full function PCs might surpass anything on the HDTV market. Significantly, both the PC Theaters and Broadcast PCs only process HDTV's progressive scan format. Interlaced HDTV is not supported at all. Microsoft, Intel, and Compaq absolutely believe that the PC's rich, interactive functionality (even just 80% of it) is more than ample compensation for any diminished HDTV picture quality.

In addition to chucking out the TV industry's much beloved interlaced formats, high definition 1080 line broadcast support is also out the TV PC window. Whether it be a PC Theater or a Broadcast PC, they initially will only display 480 line HDTV pictures. (This 480 line figure is often incorrectly expressed as being 525 lines in press accounts.). However, when displaying 35mm films, the PC's TV resolution goes up to 720 lines. But if Compaq's new PC Theater is any indication, when the big TV's scan rate is upped to 720 lines for 35 mm film, it's still only equivalent to 540 TV lines. This is because its computer progressive scan is happening at the VGA's 480 pixels (720 x 480 = 540 equivalent TV lines.) Thus, all of these TV PC numbers are considerably less than the 1,080 line high definition pictures the TV industry says that consumers want.

This product perception fault line is what divided these colossal industries at the recent NAB convention. Once this marketing crevasse fully shears in 2006, it's either the TV industry that splits off and sinks into the ocean of obsolescence, or it's the computer industry which dies gasping for new market air.

Make no mistake, the digital TV war wagons are already rolling. Come 1999, most of this country's 1,600 TV stations will be offering terrestrial digital TV programming. Moreover, some within the cable industry have also pledged support for the ATSC formats, (which ones are another story). To facilitate this rapid transition from analog to all digital, the FCC has "loaned," at no cost and under much TV lobby pressure, a second UHF frequency channel to the broadcasters. This new 6-MHz spectrum slice is to be dedicated to digital broadcasting.

This airwave "gift" by the FCC caused much gnashing of teeth in numerous quarters, as many said (including former senator Bob Dole and senator McCain) that these valuable channels should have been auctioned off, like the PCS spectrum was a while ago. (That many of the PCS auction "winners" are either now in default on their payments to the government, or are significantly delaying their spectrum rollout for lack of development funds seems to have been lost on these "No-Free-TV" sloganeers.) In 2006, this second channel becomes the primary TV broadcast medium, as the analog channel is permanently switched off

However, the broadcasters are getting quite piggy. They are now maneuvering like mad to make sure that they never have to give back their current analog TV airspace to the government. In a nutshell, they think if they can get legislation passed which more or less states that so long as one TV set in America is incapable of getting one of these new digital signals (e.g., granny refuses to part with that ancient Philco) then the broadcasters should be able keep .BOTH the analog and digital spectrum -- without paying a dime for either.

On this new digital TV channel, the broadcasters have a choice of transmission formats. They can either broadcast five simultaneous digital channels of standard definition, 525 line TV programming, or, they can use all of the channel's available bandwidth for transmitting just one 1,080 line high definition broadcast.

Of course, because it's an all digital transmission system, it can be used for much more than just broadcasting HDTV reruns of Gilligan's Island. The TV broadcasters can also utilize it for delivering new digital services, like paging, e mail, lightning fast web page downloads, etc. For that manifold use reason, the PC industry vehemently argues why waste all that new and valuable digital bandwidth for transmitting just one pretty picture? Better, it says, users should only receive five lower definition channels, and thereby gain new types of interactive, digital services. And thus, the computer industry says its data-friendly formats and richly interactive TV PCs are the best and only way to go.

Obviously, the TV broadcasters, consumer electronics industry, and Hollywood tyros don't believe that consumers are ready to convert en masse from being couch potatoes to mouse potatoes.

So how will the consumers vote their pocketbooks on this 230 million set ballot: Will the "eyes" have it, and thus vote for proposition 1,080? Or will buyers elect new services that are less than picture perfect? But before you pull that lever, consider the following: All 35mm film programming, which includes all movies, some TV shows, and 80% of all prime time TV programming will automatically be broadcast by the new digital TV equipment in high definition, 1,080 line format, in either progressive or interlaced mode. Moreover, you can expect that premiere sporting events, which always offer the best quality pictures, will be broadcast in the 1,080 line format. (Can you imagine the Super Bowl producers and its advertisers settling for just "standard definition" viewing?)

These impressive stats make the HDTV game look increasingly like an ATSC/GA blow out. Or will Mary and Joe Consumer cause an upset, and decide that an MS Windows PC Theater is worth the supreme Super Bowl sacrifice? But how will they view any of this high definition content on their PC Theater if its Windows system is only designed to process 480/720 line images?

