To: John Rieman who wrote (25935 ) 12/1/1997 3:34:00 PM From: DiViT Read Replies (2) | Respond to of 50808
"..the most significant thing that emerged from the IBC convention.." MPEG-2 comes of age Paul McGoldrick ÿ 11/30/97 Broadcast Engineering Copyright (c) 1997 Intertec Publishing Corporation. All rights reserved. ÿ To me, the most significant thing that emerged from the IBC convention in Amsterdam was entirely expected, but welcome. C - Cube gave the first public demonstration of one of its DVx chips simultaneously decoding two MPEG-2 streams. This is probably one of the last developments needed for MPEG-2-based non-linear editing systems. The result is likely to be major improvements in picture quality, higher compression ratios and improved end-to-end interoperability, compared to JPEG systems. The flexible DVx architecture is based on a 32-bit RISC processor extended for MPEG's compression-specific instructions. Encoding and decoding can be 4:2:2 or 4:2:0, and an adaptive field/frame algorithm allows for either field- or frame-type data. From a professional/broadcasting viewpoint, a single chip that decodes two streams is a major breakthrough. Although the more important chip in the DVx family may be the one with an encoder and decoder on-board and both are real-time (at least within the limits of acceptability of the phrase!). The choice of Motion-JPEG in current non-linear editing systems is almost entirely driven by the availability of single-chip codecs. Having the same functionality in MPEG-2 is an essential part of a switch in standards. Of course, if there was to be a chip with two streams of encoding and decoding on board What will happen with existing streams using M-JPEG and the DV (DVS) standards of DV consumer (4:1:1, 25Mb/s) and DV prosumer (4:2:2, 50Mb/s)? Clearly, technology decisions and market decisions do not necessarily track well. With the higher compression rates - saving valuable storage space _ and better picture quality, MPEG-2 systems should, if the prices are comparable, be the no-contest victors. In practice, the market forces associated with consumer-type standards are too intensive and too complicated to be called by average humans. It was staggering to read that after the debacles associated with the next generation of DVD , there is now a market alignment (i.e., war) set up for the next generation of audio compact discs. One side is the familiar Philips/Sony party and they are against everyone else. The M-JPEG/DV systems will be usable with the MPEG systems as connection between the existing systems, and a transcoder can easily be accomplished on a Firewire (IEEE 1394) system. Unfortunately, this method would always involve decoding the M-JPEG/DV signals and re-encoding in MPEG. Non-editing uses The importance of these and future products are not limited to broadcast. Doors are opening into other areas. Take the case of recordable DVD . With encoders costing upward of $20,000, it is unlikely that the average household would have a recordable DVD . The average ratio of the number of MPEG-2 decoders to encoders sold is probably on the order of 10,000 to one, and that is before real adoption of the standard in the home. It has to be expected that, with the advent of the codec chip and with the inevitable price reductions that result from volume, the first mass-market implementation will be in high-end PCs. It should take only a couple of years for the usual trickle-down through medium- to low-cost PCs to take place. By the turn of the century, every machine will likely be fitted with an MPEG-2 codec. The multiple-to-one ratio of MPEG-2 decoders to encoders is going to be quickly replaced by a one-to-one ratio. Even more than recordable DVD , and the authoring that goes with it (even I might be tempted), is the fact that we have reached the point that consumer digital VCRs and disc-based camcorders are really rather close. The quality of these devices is going to be limited only by the lens/microphone system and any deliberate degradation that might be introduced, allowing for future "improvements." In some cases, the acquisition of signals is going to be dramatically different, too: Texas Instruments and Motorola have made announcements that turn our current CCD detection systems on their heads. How professional the all-CMOS sensor from TI can be, for example, is not yet known, but the integration possibilities are staggering.