back to home pageby Erica Schulman
originally published in Piano and Keyboard,
Issue 188, September / October 1997
Trying to hear a good concert can be a frustrating experience. On any given day there may be several recitals that you would like to attend, but they are too far away. Your local radio station is not covering them. What can be done? If only there were a way to hear music from any location. While travel remains the only solution for most cases, this year, people who wanted to hear a recital by pianist Earl Wild or performances at the Van Cliburn competition were in luck — the concerts could be heard anywhere in the world. Using computers and the international network of the World Wide Web, audiences could tune in from any place with an Internet connection. The web originally became famous for sending text and graphics; now, at last, it has developed an efficient means of transmitting sound.
Like text and graphics, sound is a type of information. It may be represented as amplitude versus time. Our ears translate fluctuations in amplitude into pitch; the higher the frequency of the fluctuation, the higher the pitch. When sound is recorded using analog methods, arbitrary amplitudes or frequencies can be stored. For example, sound waves will cause a diamond needle traveling on a blank vinyl record to vibrate and cut a continuous groove. By contrast, a digital recording takes a series of rapid snapshots. The result is discrete in both amplitude and time, causing loss of data. In theory, analog should allow for the most realistic capture of sound.
| Yet the digital format has become the standard for music, packaged as compact discs or DAT tapes. Compared to vinyl records, CDs are smaller and more durable. Their lack of surface noise allows for a greater dynamic range in recording. Digital is the language of computers, and it was inevitable that the growth of digital recording would allow music to appear on the desktop. A variety of computer formats became available, including MIDI, AU, and WAV, each requiring different software. The ability of the World Wide Web to share files with multiple users allowed audio to be transmitted easily, and web authors were soon adding audio to their pages. But no matter what format was used, the fundamental limit remained: the fewer the bits, the poorer the sound quality. In order to prevent the discrete nature of the digitized sound from becoming apparent, very high sampling rates and precise measurements of amplitude are used. A typical classical CD may feature a 44.1 kHz sampling rate, 16-bit sound, and stereo. One minute of music requires 44,100 x 60 x 16 x 2 = 84,672,000 bits, or over 10 Megabytes of storage. 10 Megabytes is fairly large. In order to listen to standard audio files, it is first necessary to download them. A user with a 28.8 Kbps modem will take over 49 minutes to get that one minute of music. CD quality sound becomes impossible. But even with a 22 kHz sampling rate, 8-bit sound, and mono, one minute of music equals 10,584,000 bits or 6 minutes of download time. While short, low-fidelity clips are certainly feasible, longer works are not practical. | Earl Wild on the NetOne of the first musicians to take advantage of Internet audio was pianist Earl Wild. On March 7 he broadcast a performance that was billed as the first solo piano recital over the web. Wild, a concert veteran, also holds the distinction of being the first live pianist on television in 1939. The recital was given in Carnegie Music Hall, Pittsburgh, Pennsylvania, and was broadcast live on WQED-FM in the local area. The program included works by Beethoven, Brahms, Chopin, Liszt, and Medtner. WQED broadcasts over the Internet using RealAudio, so any user who downloaded the free player could listen in. The setup in the concert hall appeared perfectly standard for radio broadcasts, with the usual microphones, but a live feed was also sent to the computer server. The concert was made possible by a collaboration between the Carnegie Mellon Department of Music, WQED-FM, the Carnegie Mellon Network Development, and Progressive Networks (creator of RealAudio). |
Then again, why should one have to wait to download an entire sound file in order to hear any of it? It is only necessary to know what the sound should be at any given instant in order to play music. Several companies have developed tools to take advantage of this. The music can be played while the file is downloading, a concept known as streaming. This eliminates tedious waits and allows works of arbitrary length. However, it limits the playback rate to the speed of the modem. The player requires a certain amount of time to process the information through the computer, slowing things down even further. In order to sustain a steady flow the music must be highly compressed. Much of the development effort has focused on improving these compression algorithms and the latest versions are fairly good.
Like many businesses that function over the Internet, audio streaming companies offer free products to individual users in the hopes of establishing themselves as the standard. If you have a web browser and a sound card, you can download them and start listening (see box).
One branch of the music industry has long been involved in freely sending music out to listeners, namely radio stations. It is not surprising that they were among the first to embrace the new technology. One of the earliest to begin broadcasting over the Internet was classical station KING-FM in Seattle, located in the same city as the company that produces RealAudio. KING and RealAudio began collaborating in the fall of ’95 to determine the optimal equalization for signals sent through the Web; frequencies at high or low extremes will cause the encoder to slow down. The resulting signal is about 2/3 the size of the raw feed, and runs more smoothly than most Internet broadcasts. An estimated 2,500 users per day tune in via the Internet, some from outside the United States. Although the signal is commercial-free and therefore does not produce any direct income, it provides a useful promotional vehicle for the station. Given the ease with which statistics can be collected on number and location of listeners, advertising rates on an Internet link could be much more precise than those sent out to an estimated FM audience.
So far radio Internet broadcasts have been performed by existing stations. But the ability to broadcast worldwide without purchasing an expensive FCC license or transmission tower is likely to lead to a proliferation of new web stations. The extended reach of these stations means that almost any format can be profitable, even if it only appeals to a small percentage of the potential audience. When cable expanded the number of television channels, specialized ones dealing only with nature, history, cartoons, science-fiction, or sports were created. We can expect the same thing to happen with radio — and that is good news for classical music. Can an all-piano station be far behind?
Individuals can also use the new technology to promote their own work. Record labels no longer hold a monopoly on distribution. Sammy Hagar, former member of Van Halen, recently released the first single from his new album over the Internet. Some less famous but ambitious composers have their own works available on their home pages. More unusual users include a Christian tutorial service, a recital series from the University of Miami, and an early instrument ensemble.
Someday FM radios and record labels may be things of the past. For now, audiences can begin to listen to more of their favorite music — or record some for the rest of us to hear.
Erica Schulman is a physicist and web author who specializes in sites for classical artists and recording companies. Her most recent commission was a site for pianist Earl Wild. Her web pages may be found at http://www.netcom.com/~ens .