Worse yet for PC Theater owners, the TV industry has decided that it will primarily be broadcasting its 1,080 line images in the computer-foreign, interlaced format (usually called "1080-I"). The new PC Theaters may be going dark all over the country. And to make sure that it drives the nail into the image-challenged PC TV coffin, the Advanced Television Systems Committee (ATSC), which sets broadcast technical standards, intends to put in place a branding program. New HDTV sets can only bear the ATSC seal of approval if they are certified to receive digital transmissions in all 18 formats available under the "standard." Which means, of course, PC TVs need not apply.

This branding can become a powerful consumer selling point. Minimum wage, low tech sales people (i.e., the vast majority), will have no problem simply pointing at a big flashy sticker, and saying, "It's OK to buy it. It's got Atsee approval." Even if all the PC TV vendors put easy to use interfaces on their systems, like Compaq's new PC Theater (which uses "Star Sight" TV menuing derivation), it's still no guarantee of an easy sale. In an age in which pictures of hamburgers replace confusing numbers on cash registers, minimally knowledgeable, ever job-changing sales clerks will have a tough time coping with the technical demands of a TV PC. Finally, when Mary and Joe go cruising down the department store aisle, which do you think is going to stop them in their tracks: A stunning high definition movie on an ATSC set, or e mail scrolling across a PC Theater's wide screen?

Has the TV industry called it right? If so, what could possibly be driving such totally contrary behavior on the part of the computer industry? Easy answer: Greed, the same thing that's driving the TV folks. For he who wins the HDTV viewer sweeps gets a King Midas bundle. If you add everything up, including the $150 billion for new sets, the new content creation tools, the new content licensing, the new software system sales, as well as the revenue from new digital services, then the whole HDTV enchilada is probably worth close to a trillion dollars come mid-21st century. With such fantastic sums at stake, did Gates and Grove have any choice other than to throw all business caution to the winds, and just got for it?

Moreover, having conquered nearly all the computer desktops on the planet, where else can these two earnings-hungry companies expect to find such an explosive new market? Forfeit this HDTV fight, and Intel and Microsoft, while far from becoming paupers, will see their visions of total computer global conquest disappearing right in front of their standard definition eyes. HDTV has thus become high stakes poker, and Microsoft and Intel are gambling heavily that they can bluff and outmaneuver the TV industry, and thereby gain complete control over this rich new medium.

And should the TV people lose their massive bet, a supreme irony will have come full circle. Originally, this new digital system was, for all intents and purposes, a TV industry smoke screen. They never intended for this thing to happen in the first place. Ten years ago, in 1987, the U.S. government was making some serious noises about giving away valuable, vacant UHF space to mobile communications companies. In a highly self-serving move, the National Association of Broadcasters said, oh no, you can't give those previous airwaves to those silly paging companies, not when we plan to use those unused UHF channels for "High Definition Television."

Huh? This was the first time almost anyone had heard of this new industry plan. The NAB lobby went right down to the mat with Congress-- even so far as bringing in the ill-fated Japanese HDTV (analog) Muse system for a demonstration on Capitol Hill. The result of this carefully orchestrated, "The Japanese are Coming! The Japanese are Coming!" flag waving paranoia was that the communications companies were effectively blocked, and left to go looking elsewhere for their vacant airwaves. For the next three years, two dozen entities, from all corners of the broadcast industry, trotted out one sad analog dog and HDTV pony show after another. Digital broadcasts weren't even considered, as all the experts said such an HDTV system was technically impossible. Besides, the TV industry was in no hurry, as it had already achieved its airwave-glomming aims.

However, in 1990, from out of nowhere, General Instruments, one of the biggest makers of consumer cable TV boxes, shocked everyone by demonstrating an all digital HDTV system. (Actually, it was developed by MA/COM, which was later acquired by GI.) Stunned into silence by their collective ignorance, and the high quality of General Instruments' pictures, all the previous naysayers were left with no choice but to immediately jump on the digital TV bandwagon, including the Europeans and Japanese. Out of this Mad TV grab bag came the Grand Alliance (GA) consortium, consisting of Zenith, Philips, Thomson, General Instruments, the Korean-majority owned Zenith, M.I.T., the David Sarnoff Research Center, AT&T's former Bell Labs (now Lucent Technologies), Texas Instruments, and the General Electric-owned NBC.

In 1993, after much work and effort, the GA had a viable digital TV system specification. It was comprised of 18 different formats which embraced interlaced and progressive scanning devices, supported 525 line and 1,080 line resolution, used Dolby Digital (formerly called AC-3) encoding for audio, and sported a 16:9 screen aspect ratio (your present TV has a 4:3 ratio.) Lastly, image delivery rates of 24, 30 and 60 frames per second were supported. Overkill perhaps, but at least there was something there for everybody.

Now enter Reed Hundt, the new (and now departing) chairman of the FCC who had come in under Clinton's Democratic regime. Politically speaking, he had no real interest in seeing the TV industry's Reagan-Bush-inspired HDTV system succeed. It quickly became clear that the Grand Alliance had a powerful new enemy at the FCC. Moreover, the U.S. Congress was once again getting into the HDTV act, further clouding the digital TV picture. As if all this messy politicking wasn't enough to sloppily stir the HDTV pot, in the summer of 1996, Gates, Grove, and Hollywood descended en masse on the nation's Capital.

