History of FM and High Fidelity
The decades surrounding the Second World War transformed how the public listened to music and the spoken word. Two developments stood at the center of this transformation. The first was frequency modulation, a method of radio transmission that offered far cleaner sound than the amplitude-modulation broadcasting that had dominated since the 1920s. The second was the high-fidelity movement, a broad effort by engineers, manufacturers, and enthusiasts to reproduce recorded and broadcast sound with an accuracy approaching the experience of live performance. Together they reshaped both the technology and the culture of listening.
Neither development arrived easily. Frequency modulation was the work of a single determined inventor, Edwin Howard Armstrong, who struggled for years against commercial and regulatory obstacles, and whose system reached its full potential only after his death. High fidelity grew out of advances in amplifiers, loudspeakers, magnetic tape, and the long-playing record, and it gave rise to a new consumer category of separately purchased audio components. This article traces the invention of frequency modulation, its long contest with amplitude modulation, the introduction of FM stereo broadcasting, the rise of high fidelity and the LP record, the arrival of stereophonic sound, and the audio-component era that followed.
Edwin Armstrong and the Invention of Frequency Modulation
Edwin Howard Armstrong ranks among the most consequential inventors in the history of radio. Before he turned to frequency modulation he had already devised the regenerative receiver, the superheterodyne receiver, and the superregenerative circuit, three contributions that underpinned radio reception for decades. His search for a means of eliminating the static that plagued amplitude-modulation broadcasting led him to a radically different approach to transmission.
In amplitude modulation, the strength, or amplitude, of the radio carrier wave varies in step with the audio signal. The difficulty is that most natural and man-made electrical noise, including atmospheric static from lightning, also manifests as variations in amplitude, and a receiver cannot readily distinguish such noise from the desired signal. Armstrong reasoned that if the information were instead carried by varying the frequency of the carrier wave, while its amplitude remained constant, then a receiver could be designed to ignore amplitude variations altogether and thereby reject most noise.
Wideband FM and Noise Reduction
Earlier theorists had considered frequency modulation and concluded, on the basis of narrowband analysis, that it offered no advantage over amplitude modulation. Their analysis was correct as far as it went: a narrowband FM signal occupies about the same bandwidth as an AM signal and delivers about the same ratio of signal to noise. Armstrong's insight, developed through careful experiment in the early 1930s, was that the trade is not fixed. If the carrier is allowed to swing across a band far wider than the audio signal itself, the recovered audio improves faster than the noise does, and the ratio of signal to noise rises with the modulation index. Bandwidth, in other words, can be exchanged for quieting. Armstrong received four foundational FM patents in December 1933, conducted field tests from an antenna atop the Empire State Building, and demonstrated the system publicly to the Institute of Radio Engineers in November 1935. The demonstration showed that FM transmission could deliver audio of remarkable clarity, with a wide frequency range and a low noise floor that amplitude modulation could not match.
The advantages of wideband FM were not limited to static rejection. Because the information rides in the timing of the wave rather than in its size, an FM receiver can pass the incoming signal through a limiter that clips every peak to a uniform amplitude, discarding amplitude noise outright before a discriminator or ratio detector converts frequency deviation back into audio. The system also tolerates interference between stations more gracefully, because a stronger signal tends to suppress a weaker one sharing the same channel, an effect known as the capture effect. These properties suited FM to high-quality sound broadcasting and, in later decades, to two-way land mobile radio and other services in which clean reception mattered more than economy of spectrum.
The FM Broadcast Channel
The standards that grew out of Armstrong's work still govern FM broadcasting. A station in the United States occupies a channel 200 kilohertz wide, and at full modulation the carrier is permitted a peak deviation of 75 kilohertz on either side of its center frequency. Channels are centered on odd tenths of a megahertz, from 88.1 to 107.9 megahertz. The program audio itself extends only to 15 kilohertz, so the transmitted bandwidth is many times the bandwidth of the information it carries. That extravagance is precisely the exchange that buys the quiet background.
Broadcasters improve the result further with pre-emphasis. Noise recovered from an FM detector rises with audio frequency, so the transmitter boosts the treble before modulation and the receiver applies a matching cut afterward, restoring flat response while pushing the residual hiss down with it. North America standardized a time constant of 75 microseconds for this network; much of the rest of the world uses 50 microseconds. Wide deviation, limiting, and pre-emphasis together gave FM the quiet background that made it the natural home of high-fidelity broadcasting.
The Struggle Between FM and AM
Despite its technical merits, frequency modulation met formidable resistance. The established broadcasting industry had enormous sums invested in amplitude-modulation stations and receivers, and a thriving FM service threatened both that investment and the dominance of the major networks. The Radio Corporation of America, which had licensed Armstrong's earlier patents and whose principal business lay in established radio and, increasingly, in television, had particular reason to slow him down, for the spectrum FM needed was spectrum television also wanted. When RCA offered Armstrong a lump sum for a nonexclusive license, he refused. An outright payment carrying no royalties would have undercut the terms he had negotiated with the smaller manufacturers who were backing the new service.
