Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : C-Cube -- Ignore unavailable to you. Want to Upgrade?

To: BillyG who wrote (27938)	1/10/1998 12:51:00 PM
From: DiViT	Respond to of 50808

At CES : Satellites to carry HDTV signals -- Designs diverge at dawn of digital TV Junko Yoshida 01/12/98 Electronic Engineering Times Page 01 Copyright 1998 CMP Publications Inc. Las Vegas - The first-generation digital-TV prototypes on display here last week at the Consumer Electronics Show ( CES ) revealed diverging design strategies that could lead to a diverse but potentially fragmented market. Major consumer-electronics companies, looking to product differentiation to establish early market leads, showed divergent architectures based on markedly different chip sets, many secured from in-house IC divisions. Philips Electronics showed a 64-inch, wide-screen (16:9 aspect ratio), rear-projection, digital high-definition TV set with 1,080 interlaced video. Sony Corp. demonstrated a flat-display Trinitron TV that can drive VGA, standard-definition TV and HDTV images. Matsushita Electric Industrial Co. brought a standalone DTV decoder set-top that allows various CRT monitors and projection TVs to reproduce digital signals in the format native to each display. And Thomson Consumer Electronics showed a 61-inch rear-projection digital HDTV, jointly developed with Hitachi, that incorporates a digital satellite system (DSS) decoder. Thomson and Hitachi were not the only ones looking skyward. Indeed, for many in attendance, the stunner of the show was DirecTV's announcement that it intends to dedicate two digital satellite channels for 1,080i digital HDTV broadcast before the year is out. "While a lot of terrestrial TV broadcasters are still unsure about which video-transmission formats they may use in their DTV broadcasting, we are making a statement that HDTV digital broadcasts are worthy of our dedicating extra bandwidth," said Eddy Hartenstein, president of DirecTV (El Segundo, Calif.). DirecTV's HD-DSS programming will primarily comprise "pay-per-view types of services," Hartenstein said, though he added that programming deals are "still in the making." Existing DSS channels won't be sacrificed to accommodate HD-DSS. Thomson last year became the first consumer-electronics manufacturer to commit to incorporating an HD-DSS decoder in every HDTV set it makes. But Hartenstein said the Thomson HD-DSS deal with DirecTV is not exclusive and that his company expects all first-generation digital HDTV receivers to incorporate a DSS decoder, especially in light of DirecTV's aggressive commitment to supporting digital-HDTV broadcasts. Since DirecTV's announcement caught most CES showgoers off guard, however, it was unclear last week how HD-DSS might affect current road maps for DTV systems and silicon. Ease of integration David Badger, manager of digital product design at Thomson, downplayed the design issues of HD-DSS integration. System vendors, he said, will need "conventional DSS-receiver and demux chips, which should be added to a complete HDTV chip set, composed of such key components as a VSB [vestigial sideband] demodulator and an MPEG-2 Main Profile @ High Level decoder." Such an approach, however, may prove costly, since it would essentially mandate that every HDTV incorporate a quad phase-shift keying (QPSK) demodulator and non-MPEG-compliant DSS demux, on top of an already costly HDTV chip set. Indeed, one of the few elements common to the first-generation DTV prototypes is high cost. Philips and Thomson both estimated that rear-projection HDTV systems larger than 60 inches will be priced between $6,000 and $8,000 when they hit the market in the fall. Another common element is the emphasis on 1,080i decode and display. The systems shown by Philips, Matsushita, Thomson and Sony all handle that high-definition format. And on the show floor, many system vendors used special over-the-air feeds, transmitted from Las Vegas-based network affiliates, of 1,080i-based programming. "We firmly believe that HDTV sets should produce some 1 million pixels for true HDTV performance; 480p [progressive scan] is not HDTV," said James Meyer, executive vice president and chief operating officer of Thomson (Indianapolis). Beyond that, however, there