In the purported interests of consumer price protection, Gates and Grove maintained that supporting the interlaced format on PCs would be much too costly. Moreover, putting in those 2 megapixel, frame-buffer memory chips needed for processing the High Definition formats would send PC prices skyrocketing. In truth, this computer industry ploy was no less cynical than the wily maneuverings of the TV industry when it body-slammed the mobile communications companies back in 1987. According to Philips, only about $50 would be added to the total PC build cost if the computer makers were to support all of the GA's 18 formats. This translates into about $125 to $150 at retail. Moreover, like all semiconductor prices, it is expected that GA chip set costs will quickly plummet to just several dollars in as many years.

Thus, consumer PC price protection had nothing to do with Gates' and Grove's FCC pleas. Rather, it had everything to do with dumping that bandwidth-gobbling, 1,080 line high definition format. Once dumped over the regulatory side, Microsoft and Intel could get down to business. Huge new markets lay ahead in providing new systems and software products that utilized those five progressive format data channels. (Indeed, much of the well publicized hue and cry about giving away that "free" HDTV channel was a similarly cynical PR hustle. The cash-overflowing Microsoft surely would have bought up some of those precious airwave licenses.) And finally, why should Grove and Gates give the rival TV industry a break, and let it make money off of their systems?

At the same time that Gates and Grove were weeping about poor consumers, the Hollywood contingent was crying to Congress and the FCC about the movie-butchering, 16:9 aspect ratio set forth in the GA's HDTV spec. Spielberg, et al, clamored for a more director-friendly, "image-pure" 2:1 aspect ratio. Once again, these were crocodile tears. Movies are always shot so they can be displayed in several aspect ratios at different types of movie theaters, from the shoebox-sized foreign movie houses to the ultra big screen Star Wars jobs. Once more, the reality distortion field known as HDTV was transmitting full blast.

But when Al, "Information-Highway," Gore threw his support behind the computer folks, coupled with Hundt's open animosity to the TV industry, it was easy to see that the GA's plans and hard work were about to become totally unraveled. After a series of Byzantine industry/Washington meetings in October and November of last year, the FCC finally said OK, here is what we'll do: There will be no HDTV standards at all! Choose whatever frame rates you want, go with either progressive or interlaced, and you decide what screen aspect ratio you like. The only things you have to keep are:

1. Digital video compression based on the ubiquitous MPEG-2 system, and using its Main Profile parameters. The profile parameters define whether the picture is coming at you in high definition mode (MPEG-2 Main Profile@High Level) or in some other definition mode, like 525 line standard. "B-Frames" (bi-directional frame motion compensation techniques used for improving picture quality) must also be supported for live video capture.

2. The Zenith-developed VSB digital modulation transmission system for broadcast and cable. This technique is intended to give broadcasters the widest possible coverage area, yet without clobbering existing analog signals, or fouling up digital signals.

3. The packetized transport system for carrying video, audio, and data

4. And lastly, Dolby Digital, six channel surround sound (front left and right, center, two rear surrounds, and a special subwoofer low frequency "effects channel" (now, when Arnold blows something up on HDTV, the neighbors can share in your viewing experience.)

With this non-ruling, the FCC essentially said, "We will let free market forces do their thing, and may the best digital TV technology win." This FCC pronouncement was government farce at it's best. The TV industry, spearheaded by the Grand Alliance, had absolutely no intention of abandoning a decade's worth of very expensive R&D. In response to the FCC's non-decision, the GA members said fine, we will just go ahead as originally planned. We will build TV sets and transmission systems that support all of the original 18 formats. And the computer industry can do what it damn well pleases. Which, naturally, it quickly set out to do. As a result of all this maneuvering, the HDTV moniker, which implied just a high quality picture, has now given way to just "D-TV," digital TV.

Is it any wonder, then, that the ATSC's Graves and Compaq's Stearns were talking at such cross purposes during the 1997 NAB convention? Each represents two seemingly irreconcilable business agendas. Their intractable industry positions, plus all the recent political decisions mandating quick to market D-TV broadcasts, made the Las Vegas NAB a swirl of excited confusion. Everyone was revved up, and raring to go. But go where, and how? Many of the TV stations were plain puzzled at how quickly this whole D-TV thing happened, accustomed as they were to delay, delay, delay. The digital sky had finally fallen, and many still couldn't quite believe it. But where there is fear, uncertainty, and doubt, there is also a computer industry opportunity.

Consequently, at the NAB convention, Microsoft, Intel, and Compaq made an aggressive pitch for progressive-only, no high definition pictures to the TV broadcast industry. Bottom line, said these Computer Capos, either get on board with our PCs as TVs, or go ahead and sleep with the fishes as you drown tied to that ATSC boat anchor.

