All About Dolby
June 1971 Radio-Electronics

June 1971 Radio-Electronics [Table of Contents] Wax nostalgic about and learn from the history of early electronics. See articles from Radio-Electronics, published 1930-1988. All copyrights hereby acknowledged.

This photo of Ray Dolby holding one of his prototype noise reduction circuits is probably the most widely published of him and therefore the most iconic of the Dolby noise reduction system. Audiophiles of the era (and today for that matter) immediately recognize the man who took the hisses and pops out of their beloved music. I always like to keep in mind when reading articles like this one in a 1971 issue of Radio-Electronics magazine is that when it was originally published, Dolby had not yet become a household word and news of his accomplishment was just getting out. Many articles, books, and research papers have been written on how the Dolby system works. At least five of them from the groundbreaking era have been posted here on RF Cafe, so you can get some insight into the excitement. The technical term "companding" (compressing and expanding) was being seen in print for the first time. Unlike today when such signal processing is handled in software or in an ASIC, Mr. Dolby had to design and build his conception from discrete components (see the Bob Pease-style breadboard in this photo).

Other Dolby articles from vintage electronics magazines include: "The Dolby System - How it Works" (October 1969 Radio-Electronics), "All About Dolby" (June 1971 Radio-Electronics), "The Dolby Technique for Reducing Noise" (August 1972 Popular Electronics), "The Dolby Noise-Reduction System" (May 1969 Electronics World).

For lower noise on tape playback and FM stereo broadcasts, Dolby may be the answer

By Steve Leckerts

Every way we have to store or send messages, also stores or sends noise along with the message. There is no known way to eliminate noise completely from any message. However, the degree to which noise actually interferes with the messages is not the same for all media. For example, we can readily ignore the "noise" of small marks and defects in a newspaper while we read an article. On the other hand, even loud breathing by a neighbor at a concert is enough to destroy the mood or sensation that the composer and performers are trying to create.

Since all noise is caused by real physical disturbances, noise cannot ever be completely removed. However, the amount of noise heard by a listener depends upon certain conditions, some of which can be controlled to reduce noise.

In magnetic tape recording, for instance, the predominant noise heard is that produced by the tape itself. This tape noise (hiss), in turn, results from the nature of magnetic recording-magnetizing particles of a powder with which one side of the tape is coated. When the sound being recorded is a loud one, more particles are magnetized. When the sound is soft, fewer particles are magnetized. Because the number of particles varies slightly from one place on the tape to the next, the intensity of the recording cannot be made perfectly uniform. When the tape is played back, there is a slight fluctuation of loudness heard by listeners as a steady hiss super-imposed on the recorded program. During quiet passages or intervals in the music this hiss can be quite noticeable and rather disturbing.

One way to reduce this tape noise is to record at the highest possible program level - to increase the power of the electrical signal before recording so we can magnetize more particles for a sound of given loudness. Then, when the tape is played back, the sound from the speakers will be louder. When the listener reduces the volume to restore the original level, he also reduces the noise of the tape, since the noise remains the same no matter how the recorded level is varied.

Obviously this technique has limits. At some level, when the program becomes quite loud, all the magnetic particles on the tape are magnetized. Now it is impossible to record a sound which is even slightly louder, for it still gives the same result - all of the particles are magnetized. In the best recording tape used in professional recording of classical orchestral music, this maximum level is often reached at a point at which the quietest sounds of solo instruments are barely audible above the noise. For this reason, the dynamics of recorded music must be sacrificed if the listener is to hear the music without exceeding normal tolerances of distortion and noise.

Much of the disturbing noise of a tape recording is in the higher frequencies. This has led to recording high frequencies at an increased level, and then attenuating them during playback. We call this equalization. The technique works because high-frequency portions of the program are normally at lower levels than other parts of the original program, so there is not as much danger of saturating the tape. However, this is only a statistical truth. It is usually, but not always applicable. Because of equalization, loud cymbal crashes or other passages that have loud high-frequency sounds may be distorted even on the best recordings, unless these instruments are positioned at a distance from the microphones during the performance.

The Dolby Noise Reduction System

The Dolby noise reduction system takes advantage of psychoacoustic phenomena in a way that enables a substantial reduction of noise without any other audible effect on the program. Intrinsic to the system is a reliable method of electronic control. Large numbers of noise reduction units can be built that are fully and exactly compatible with each other. Any recording or broadcast made with one Dolby-type unit can be played or received through any other Dolby-type unit.

There are two versions of the Dolby System. One, the "A" system, is designed for professional use, and suppresses noise, hum and other disturbances over the entire frequency range. The simpler "B" system has been specifically developed to reduce noise in broadcasts and recordings for home listening. The entire "B" system circuitry, completely assembled on a circuit board, can be made by a manufacturer for less than $10. Built in as a part of a receiver, a Dolby "B" system might raise the list price by approximately $50.

A lot of research has been aimed at investigating the psychoacoustic effect, known as masking, by which a louder sound conceals from the listener the presence of a different, softer sound. This effect plays an important part in the operation of the Dolby noise reduction system. Noise is only disturbing to the listener when the program level is low enough to permit the noise to come through and be heard. A trombone playing at high volume, for example, masks the sound of a tambourine being played softly at the same time, or the sound of noise caused by recording tape.

