Chapter7 Making Digital Audio a Reality

Making Digital Audio a Reality

In 1971, Heitaro Nakajima resigned from his post as head of NHK's Technical Research Laboratories and joined Sony. Four years earlier at NHK, Nakajima had commenced work on the digitization of sound and within two years had developed the first digital audio tape recorder. He was struck with the idea of digitizing sound when trying to improve the sound quality of FM broadcasts. Nakajima thought that by using digital technology, which had only been used in computers and long-distance telephone transmission, the quality of recorded sound could be improved.
Spurred by the 1964 Tokyo Olympic Games, the Japanese government as well as private industries had aggressively invested in R&D activities related to broadcast technology. This drive induced growth and development of the domestic broadcast industry. For example, satellite broadcasting was developed and color transmission followed closely behind. Until then, however, these advances in video and audio recording were restricted to analog technologies.

Nakajima was one of the first to actually produce digital sound. He achieved this by sampling sound waves at defined intervals. Each sample was then converted into a binary number that could be recorded as a series of pulses on magnetic tape. This was the basic process used to digitize sound. At the time, no one could really see any future in Nakajima's large and expensive digital tape recorder, which reproduced static noise.

When Nakajima joined Sony at the invitation of Shigeo Shima, audio technologies were still primarily analog. In addition, there was considerable negative sentiment within Sony toward digital technology, partly because Sony had decided to withdraw its only digital product, the SOBAX electronic desktop calculator.
Following the 1967 launch of the first SOBAX model, the ICC-500, the development of compact, low-priced electronic calculators progressed rapidly. This was due in large part to the development of such semiconductor-based components as ICs, which made electronic circuitry more compact, and LSI (Large-Scale Integrated) circuits, which were even more compact. Sony and other manufacturers developed LSI circuits based on MOS (Metal-Oxide Semiconductors). However, these proved to be quite expensive for Sony, and the company could not keep pace with the fierce competition. Sony thus decided to halt production of the SOBAX calculator in 1973. Because of this decision, Sony's digital technology, which had grown as a result of the development of the digital calculator, was temporarily stalled.

Despite these setbacks, Nakajima, unable to forget the first time he heard digital sound, did not give up. He earnestly believed that digital technology would be valuable in the future. Ten years after he started his research, he still held the belief that "digital technology would be valuable within ten years." "I'll make it so," he said. Achieving the best sound quality was his lifelong dream.

Saved by Betamax

Sony's first PCM digital audio recorder, the
X-12DTC, used a 56-channel fixed head.
Shown here is the tape transport mechanism
during its development stages.

Two years after joining Sony as chief of the Audio Technology Center, Nakajima together with Jun Takayama and Hiroyuki Suzukawa began work on the development of digital audio. The building in which they worked was some distance from the Sony head office in which Ibuka worked, making it difficult for Ibuka to see what Nakajima and his colleagues were working on. The distance between the two buildings was enough to allow Nakajima to push ahead freely with his endeavors, unhindered by constant inspection. Nakajima forged on, always mindful of the motto, "Continuity is strength."

To code an analog signal or to establish a digital modulation system, Nakajima continued to work with the PCM (Pulse Code Modulation) that he had used during his time at NHK. This system was based on a concept introduced in 1939 and was the same one used for computers, long-distance telephone transmission and broadcasts to Earth from Apollo space-crafts.

The amount of tape space taken up by digital audio recording is about one hundred times that of an analog audio recording or the same amount of space on a tape used by a broadcast-use VTR. Therefore, conventional analog audio devices could not be used because the tape could not record enough. When a mechanism is made from scratch, this machinery becomes unavoidably large.

Nevertheless, Takayama and Suzukawa toiled to build Sony's first PCM digital audio recording machine, the X-12DTC, was announced in 1974. It used 2-inch wide tape and a fixed head with 56 channels. Although it reproduced sound, the X-12DTC recorder was roughly the size of a refrigerator. The transport unit alone weighed approximately 250 kilograms. Although overly bulky, the creation of the first machine marked the beginning of Sony's history in digital sound recording. The machine was transported to and from various venues to make test recordings of orchestral music. The recorder was also exhibited at the 1974 Audio Fair in Japan. Some audio specialists remarked on the clarity of the machine's sound. In the end, however, the recorder was not marketed, even though producing digital sound through a PCM system with fixed heads represented a revolution in recording technology.