Despite Compaq's PR spin to the contrary (see article beginning) this troika's offer you can't refuse mostly fell on deaf, if not incredulous, NAB member ears. Rebuffed in Las Vegas, Gates and Co. thereupon rolled up their sleeves, and set about to conquer the incredibly powerful NAB membership, including the rich GA companies. And like so often before, Intel also sidled up alongside Microsoft for the upcoming fight. But this time, Compaq also joined in, making it a triple threat effort.. Their collective response is HD0--High-Definition Digital Television Level 0. (Is the consumer really ready for this acronym overload so beloved by the computer industry?)

The thrust behind HD0 is nothing less than to totally trash ten years work of the GA, and in particular, its interlace and 1,080 line standards. HDO is the first of three PC industry salvos against the ATSC spec. Level "0" does away entirely with 1,080 line, high definition interlaced images, and replaces them 60 frame/sec, 480 line, progressively scanned pictures, with B Frames. However, HD0 also offers a 720 line, 24 frame/sec, progressive format expressly for film sources, but no B-Frames. If HD0 goes over with consumers, then in the year 2000, HD1 arrives, which will up the 480 line resolution to 720 lines, and boost the 720 film format to 1,080 lines -- but still having no B-frames. Finally, come 2001, HD2 makes its debut (HAL, where are you when we need you?), and all PC TV broadcasts are then sent out and received at 1080 lines.

HD0 cuts out the bandwidth gobbling "High Level" MPEG-2, and replaces it with "Main Level" encoding and decoding. The main level coding scheme, in addition to its purported cost savings, only consumes about half of the digital channel's 19.3 megabit/sec pipe, leaving ample room to download the Encyclopedia Britannica while watching the new PC-interactive version of Jeopardy. In addition, all of the computer industry's proposed HD levels would support IP multicasting embedded in the MPEG stream, making data services broadcasting possible. The HD0 standard will also offer content developers new ways to produce and market their products, provide text with video for story background information or on-line chat, as well as supply hooks for Internet commerce sites to tap into an ongoing broadcast. Assuming this new standard catches on (which is still very much in doubt among NAB/ATSC-backers), the first HD0-equipped PCs should hit the streets by the end of 1998.

Interestingly, a HD0 card and its digital TV tuner is expected to add about $100 to the retail cost of a PC; not very far off from what the cost would be if PC makers were to provide full GA format support. However, all of this money now goes into Intel's pockets, and not Philips' or some other GA chip set vendor's. More important, the computer industry now controls and licenses the D-TV receiver spec, and not the GA/ATSC.

But HD0 is only the tip of this juggernaut iceberg. To accomplish its ambitious goal of sinking the titanic NAB/ATSC ship, Microsoft will do what it has always done so well in the past-- "embrace and extend." The key to Microsoft's awesome marketing power has always been its operating system franchise. And more to the point, the company's total control over its APIs. By defining the APIs and the layer of underlying services, Microsoft has consistently rallied huge numbers of software developers around an ever evolving set of operating system features. All those new applications, now locked into the MS APIs, also serve to lock the user into Microsoft's operating systems. (This after all, was what was behind Sun's Java -- sever the application from the underlying OS/APIs and you also sever Microsoft's almost total desktop monopoly.) No surprise then, that we see Microsoft once again playing to its API hand in order to conquer D-TV.

Microsoft has already set the D-TV stage with some recently announced enhancements to its DirectX family of APIs, now called DirectX 5.0. This new rev of DirectX includes support for Java, streaming video, and object-based multimedia formats. 5.0 further marks the beginning of a profound change in the desktop, as multimedia objects and elements now start to become integral to the look and feel of the user interface. DirectX 5.0 is also Microsoft's first truly unified API, via which CD ROMs, telephony, and web-based publishers can draw on a single source of library functions and services. By doing away with having to decipher what are APIs, which are libraries, and what are controls, 5.0 (hopefully) eliminates the confusion which has always plagued DirectX developers.

In addition, the 5.0 API supports the 12 megabit/sec Universal Serial Bus (USB) and the 400 megabit/sec 1394 serial bus (aka, "FireWire.") Also included is support for Intel's new 57 MMX instructions, the Accelerated Graphics Port, and Microsoft's Talisman multimedia architecture (see the Spring, '97 issue of 21st, "Intel MMX vs. Microsoft Talisman," http://www.vxm.com).

But much more fundamentally, the old way of dealing with multimedia frames as consisting of 3-D and 2-D images, which were then stored and manipulated as a rasterized image and described on a pixel by pixel basis, is being done away with by rev 5.0. Instead, such complex multimedia objects become collections of autonomous objects, stored in their own individual containers, which can then be individually manipulated. Video streams can also be put inside a software object container. All these varied containers can be dynamically marshaled as required, thereby offering nearly limitless multimedia content creation possibilities.