An important peculiarity of the masking effect is that it does not occur when the two sounds concerned are very different in pitch. Therefore, a trombone may mask the sound of a tambourine, but even a very loud note of a bass drum will not, because tambourine and bass drum are so different in pitch. The same effect occurs with the masking of noise during the playback of tape recordings or FM broadcasts, since much of the noise which is heard by the listener is high-pitched noise in the form of a steady hiss. The noise is. therefore, more easily masked by high-pitched sounds than bass notes.

As discussed earlier, increasing recording level and reducing playback level can be used to reduce noise. However, the effectiveness of this approach is limited, since at some point the recorded level simply can no longer be increased. In disc recording, too, a maximum level is reached, when the grooves run into each other. In FM broadcasting, the limit is the point at which overmodulation (i.e., excessive deviation) occurs.

The Dolby System integrates the ideas of masking and automatic level control. The system automatically increases the recording or broadcast level of quiet musical passages which could not mask noise, and then reduces the level of the same passages during reception or playback. In the process, the original sound is exactly restored, but noise which would otherwise be audible is greatly reduced. The encoding, during broadcast or recording, and the decoding, during play-back, are done by circuits which are nearly identical and can, in fact, perform either function if appropriately wired or switched. Because the system can analyze the program so quickly as to make its operation inaudible, noise is suppressed without changing any other program characteristic which can be heard.

Since the encoding process has no effect at all upon loud parts of the program, it cannot cause excessive levels to be reached during recording or broadcast. Its effect upon the program sound is so subtle that listeners who do not have decoding devices often cannot even tell that the program they are hearing has been processed in some way. They may even believe that it has been improved over the original, especially if they listen with inexpensive equipment. When the encoded signal is heard through wide-range equipment without Dolby circuitry, it is noticeably brighter in overall sound. However, if the treble tone control is adjusted slightly, even this effect disappears. The system is therefore compatible, in the sense that its use does not require that listeners own playback equipment with Dolby circuitry unless they wish to gain the advantages of noise reduction which the system provides.

There has been no noise reduction technique in the past which has been considered acceptable by any significant number of recording engineers or artists. The Dolby System, however, is now used by more than 300 recording and film companies. Its application is mandatory in classical recordings at many record companies.

Problems of FM Broadcasting

From the standpoint of the high-fidelity listener or broadcaster, FM presents several serious problems. The problems arise largely because the potential of FM broadcasting as a medium of the highest technical quality, potentially superior to any form of recording available to the home listener, has been eroded by a series of technical changes. As a result, the exceptional channel by which FM gains its special identity has not been fully preserved.

Much has been said and written of the equalization used in FM broadcasting, particularly of the restricted level of high frequencies imposed by the standard. The transmission of high quality recordings or live broadcasts is often made impossible without artificial limiting of program levels.

The system of stereo multiplex broadcasting in general use has degraded reception considerably. Even the theoretical minimum increase in noise produced by stereo broadcasting is nearly 24-dB per channel, and this figure is attained by only the very best circuits.

SCA background music services which are economically necessary for the survival of many FM stations, impose technical requirements on tuners for home use which are rarely met. Even tuners of very high quality (and cost) often emit high-frequency "chatter" when both SCA and stereo multiplex broadcasts are transmitted.

Experiments show that the Dolby "B" system has pronounced effects upon broadcast reception because of the system's highly effective noise reduction.

When the "B" system is used during broadcast, and listeners equipped with decoders, the result is approximately 1 0 dB of noise reduction. Ten dB is the difference in level when power is changed by a factor of ten.

The improvement can be interpreted in various ways. It is, for example, the same additional quieting which listeners in various locations would obtain if the power of a 50,000-watt ERP (Effective Radiated Power) station were increased to 500,000 watts. On the other hand, it is the increase in quieting which would result if the sensitive figure of all tuners, usually given in microvolts, were divided by slightly more than 3. Thus, a tuner with about 10-μV sensitivity for 20 dB quieting would now achieve that degree of quieting at about 3-μV. The result is extending coverage to an area much larger than that which the station could reach formerly.

It is in the indirect effects it may have that the Dolby system offers significant benefits to broadcasters and listeners. Obviously the availability of "A"-type tape recordings brings a new standard of quality in musical source material to stations even without any other change in technique. Several FM stations have already broadcast such "A"-type tape recordings. In the fall of this year, Dolby Laboratories and Decca (U.K.) will make available to FM stations a series of "A"-type master tape recordings made during the past few years by Decca. These performers have never before been available to broadcasters in "A"-tape form. If these stereo programs are broadcast and received using the "B-system, listeners will be able to obtain quality impossible to achieve today.

The characteristic proposed by Dolby Laboratories for FM broadcast noise reduction is the same as that of the "B"-system now used in tape recorders and pre-recorded tapes. The Model 320 noise reduction unit, now used to encode such tapes, could also be used for the FM broadcasts by radio stations, and the products that use the "B" system, that are available to consumers can be used to receive the broadcasts in decoded form. All in all, it looks like better FM music reproduction is on its way.

Posted June 5, 2024
(updated from original post on 5/14/2017)