was little commonality among the systems shown. There had been pre-show predictions that straightforward projection TVs would be the most common first-generation DTV platforms. But the systems shown were of a decisively proprietary bent. Matsushita leveraged its own single-chip DTV decoder/display processor IC (see Dec. 22/29, page 4) in a demonstration set-top, slightly smaller than a DSS-decoder box, that can decode all 18 video formats specified by the Advanced Television Systems Committee. It can also flexibly display images in a variety of raster formats, including 1,080i, 720p, 480i and 480p. Indeed, in its demonstration, Matsushita used the same DTV set-top to drive three systems: a digital-ready SVGA TV/monitor; a prototype 36-inch wide-screen direct-view DTV; and a 56-inch wide-screen projection TV. "Software smarts reside in the decoder/video-processor chip inside the set-top, allowing different-resolution monitors to display signals [in the format] most friendly to each display," said Sai Naimpally, vice president of Matsushita's Panasonic AVC American Laboratories (Burlington, N.J.). The set-top is the first consumer video product to feature so-called Y Pb Pr output, the component video standard commonly used in today's professional video equipment. The standard can accommodate multiple line resolutions and sync signals for output from a set-top to a DTV display, said Naimpally. The set-top also accommodates NTSC output so that it can be linked to a conventional TV. Further, the box incorporates a 1394 interface, "for connecting to D-VHS or other storage media in the future," Naimpally said. But Norio Kawaguchi, general manager in charge of engineering at Matsushita's television division headquarters, said that accommodation of 1394 output in the final product "is tentative" and is dependent on the achievement of industry consensus toward a solution for digital-copyright protection. Sony, meanwhile, made it clear last week its first DTV model will be based on a flat Trinitron display with a superfine aperture pitch. The company will incorporate two additional proprietary technologies-Multi-Image Driver and Digital Reality Creation (DRC)-in a bid to maximize the upconverted picture quality of its first-generation DTV, said John Briesch, president of Sony Electronics' Consumer Audio/Video Products Group. Multi-Image Driver can upconvert NTSC to the full VGA format and can simultaneously display any analog or digital TV signal via a picture-in-picture func- tion, ensuring what Sony says is the the highest image quality possible from both sources. It's designed to handle both entertainment programs and such data applications as Web sites. High-res mapping Digital Reality Creation (DRC), developed by researchers at Sony's Algorithm Research Center in Tokyo, is a coding method that maps out a high-resolution signal directly from a standard television signal. By using a new algorithm that's entirely different from the motion adaptive linear interpolation method, DRC codes the essence of the signal patterns-including horizontal, vertical and time elements-of the original image and directly maps the code into a new picture. The goal is to enhance the image quality of digital broadcasts as well as programming from such standard video sources as cable boxes, VCRs and DVD. Philips' projection TV, meanwhile, is driven by Philips Semiconductor's TriMedia 1 chip and a newly developed ASIC. Thomson has similarly turned to an in-house chip source, using an MPEG-2 Main Profile@High Level decoder chip designed by SGS-Thomson in its first generation HDTV projection TV. And both Sony and Matsushita are employing DTV chip sets developed by their in-house semiconductor divisions. The 61-inch wide-screen projection HDTV shown by Hitachi and Thomson is based on Hitachi's rear-projection TV technology. Both companies plan to market the set starting in the fall. The implementations at CES may have differed, with each company turning inward for design, but all are banking on DTV to boost consumer electronics in general. Doug Dunn, chairman and chief executive of Philips Consumer Electronics , summed up the sentiment by saying DTV is "what's needed to revitalize this industry."-Additional reporting by Yoshiko Hara.