DirectX has also become a deeply layered architecture. At the bottom-most layer is the hardware itself. Next up are the familiar DirectDraw, Direct3D, DirectSound, as well as the newer DirectSound 3D, DirectInput, and DirectXSetup. Microsoft now calls this API set, "DirectX Foundation." For maximum application performance, like for games, developers will want to get closer to the metal and use the Foundation services. The next layer after Foundation holds the DirectX media class of APIs. This encompasses animation, networking and streaming services, Direct3DRM, and Direct Play. Tool vendors would probably use this API level. These two layered APIs, taken as a collective whole, are termed by Microsoft as the DirectX multimedia system services. Drawing upon this API assemblage are NetShow, NetMeeting, VRML, ActiveMovie, Interactive Music, and others. Finally, sitting on top of this system service wedding cake are the components, like ActiveX Controls, and Java applets. Taken altogether, they represent a comprehensive set of APIs and services for multimedia development.

Not surprisingly, Microsoft is squarely aiming its new multimedia and video-capable system, along with the NT platform, at the professional broadcast-video and animation markets. These segments have long been a Macintosh and SGI UNIX stronghold. Microsoft is also reportedly making DirectX deals with professional video development vendors, like Avid. However, it is not content with just owning the professional broadcast market. Microsoft also intends to support the new flavors of DirectX on video arcade games, set-top boxes, and hand-held DVD players.

Thus, for D-TV, Microsoft is already well on its way to establishing its familiar pattern of embrace and extend. First get the producers of video and animation content locked into this rich set of operating system APIs. Next, port these APIs to as many different platforms as possible, including PC TVs. And finally, when the new DirectX multimedia content is broadcast, it will go looking for the appropriate WinOS hooks to execute. But as these API hooks wont be on a GA/ATSC D-TV, nothing happens, Obviously, the trick is to get thousands of new API-bound services and applications flying over the D-TV airwaves. Do that, and GA set customers will eventually get mad as hell looking at a blank TV screen. And if Gates can get enough frustrated consumers on a rampage, then the broadcasters will be forced to abandon the current ATSC standards, and tightly embrace the Microsoft religion. Or so Gates' logic goes. Well, after all, haven't variations on this strategy worked successfully in the past? Then why not for PC/D-TV, as well?


A Brief Digression:

Apart from D-TV, Microsoft has a number of other PC TV irons in the fire, all designed to provide "rich, interactive content." For example, Microsoft's major new OS upgrade, WIN98, ak"Memphis," will support multiple types of data pipes; the idea being that the digital content provider can leverage whatever bandwidth happens to be available to the PC. Via Memphis, the digital content delivery stream can be dynamically upgraded in its overall speed, depending on what you have at the PC end, be it a PC/TV receiver card, a direct satellite broadcast receiver card, or, a broadband modem hooked into a TV cable system. There are several types of Internet + "video" broadcasting data delivery systems that Memphis will be supporting:

1. Data over VBI: This is the Vertical Blanking Interval technology as already used by broadcast television. This will get you up to 115 KB/sec. There is nothing new here, as Intel has been pitching its VBI system for some time now. Essentially, VBI gives you closed captioning or subtitling capabilities. But the VBI system died a thousand deaths several years ago, when it was known as Interactive TV. So, no news here, and really, what's going to make the big difference this time around? Moreover, even at D-TV's high definition 1080 data rates, which use a 19.3 megabit/sec pipe, only a portion of that pipe is being used at any particular time. If there is no action on the screen, e.g., just talking heads, less than half of that bandwidth is being used, leaving almost 10 megabits/sec free for downloading data. Even during intense action scenes, like when things start blowing up, there is still 1 - 3 megabits/sec available for shoveling data downstream. In short, VBI in the era of D-TV is a dead technology.

2. Data over Cable. Although data over TV cable is widely touted as being a 10 to 30 megabit sec/ pipe, don't believe it. This is a party line system. The more users pile on, the slower you go. In a heavily loaded neighborhood, all it will take is one yahoo to start downloading full bandwidth video clips to drag every down to dial up modem speeds. Moreover, there are no standards for these TV cable broadband modems and data delivery systems. Expensive to implement, problematic, and proprietary, but potentially useful in certain environments. A restricted technology, still in search of Daddy Warbucks (like Comcast & Microsoft.).

3. Data over Satellite. DBS holds forth the promise of fairly high downstream data rates, up to 400 Kbs, which service you can buy today in some areas.. Some (Hughes) are claiming data rates as high as 30 megabits/sec, but this remains to be seen. Today, current DSS transmissions for TVs are typically less than 5 megabits/sec. Regardless, of all the systems for inexpensively delivering high downstream bandwidth to the widest number of users, DBS holds the most potential.

