If you follow the common wisdom on this issue, you might assume that replication companies worldwide should start mothballing their stamping lines and be looking for opportunities in the broadband arena. After reading recent press releases, reviewing the specifications, and reading about the capabilities of the new Windows Media development environment, you could logically draw the same conclusion. But it seems to me that current bandwidth demands—coupled with the coming age of high-resolution audio and video—are going to require more of the system than Sprint and WorldCom can hope to cover in the next 3-5 years.
It's no secret that the recording industry has had a particularly bad couple of years. The industry says sales are off almost 25% from the previous year. Record companies and their trade association, the RIAA, are scrambling to figure out how to deal with the dramatic decrease in revenues. One reason cited for lackluster sales is the proliferation of "music-swapping" through the Internet. The movement of copyrighted tunes between music fans, especially young people, has transformed how music is delivered, and has forever changed the paradigm of how music is valued. The bandwidth for delivering music—via downloading or streaming—is already with us. But is the quality sufficient? As the "ripping" generation migrates from the dorm room to apartments and, eventually, their own homes, will "CD-quality" MP3 files provide enough fidelity or channels? I don't think so.
It's true I'm an audiophile. I enjoy the sound of a well-recorded stereo CD, but I'm captivated by the immersive experience of a 5.1-channel surround mix delivered on a DVD-Audio disc. Music has taken a quantum leap in quality in this new format, and as a result, there is a tremendous leap in the storage and bandwidth required to handle it. Even taking into account the amazing capabilities of state-of-the-art "lossless" compression algorithms like Meridian Lossless Packing, the increase in bandwidth is about seven times over that needed for a compact disc.
This calculation was done assuming 96kHz/24-bit audio in 5.1 channels. What happens when the sampling rate is increased to 192kHz or even 384kHz? We're already on the way. The first computer sound cards were capable of 22.05kHz/8-bit sound. My studio just completed work on a bundled disc for the latest Creative Labs Audigy 2 Sound Blaster. It handles all of the DVD-Audio sampling rates and word sizes (up 192kHz/24-bit) and includes the MLP algorithm. Does the world of computer gaming really need that level of fidelity? Maybe not—but engineers will always take advantage of the additional horsepower offered by improvements in processors and computer designs and use it to distinguish their products from the rest of the market.
Accompanying the increased resolution of each individual channel of sound will be an increase in the number of those channels. First there was monophonic, then stereo, and currently 25 million households listen to their favorite DVDs in 5.1-channel surround. However, some proponents are already moving that number higher. Tomlinson Holman has demonstrated a 10.2 system. I've heard it and can attest to the effectiveness of the added "height" speakers. If the goal is to surround a listener with the sound of a live performance space, then why not increase the number of speakers beyond that number? Maybe a new generation of music consumers will be enjoying a 20.4 system in their new SUVs. I recently visited the Los Angeles R&D facilities of Alpine, one of the world's foremost automotive sound companies. They are equipping automobiles with lots of speakers, as many as 16-20 individual drivers. Each speaker isn't receiving it's own discrete signal—but it's just a matter of time.
On the other side of the A/V equation, what about the bandwidth requirements for video? Will the demands of larger aspect ratio screens and greater resolution impact outstrip the bandwidth of the fiber-optic cable currently carrying your favorite sitcoms and news programs? The emerging HDTV market moves the number of vertical lines to 1080 (or 720 depending on which station you're watching) and the number of pixels increases because of the migration to widescreen formats. This is what the next generation of viewers is going to demand from their entertainment experiences. It's not until we finally are able to achieve the interactivity and resolution of the "holodeck" (the completely real "virtual" experience popularized by the Star Trek series) that designers and engineers will cease their quest for more bandwidth, and we're nowhere near there now.
The counter argument claims that better compression algorithms and advanced schemes of psycho-acoustic modeling will keep the demands for bandwidth within reasonable limits. Certainly, Microsoft has some new tricks up their sleeves with the release of the new Windows Media tools. Some of the quality claims are astounding and undoubtedly, its compression efficiency will delay the need for larger delivery pipes. However, there are a lot of smart people designing higher-capacity optical discs (Blu-Ray and Advanced Optical Discs) for use when high-resolution audio and video are the norm in home theaters and automobiles. From our perspective here at the opening of the 21st century, it's hard to imagine why we might need a terabyte hard drive and 10Gbps transfers. But think back to the days when a 20MB hard drive seemed huge, and try again.