Bob Perry’s 1394 vs. DVI Article
Dear Gary: It’s disappointing that Bob Perry’s article promoting his company’s decision to embrace IEEE 1394 is so shortsighted. First, he tries to explain how IEEE 1394 will cost less. He grossly exaggerates the cost of HD MPEG-2 decoding by stating today’s end-consumer cost is at about $1,000. That’s true but isn’t relevant. Instead, the incremental cost to the consumer electronics manufacturer is a reasonable cost to cite. The bill of materials, dominated by a HD MPEG-2 ASIC and a RAM chip, is probably closer to $20, and will drop over time. He goes on to explain how only one HD MPEG-2 decoder would be needed in an IEEE 1394 system. That’s simply not true—can you imagine buying a DVD player with IEEE 1394 as your only video output option? Each video source will have an MPEG-2 decoder so that it can drive at least one other video output. IEEE 1394 may be even more expensive than competing solutions because it will require an additional HD MPEG-2 decoder in the display! Secondly, and far more importantly, IEEE 1394 will limit consumer video quality to today’s 720p or 1080i resolutions and compression method for the foreseeable future. The display manufacturers will build their IEEE 1394-connect, HD MPEG-2 decoders to handle 720p and 1080i, nothing more. And consumer video sources wouldn’t try to break past that quality barrier, because no current display would be able to play it. What happens if an MPEG-4 based DVD becomes standard? Or a higher bit rate MPEG-2? What if another new compression standard is made available, with clearly higher quality? The IEEE 1394 displays we buy today won’t be able to handle it. The video sources will have to “dumb down” the video for playback on IEEE 1394 displays. Please don’t lock us in to this dead-end standard. Keep the displays agnostic from the video sources. P.S. Thanks for such an in-depth magazine. Nothing else comes close for exposing the technical issues related to making home theatre its best! I just subscribed for two years; looking forward to more great stuff.
Jeff Irwin
mailto:jjirwin@rescomp.stanford.edu
Mitsubishi’s Bob Perry Comments
First and foremost, thanks for reading the article and taking time to comment. Obviously, I disagree with your opening assessment, however please allow me to provide insight by answering your points below. With regard to pricing and cost, it’s important to remember that raw bill-of-materials sometimes is almost unrelated to final end-consumer cost. A chicken sandwich may have a bill-of-materials of 20 cents, but it’s not reasonable to expect to buy it for 20 cents. As for the HD MPEG-2 Decoder ASIC and RAM chip costing $20.00—not on this planet. But you are welcome to cite a source for a High-Level, Main Profile MPEG-2 and 128 MB of RAM for “closer to twenty dollars,” and I will call them. It is correct to assume that all electronics decline in costs over time for the same capability, but consumer’s expectations and demands increase, so over time capability and sophistication increase, quite often leading to moderate or no final cost declines in final products. I disagree with your assumption and analysis. A DVD player—connected to the IEEE 1394 network—can drive every display device simultaneously that is connected to the network. You only need decoding at the display, not at the source. If you consider today’s DVD players they do not drive two displays, just one, and they have the SD grade MPEG-2 decoder in them— wasted cost in a IEEE 1394 network. Once again, I disagree. The IEEE 1394 can handle resolution levels above human comprehension. For example, a standard IEEE 1394 connection operates at 400 Mega-bits-per-second (Mbps). A 1080i MPEG-2 stream has a data rate of 19.3 Mbps. A full 1080p MPEG-2 bitstream is less than double (because of compression efficiencies) at 28 Mbps. A 2160p MPEG-2 bitstream (were such a format to exist) would be below 40 Mbps. A standard IEEE 1394 connection could carry 10 such theoretical streams simultaneously without degradation to all the devices on the network. Without question, the limiting factor is NOT the IEEE 1394 connection. The real limiting factors in the point about video quality raised by the the reader are multiple: 1. How much bandwidth is allocated by the program supplier for transmission to the home? 2. What is consumer desire and marketability of higher resolution levels? 3. Can the content industry actually record and store much higher resolution levels? The highest bandwidth connection for content to the home today is the 6 Mhz broadcast channel, with a maximum data rate of 19.3 Mbps. Until the American public is ready for another change in this system, or the obsolescence of all digital receivers at that time by changing the system or compression method, this will not change. With regard to consumer desires, when a large 60-inch diagonal display is viewed from 12 to 15 feet away, what is the capability of the human eye? Many experts believe that 1080p is probably the highest resolution level appropriate for these screen sizes and distances. MPEG-4 is an interesting digital compression system because the level of compression can be adjusted to meet bandwidth requirements—i.e., DVD level video can be recompressed to lower resolution to fit on a CD-ROM. But the infrastructure of a broadcasting system—over-the-air, satellite, and cable—tends to be fixed, because a change requires a huge changeout of the installed base of receivers (including satellite receivers and cable boxes). So changes take years and years. Higher bit rate MPEG-2 can also occur but will probably be limited to a HD Playback Only DVD format that is expected in a couple of years. Once again, the IEEE 1394 will not be a limiting factor. And today’s integrated digital televisions generally can decode up to 1080p for display. Without question, displays that do not include some form of digital interface, or the ability to upgrade to one will be dead-ends in the years to come as digital copy protected connections are demanded by content providers and delivery systems. Thanks again for the questions, I enjoy the debate!
You can E-mail Widescreen Review @ mailto:editorgary@widescreenreview.com