But always bear in mind that for the data to be injected into the video stream in any of these delivery systems, the VBI or Internet host must be hardwired to the transmitter or cable end. So, he who owns the head end owns the content delivery service. Know we know why Microsoft just shelled out a cool $1B for a piece of Comcast. As for WebTV, it was a business loser when it was independent, and there is no reason why it wont stay that way after being bought by Microsoft At most, Microsoft picked up a grab bag of interesting video/data technologies, which, ultimately, will find their way into other MS operating systems. And lastly, we have,

4. Video over the Net, the great hope for breaking the stranglehold of the major TV networks. If all the networks' broadcast pipes could be bypassed, and video was pumped right into your PC over the Net, then it would mark a watershed in TV communications. Microsoft has already placed its bets by buying 10% of Progressive Networks, Inc., the makers of RealAudio, and just recently it bought outright Vxtreme, a startup which provides for Net streaming video and audio.

However, the likelihood of the Net becoming a viable competitor to either terrestrial, cable, or direct satellite TV broadcasts is slim to none for the foreseeable future. And the roadblock, is not, as you might suppose, those creaky dial-up modems. Rather, it appears to be in the Net infrastructure itself, and thus, is a much more serious problem.

Even when spinal tapped straight into the Net backbone via a T1 link, it appears the best you will do is little better than cheap modem dial up into your ISP. In the July '97 issue of Boardwatch magazine, they did an exhaustive analysis and packet passing test survey of 29 backbone providers in North America. The providers ranged from tiny SAVVIS to big mother Bell. Working in collaboration with Keystone Systems, a company that offers a service for Web Admins to monitor Net performance, Boardwatch published some astonishing test results.

For each of these 29 backbones, Boardwatch measured how long it would take to download a 50 KB Web page file. And let's underscore this: The T1 enabled testing was done straight off the Net backbone, with no data rate robbing dial up systems involved. All these test measurements were done over a thirty day period, with measurements taken every 15 minutes, 24 hours a day, across 27 cities in North America. And the result? The average time measured to snatch this 50 KB web page off the Net via a 1.544 Mbps connection was 9.928 seconds! So on the whole, the Internet; can do no better than 5 KB/sec!

Now let's stuff Java, ActiveX, 'push" content, streaming multimedia, teleconferencing, and all the rest of the rapidly expanding, quickly fattening Internet menagerie onto this already over-saturated wonder of the millennium, and what do we have? A meltdown of Chernobyl proportions. Worse, the more popular a new backbone becomes, even one using the highest speed ATM switches, the more it slows down, just like that four lane freeway during rush hour. Given that the Boardwatch test results hold up, it's clear that the Net infrastructure is still a long ways off from supporting the MS TV NETwork, or anyone else's.

And even with satellites beaming down Net streaming video, the user still has to communicate back to the Internet, which will still be dial-up in most cases, which puts you right back on the rickety backbone. Moreover, the most downstream bandwidth you can expect from that satellite will probably be about 400 KB/sec to 800 KB/sec, or just a fraction of what D-TV can do. Of course, one could forgo the Net's interactive qualities altogether, just do a straight broadcast feed, and thereby pump up the downstream Net video to DBS rates. But doesn't that defeat the whole notion of "rich, interactive, experiences?"

In sum, all the foregoing brings us right back to why high bandwidth D-TV is so important.

Now we can go back to our main story....


So, the big question is, do E.R., Seinfeld, and the Super Bowl really need all the capabilities of DirectX, and/or Win98, to draw in the TV viewers? Most likely not. If enough viewers feel this way, then the ATSC backers will be safe in their industry positions. However, Microsoft, along with Intel, are also trying to cover that viewer-preference eventuality. Their comeback strategy is to set all the new standards describing how D-TV video, audio, and data signals are to be broadcast (which then will be folded into the WinOS du jour).. In other words, they are going right for the NAB/ATSC jugular. If they manage to cut that industry life-giving vein, then the ATSC backers stand to lose everything.

Some powerful folks in Washington like this new computer-inspired broadcasting plan. Indeed, Hundt sees the computer industry as the TV viewer's savior. This departing FCC chairman has a tough time understanding why all those broadcasters are so hung up on sending just good looking, 1,080 line pictures. With such powerful government backing, is it any wonder that the computer industry is feeling flush?

The computer industry has thus made no bones about its position. If the broadcasters accept all these new computer-driven standards, then many millions of inexpensive PC TVs will already be sitting in people's homes, eagerly awaiting those Microsoft/Intel-specified D-TV signals. Helping to drive Gates' low cost argument is the fact that D-TV will add about $1,000 to $2,000 to the cost of an ATSC/GA set. According to this economic theory, if the broadcasters insist on sticking with those outdated NAB standards, then far fewer of those costly D-TV sets will be out there to get their ATSC signals. So why not go with the PC numbers?