Encouraged by Nakajima's enthusiasm, Toshitada Doi and his colleagues joined the development team and began work on constructing a smaller and cheaper recorder. One of the group members, Akira Iga, had an inspiration. The frequency band used to reproduce video images was more than 200 to 300 times the frequency band of analog audio. Theoretically, this would be sufficient to record the large amounts of information necessary for digital audio. Iga wondered what would happen if he tried to make a digital audio recording using the Betamax home-use VCR, launched in 1975. Nakajima had him try immediately.

Nakajima and his team designed a PCM circuit that would perform large-scale signal processing in order to allow Betamax to record and play back digital sound rather than video images. This was called a PCM Processor, and it appeared to hold great promise. A VCR and a PCM processor used in combination constituted a digital audio tape recording system. This system was displayed at the 1976 Audio Fair and attracted considerable interest.

In September 1977, the PCM processor was marketed as the PCM-1. Although it had to be used in combination with a VCR, it was the world's first product to be commercialized that enabled digital recording and playback. The PCM-1, which was generally intended for home-use, was priced at 480,000 yen.

Despite its high price, the PCM-1 was praised for its sound clarity. A first time listener of the PCM-1 likened its clear, digital sound to removing a thick layer of clothing from analog sound. At the same time, however, the PCM-1 drew much criticism.

The Approval of Maestro Herbert von Karajan

The PCM-1, commercialized in September 1977

It is difficult to know exactly how a product will be used until it has been launched. Depending on where and by whom a product is used, the product will either function well or encounter difficulty. Those who design a product have little influence on its success once launched.

For example, there were a number of complaints from people, who used the PCM-1 in combination with various audio units, including FM tuners. They found that it produced an unpleasant electronic humming noise. An increase in dynamic range caused the machine to pick up noise unless the unit was operated in a quiet, soundproof room. This gave rise to complaints of poor sound quality. With any new product, problems that cannot be foreseen in the laboratory are bound to occur, and Nakajima received a continuous stream of complaints.

Eventually, Nakajima and his colleagues moved to stem the tide of criticism, sorting through and analyzing the various grievances. A large number of the problems had occurred because R&D efforts had not been sufficiently thorough because of the haste to create a finished product within one year. Also, the group discovered that the operating instructions supplied with the machines were not sufficient and resulted in misuse.

Thus, Nakajima and his colleagues ascertained the reasons for the grievances. It was a difficult time for the engineers, but the experience of sorting and analyzing claims was a valuable lesson for developing the next digital audio system.

The PCM-1 engineers who had borne the brunt of so much criticism did receive some encouragement. One person who praised the PCM prototype processor as a master recorder for use in studios was the world-renowned Austrian conductor Herbert von Karajan. Maestro Karajan was a longtime friend of Morita. One day in September 1978, when Karajan was visiting the Morita home, Morita asked Nakajima, "Since Mr. Karajan has come here, isn't there something we can play for him?" "Let's play him something on the PCM recorder," replied Nakajima, happily carrying the prototype into the room.

They played a performance of Karajan's that had been secretly recorded during a rehearsal in Salzburg. Rather than scolding them for this unauthorized taping, Karajan was profoundly moved. "This is a new sound," he said. Extremely interested in machines in general and in new recording technology in particular, Karajan had his own recording studio and often edited recordings himself. He was interested not only in the quality of sound, but in future recording systems as well. He claimed to prefer the sound of the PCM system over that of analog recordings, which he was more accustomed to. The engineers were extremely happy and felt much encouragement with the approval of Maestro Karajan.