To: BillyG who wrote (27938)	1/10/1998 1:19:00 PM
From: DiViT	Respond to of 50808

Crosstalk DVD-Audio specsmanship is just so much noise 01/12/98 Electronic Engineering Times Page 50 Copyright 1998 CMP Publications Inc. The proposed specifications for DVD-Audio reported in Yoshiko Hara's story on Japan's Audio Fair '97 are very impressive, to say the least (see Dec. 1, page 1). JVC says they'll be delivering a 96-kHz frequency response, but I'll bet Sony's system will sound better, since it has a 100-kHz frequency response (cough). I think that JVC, Sony, Philips, Sharp, and Matsushita, all deserve a large collective "thank you" for making all those critical transients above 22 kHz available to all of us (cough, cough). I suppose it will be like lifting a wet blanket from the front of my speakers, though I guess I will need to buy those brand new ones that Sharp talks about (flat to 100 kHz) to get the full effect. I guess I'd be ahead of the game as a musician if I'd get good at some new instrument that operates in this ultra-soprano register. Here's one (from www.acmewhistles.co.uk/dogs.html): "The Acme Silent Dog Whistle produces ultra-high frequencies in the range of 5,800 Hz to 12,400 Hz. Used for training dolphins and killer whales at Sea World it is also famous for its effects on cats, horses, mice and even bats!" The point I've been skirting around is that the most common figure for the upper frequency limit of human hearing is 20 kHz. I have not read of a single scientific study supporting the perceptibility of sounds above 25 kHz. The sampling rate of 44.1 kHz was chosen for the compact disk because that was sufficient to encode the highest perceivable frequencies. Just because VHS is in dire need of replacement does not mean that CDs are in similar need of an upgrade. Do the Japanese know something about human physiology that the rest of the world doesn't know, or is the marketing instinct to push numbers just too irresistible? I'd like to see a set of double-blind listening tests upon a large population of humans with a wide range of listening habits in a wide range of common listening environments. I'd like to suggest this as a research topic for someone (maybe someone associated with the Acoustical Society of America?). I have two real concerns. The first is that our tolerance of distortion, i.e., our ability to extract meaning from a noisy signal, is a tribute to the intelligence of human beings. As a species, is it wise to be training ourselves to be intolerant of noise? To illustrate this concept, I like to tell the contrasting stories of two of my good friends. In high school my first friend bought a CD player as soon as they were available. In his car he still had a cassette player where he listened to cassettes made from CDs. While driving with him one day, he popped in an older tape that he had made from a vinyl LP. Upon hearing the first subtle surface noise before the first song had even begun, he angrily ejected the tape and threw it against the back window. My second friend had a cassette (I think it was an old Stones album) that he had copied from another cassette in the following manner: Using a mono cassette player/recorder with the external microphone held against the speaker of a similar unit, he pressed Record on his machine and Play on his friend's machine. To make matters worse, he was in the living room with an inexpensive stereo cassette player on weak batteries sitting on the coffee table. This player was of such poor quality that even one of my first friend's tapes would have so much wow and flutter that even I would not be able to enjoy it. But here was my friend, grooving, having no problem with the quality of the signal, singing along even. While each individual is certainly entitled to his own standard of acceptability, I cannot help wondering what my first friend's superior powers of noise-detection have gained him. I fear that we are training ourselves to demand higher signal-to-noise ratios; in effect, we are getting better at hearing noise that used to be imperceptible. And to what benefit? My second concern is that should these devices actually come to market and should people actually purchase speakers that can accurately generate a 100-kHz signal with a dynamic range of 120 dB, the sonic landscape would become very cluttered with noise that is imperceptible to humans but that might interfere greatly with the communication of birds or other creatures. We behave as if we're the only living things on the planet, playing with numbers so that affluent couch potatoes feel compelled to upgrade their systems; meanwhile, animal habitats get polluted with sonic radiation. The truth is that no one knows the side effects these ultrasound frequencies will have on animals. Incidentally, there was no talk whatsoever in the article about dynamic range. I think 24 bits makes sense for its divisibility by eight, but the players had all