However, the notion of cheap TV PCs quickly and vastly outnumbering horribly expensive GA/ATSC sets is just one more computer PR red herring, for all the following reasons:

First, you can expect to see hordes of inexpensive (probably $500 to $700 initially, then much less than $200 by late 1999) D-TV converters hitting the streets. Initially, most people will probably spring for one of these cheap and easy to install converters, rather than throwing out their existing analog sets. However, these converters will put the wide screen D-TV pictures into "letter box" format with black bars at the screen's top and bottom in order to squeeze the new pictures onto current sets. In addition, these converters will not provide 1080 line, true high definition reception, as your current NTSC TV can only put out a maximum of 525 lines.(and even that line number is doubtful for most consumer sets.) Regardless, these cheap devices may still put a significant dent in that $150 billion set replacement figure so often bandied about.

Second, almost all of the new PC Theaters will come with both the PC and a big screen. E.g., Compaq's recently announced PC Theater system, with its Thomson-RCA 36" TV tube. Marketed only as a combo PC/TV unit, it is expected to retail for about $5,000; not much different from what the GA/ATSC set makers will be getting for many of their high end systems.

Third, being the consumer electronics industry, the prices for GA/ATSC sets will drop like a stone, far more quickly than even a PC Theater. Although a couple of ATSC vendors are making noises about their initial sets costing upwards of $10,000 to $11,000, you can reasonably expect to see $500 to $700 ATSC-certified sets within two to three years of initial D-TV broadcast rollout. (A little discussed factor in all the current D-TV pricing rhetoric is the recent trade treaty between many Asian countries and the U.S. which will significantly knock down trade tariffs. Suddenly, very low cost Asian "Tiger" producers have a shot at the gold D-TV ring; which will be much to the revenue consternation of high cost producers like Sony, Mitsubishi, Thomson, et al.) At this consumer commodity price point, a PC, no matter how stripped down, just can't compete. (For the same reason, you can also expect to hear charges of dumping being levied by the computer makers against the Asian off-shore set makers.)

Fourth, just as you will see analog NTSC to D-TV converters, you can expect a flood of $150 or less PC add-in cards that will support all eighteen of the GA standards. These new cards will turn PC Theaters and Broadcast PCs into ATSC broadcast-loving Trojan Horses. In fact, Intel is already making noises about producing such a add-in card that would receive 1080-I analog signals, and then convert them to a PC's progressive scan format.

Fifth, Microsoft still makes most of its money from selling operating system licenses to OEMs, which typically fetch anywhere from $20 to $70, depending on system type and quantity. But in consumer electronics, component part costs are discussed at the level of nickels and dimes. Even an OS that costs $10 is seen as being way out of line. Thus, a consumer electronics system which uses a Microsoft-supplied operating system could actually carry a significant cost penalty to the purchaser--Not to mention what it will annually cost the consumer for constant software upgrades. If the consumer expects their PC TV to be able to get all those ever evolving DirectX-driven, WinOS-dependent D-TV programs, then they will have no choice but to buy into this upgrade business model. (You only have to look at the Web's hamster-trapped-in-a-treadmill upgrade cycle to see that one coming.)

Sixth, and finally, Intel makes it money by selling CPU chips that cost the PC makers several hundred dollars a piece. Even though the actual CPU chip costs far less than that to actually produce, Intel must somehow recoup its hundreds of millions of dollars in R&D, as well its costs to build those billion-dollar chip fabrication plants. How can Intel expect to keep those necessary high margins if it's forced to compete with the consumer electronics industry, where parts only cost a couple of dollars? For all the same economic reasons, Compaq and other PC makers will also have to charge prices for their PC/D-TV products, even for just the system box, that are several hundred to even thousands of dollars more expensive than those coming from consumer electronics companies. All of their PC-driven business models are way out of whack when compared to those of Sony, RCA, and Mitsubishi. So, like Intel, how do these computer companies expect to profitably stay in the D-TV PC TV manufacturing game? (Compaq may have already seen the budget light, as it now seems to backing away from its firm commitment to support Intel and Microsoft's D-TV blitz.)

Yet despite what may appear to be overwhelming barriers to market success, Microsoft, et al, do have one significant hole card. While the ATSC D-TV spec goes into great detail about how to digitally broadcast video and audio, it says almost nothing about how to actually transmit and receive data. If a broadcaster is looking to simulcast data-based services, or a GA set maker wants to support them, both are pretty much out to sea as to how to do it.

Thus, all those 1,600 stations looking to get into the digital TV services business are ripe customers for Microsoft's NT/DirectX development combo, and their viewers for MS PC TVs. Or at least until the ATSC gets its data/video broadcast standard together, which, realistically, may take some time. But even so, none of the NAB/GA membership have the software or developer clout that Microsoft has. It is thus highly unlikely that the NAB/ATSC/GA bunch can beat Microsoft at its own software game. This enormous data gap in its specification wall makes the ATSC's D-TV fortress quite vulnerable. And it is at this weakened point in the ATSC's defenses where Gates and friends will likely focus their all-out attack.