The Age of the Optical Disc Follows

The launching of an audio disc using the PCM

From the early stages of the PCM-1 development, Sony had been looking for a way to record and play back sound using a digital audio disc. In 1975, Sony put the Betamax VCR on the market. At the same time, Sony had been working to market an optical videodisc (laser disc) which was developed by Philips. Described as "a record that plays a picture," and having a diameter of thirty inches, the same as a conventional LP record, this "large platter" was developed by Senri Miyaoka, the same person who had developed the Trinitron color TV. Deputy President Iwama remarked to Miyaoka, "Sony must have optical technology in the coming age of electronics." Encouraged by Iwama, Miyaoka assembled a team of engineers to develop an optical recording system, selecting members from within the 2nd Development Division.

In spring 1976, Doi and his colleagues delivered a prototype of the PCM-1 to Miyaoka and his team and asked them to make a disc, which could record digital audio. This disc was probably the first digital audio disc prototype in the world.

The type of signals recorded by the Betamax and the video disc were the same, meaning that the PCM-1, which had been developed as an adapter for the Betamax, could also be used as an adapter for the video disc player. At this time, Doi's plan was to connect the videodisc player to the PCM-1, thus creating a digital audio disc. The sound quality of the Betamax was relatively high, so this seemed like a reasonable plan. When Doi and his colleagues tested the disc they had developed; however, they felt as if hit with a sledgehammer. What they heard was the complete opposite of their expectations. Far from producing a clear sound, the disc produced a poor disconnected sound against a background of static and could barely be heard.

After looking into the causes, Doi made three decisions. The first was to use the PCM-1, which they had been preparing to launch in the autumn of that year, as an adapter exclusively for Betamax. This was a simple and logical decision. The second decision was to approach the audio disc and the videodisc as unrelated items and give priority to developing audio technology. In other words, to develop a brand new way to directly record the digital audio signal onto an optical disc, instead of using a video signal from a video recording format. At the time, the videodisc was a star product, which everyone believed would prove popular throughout the world. The decision to develop a digital audio disc that used a totally different format was extremely bold. Doi's third decision was to actively develop signal error correction technology, something for which no one had any experience to draw upon. Considering this, the third decision was quite exceptional.

Audio data is converted to binary digital signals, and recorded on the surface of a digital audio disc as an array of ones and zeros. When played back, digital signals are converted to electronic signals as a light beam reads them. Signal error correction is a function whereby the machine rectifies a situation in which a bit has been incorrectly read. Signal error correction is especially important in optical discs because, compared to tape a greater number of erroneous readings are made during playback.

Under Doi's leadership, the team began to study signal error correction. At the same time, Doi ordered computer specialists to create a computer simulation system for studying signal error correction. Thanks to the efforts of Doi and his team, Sony was able to establish a firm position for itself in the field of digital audio.

At the 1977 Audio Fair, Sony and two other companies exhibited their digital audio discs and players. However, these two companies used a video signal on a videodisc, and they had a rather simple signal error correction feature. Sony's machine was an unmodified video disc player that recorded digital audio signals, not video signals, directly on an optical disc that employed signal error correction.

Sure enough, people criticized Sony saying, "Even though it's possible to meet industry standards by using a video disc with a video signal, Sony had to be different and use a different format." Responding to this criticism, Doi gave a speech during Audio Fair in which he said, "Using video signals, you can record a performance that's thirty minutes in length, but by recording audio directly you can record a performance thirteen hours and twenty minutes long." In other words, the efficiency of direct audio recording is twenty-seven times greater than that achieved using video signals. He tried to convince listeners of the technological superiority of direct recording. Ohga, who was later informed of this speech, angrily commented, "Thirteen hours and twenty minutes is an absurdly long time. Hardware works only when it's fed software. Although Sony has its own software company, CBS Sony isn't going to make any money by marketing a disc with hours and hours of music recorded on it."

Thirteen hours and twenty minutes was actually the theoretical limit of direct recording. In 1977, the playing time of a demonstration disc was an hour, and at the 1978 Audio Fair, the time was two hours and thirty minutes. Due to the structural limitations of the video disc player, the actual recording density achieved was lower than the theoretical level.