better come equipped with some easy-to-use compression, be it of the super bit-mapping variety or whatever. (How about some simple, free algorithm? The DSP involved is not that tough.) People will wise up when they find that the bottom 30 or 40 dB is always getting swallowed by the noise floor of the typical home. Either that, or speakers will be getting blown a lot more than people will like. My recommendation, for what it's worth, is that the bit depth be selectable from 16, 18, 20, 22 and 24 bits, and that linear PCM encoding be used. Sample rates of 44.1 kHz and 48 kHz (and I suppose double this, since there seems to be so much momentum behind it, though I think it's stupid) should be supported, as should all of the DVD Video formats ( MPEG , Dolby Digital). Allowing the content developer maximum flexibility in the trade-off between playback time, dynamic range and frequency response sounds like a noble primary goal, more so than trying to push a proprietary "super" file format with dubious technical merits. The authoring systems as well as the playback systems for the super audio formats will be so expensive that only the wealthy will be able to take full advantage of the formats. For the rest of the world, CD-quality is quite good enough. The DVD Forum does not appear to be interested in making DVD authoring convenient to the masses. Most people who "record" things end up taking some copywritten work and transferring it from one place to another, be it music or video or a computer program. Unfortunately, too often this involves unfair use of intellectual property-a fact which the record and film industries (including Sony) are well aware of. As a content developer, I am sensitive to the issue of piracy, but I also know that the tighter the protection is, the less convenient the medium is. I like the shareware model myself. Registered versions, free updates, manuals, access to an Internet site; these are all positive motivations for paying. Plus the advertising is essentially free, word-of-mouth (or search engine). This gives consumers the greatest access to quality programming in the most competitive environment, though it is an environment that is reasonably easy for a small producer to participate in. My prediction is that the most cost-effect audio scenarios for the consumer will involve recording MPEG audio in the DVD Video format onto a DVD-RW installed in a PC. There will be DVD players that will be able to play these DVD-RWs, though consumers will have to be smart enough to check which incarnation of DVD-RW they're dealing with. One more point: The general manager of Sony's Audio Laboratory states that the company's DSD technology, which features 1-bit quantization at 2.8224 MHz, has a theoretical frequency response of 1.4 MHz "but for marketing products we limit it to 100 kHz." First of all, at 1-bit quantization, what kind of dynamic range would one expect at a bandwidth of 1.4 MHz? The way 1-bit converters work is significantly different than the way an ordinary 16-, 18-, 20-, 22-, or 24-bit A/D works. Instead of measuring each time interval and assigning as accurate a number as possible to represent the voltage at that instant, these 1-bit converters simply return a Boolean value that effectively states that the voltage is higher than it was for the last sample, or lower than it was for the last sample. As one might imagine, this does not result in a very accurate waveform. The reason it works is that the resulting waveform is band limited to a range that eliminates the distortion. It's a DSP trick that allows an inexpensive RF component to achieve performance that otherwise would require a high-order anti-aliasing filter, which also introduces phase distortion. It should be noted that the relationship between the D/A conversion technique and the encoding scheme certainly does not forbid the use of 1-bit converters for playback of 24-bit or 16-bit linear PCM, so the argument for the adoption of their proprietary encoding format on this basis is weak. The cost would far outweigh the benefits. Recording studios and mastering houses would be required to purchase devices that license Sony's technology. Computer-based editing equipment likes linear PCM formats, and the algorithms involved for performing DSP operations on a bitstream are simplest for these linear formats. DVD is an important technology for the future of the human race, and it should always be possible to author and playback the simplest, most intuitive data formats. Nothing beats PCM for this. Why are these companies saying nothing about the potential for enormous play times (like 60 hours of near-CD-quality audio on a 4.7-Gbyte DVD using MPEG or Dolby Digital)? I believe this benefit is more easily grasped by the average consumer, and subsequently I think it would sell more units than the DVD (ridiculously) super audio proposals would. Sam Frantz Tangram Enterprise Solutions Inc. Malvern, Pa.