In sum, is it manifest destiny that Mighty Microsoft will win yet another huge battle for market dominance? Despite the many worthwhile benefits of digitally transmitted services over the D-TV air waves, this new victory is far from assured. It will be very tough to get consumers to buy into the ever system changing PC business model. Consumers are conditioned to buying a big ticket item, like a TV, and then forgetting about it.

But the new D-TV services market, like their Web-based counterparts, will always be a moving, ever evolving target, requiring perennial software fiddling on the part of users. Ask consumers to always be messing with their PC TV, and howls of beer and pretzel protest will likely come thundering off coaches across America. Of course, some may say that automated Net downloads of new software, made possible via push technology, will solve that upgrade problem. However, that presumes a stable Internet/Web ISP hookup. To date, such ISP stability is mostly a pipe dream. E.g., the AOL and MSN outages. That need only happen just once, in the middle of a crucial event, and fickle (and yes, lazy) consumers are probably lost forever to the ATSC/GA camp.

As a corollary to this laid-back consumer axiom, passive content viewing, and not interactive services, is what has historically driven the home entertainment industry. It is completely unrealistic that consumers will change more than fifty years of viewing habits over night, despite the computer industry's most fervent wishes and heavy handed marketing. And in content creation, Hollywood and the TV industry are still the undisputed Lion Kings.

They also have a deep seated paranoia about anything digital, which for them, only means opening the floodgates to wholesale illegal copying. (Look at the tortuous, slow path DVD took in getting to market because of Hollywood's digital suspicions.) The Malibu Mob will not look kindly on PC-based devices, which, thanks to their rich functionality, offer endless possibilities for digital copying mayhem. Moreover, the Internet, to which the TV PC is intimately attached, will become a treasure chest of how-to info and software for cracking most any data encryption scheme that Microsoft or Intel come up with. Lastly, it is no more sensible to say that Microsoft can replace this deeply entrenched industry than to say that it can replace Microsoft. Disney is not about to move to Redmond.

It's also important to note that numerous marketing clinics set up by the TV set makers have shown that viewers are totally blown away when they get to compare a DVD picture -- equivalent to a standard definition DTV broadcast -- with a true HDTV image. Despite what Reed Hundt may be saying these days, the differences in picture quality between standard definition and high definition are enormous. For further proof, just compare a DVD with a laser disc. DVD only improves the TV image by about 60 lines, and yet the differences are easy to see. Now imagine what a 600 line improvement looks like! Even more devastating to Hundt's 525 line holy war, it appears that none of the TV as PC crowd has ever done any consumer response testing to a standard D-TV broadcast. So when Hundt says he sees no difference in picture quality, he's apparently a test market of one.

Moreover, the U.S. Congress has even hauled Hundt's ashes across the coals for his "HDTV Sucks!" mantra. At Senate Commerce Committee hearings in September, in which ABC and Sinclair Broadcasting were explicitly warned about the dire consequences if they did not support HDTV's 1080 line transmission format, Senator Ernest F. Hollings, the South Carolina Democrat, singled out Reed Hundt for some especially scathing criticism. Hollings told Hundt, "If you're trying to change the rules and set a new (HDVT) policy, you'd better get a new Congress." Hollings then added, "This whole thing was on track until you muddied the waters." Is it any wonder that Chairman Hundt is now Ex-Chairman Hundt?

For all these manifold reasons, this fierce battle for the new D-TV market will be won by the traditional TV industry and the CE vendors. And that means true 1080 line HDTV is coming. Which also means Microsoft Intel, Compaq, et al, lose this one. It will always be easier to sell a pretty picture than complex digital services to a naive consumer.

However, also consider this PC TV upside: Much has been written of late about the need for low maintenance network PCs. Moreover, companies and organization moving their business models over onto the Internet realize that high quality networked video, audio, and animation are the keys to next generation systems success. But the current desktop metaphors, even the new Web-based ones, still don't cut it in this new multimedia, Web-based world. And unfortunately, the Java/Netscape/ActiveX wars have spawned not PC simplicity, but seemingly endless complexity of choice.

But lo, what do we have here? A low cost, simple to operate and maintain consumer PC/D-TV expressly designed to process networked multimedia content, and digitally broadcast services? And the DirectX/NT system now provides all the developer hooks to get this new information processing paradigm on the air? Yo, Compaq! How quickly can you deliver fifty million PC TVs as desktop replacements?

Thus, perhaps the ultimate irony: The consumer electronics vs. PC TV battles may have instead produced the much sought after next generation Internet PC.

As the saying goes, wars rarely end the way you think they will.

Copyright 1997, Francis Vale, All Rights Reserved

21st, The VXM Network, http://www.vxm.com

s