Armstrong therefore promoted FM largely at his own expense. He built and paid for an experimental station, W2XMN, on the Palisades at Alpine, New Jersey, and began full-power broadcasting there on 42.8 megahertz in July 1939. Equipment makers and independent broadcasters followed. The Federal Communications Commission allocated the band from 42 to 50 megahertz to the new service and authorized commercial FM operation beginning on January 1, 1941. By the end of the war some fifty commercial FM stations were licensed and several hundred thousand receivers had reached listeners' homes.
The 1945 Band Reallocation
In June 1945 the commission moved FM to a higher range of frequencies, the band from 88 to 108 megahertz that the service occupies today, reserving 88 to 92 megahertz for noncommercial educational stations. The vacated low band went to television channel 1 and to fixed and mobile services. The commission defended the change on technical grounds, chiefly a prediction that sporadic-E and ionospheric skip propagation would cause severe long-distance interference at the lower frequencies during the coming peak of the solar cycle. Armstrong and many engineers disputed both the propagation forecast and the conclusion drawn from it.
Whatever its merits, the decision proved devastating in practice. Every low-band transmitter, antenna, and receiver became obsolete. Stations had to rebuild, listeners had to buy new sets, and the audience that FM had painstakingly assembled dispersed. The service lost years of momentum at precisely the moment when postwar manufacturing capacity and consumer appetite might have carried it forward, and the resources of the industry flowed instead into television.
Armstrong's Patent Litigation and Final Years
Armstrong spent much of his later life in costly litigation to defend his patents against large corporations that, in his view, used frequency modulation without fair compensation. He sued RCA in 1948, and the case ground on for years against an opponent with far deeper reserves. The legal battles drained his finances and his health. In 1954, exhausted and embittered, Armstrong took his own life. His widow, Marion Armstrong, took up the suits and settled with RCA later that year. She pursued the remaining actions for more than a decade and prevailed in nearly all of them, vindicating his claims long after his death. The episode stands as a sobering illustration of how a superior technology and its inventor can be ground down by commercial and institutional resistance even when the underlying engineering is sound.
For much of the 1940s and 1950s, accordingly, FM remained a minority service. Amplitude modulation retained the larger audience, the dominant networks, and the popular programming, while FM struggled with few listeners and little advertising revenue. Many FM licenses existed only as adjuncts to AM stations and simply rebroadcast the AM signal. The qualities that made FM superior for music would in time prove decisive, but that vindication came slowly.
FM Stereo Multiplexing
The fortunes of FM broadcasting improved markedly with the introduction of stereophonic transmission. By the late 1950s stereophonic sound had begun to reach consumers on records and tape, and broadcasters sought a means of transmitting two channels of audio, one for the left and one for the right, over a single FM station. Two constraints shaped the search. The system had to serve the existing base of monaural FM receivers, which had to continue producing a satisfactory single-channel signal from the same broadcast. It also had to leave room for the subsidiary background-music services that many FM stations ran under a Subsidiary Communications Authorization, because those services were often the only profitable part of the operation.
Several systems competed, among them a well-regarded design by Murray Crosby. The commission released its final order on April 19, 1961, adopting a compatible pilot-tone system drawn from closely similar proposals by Zenith and General Electric. The Crosby system, whose subcarrier arrangement left no clean room for the background-music channel, lost out largely on that ground. The standard took effect on June 1, 1961, and the first broadcast under it went out at midnight that morning from WGFM in Schenectady, New York.
The Composite Baseband
The pilot-tone system builds a composite baseband signal that the transmitter then applies to the carrier in the ordinary way. The left and right channels are first matrixed into a sum, left plus right, and a difference, left minus right. The sum occupies the audible baseband from 30 hertz to 15 kilohertz, and a monaural receiver reproduces it directly as ordinary mono. The difference signal modulates a suppressed subcarrier at 38 kilohertz in double sideband, occupying 23 to 53 kilohertz, well above the range of hearing. A low-level pilot tone at 19 kilohertz, exactly half the subcarrier frequency, both announces the presence of a stereo broadcast and gives the receiver a phase reference from which to regenerate the missing subcarrier. A stereo set recovers the difference signal, adds it to and subtracts it from the sum, and so reconstructs the left and right channels. A monaural set simply ignores everything above its audio range. Stations carrying subsidiary services place them on a further subcarrier, most commonly at 67 kilohertz, above the stereo subchannel.
Compatibility and Its Cost
This compatible approach resembles the method by which color was added to the monochrome television signal, and it carried a comparable penalty. The difference channel sits high in the baseband, where noise recovered from an FM detector is greatest, and the pilot and subcarrier consume modulation that would otherwise go to the program. Stereo reception therefore costs roughly 20 decibels of signal-to-noise ratio compared with mono on the same signal. That is why a distant station that sounds clean in mono hisses in stereo, and why receivers blend toward mono as the signal weakens.
Even so, the arrival of FM stereo gave the service a decisive advantage. Amplitude modulation, with its narrow bandwidth and higher noise floor, was ill suited to high-quality stereo music, and no compatible AM stereo standard ever gained lasting traction. Regulation reinforced the shift. In 1964 the commission adopted a nonduplication rule limiting how much programming an FM station in a larger market could simply rebroadcast from a co-owned AM station. Compelled to originate their own schedules, FM outlets turned to classical, jazz, and album-oriented rock formats that rewarded the band's fidelity. Audience followed programming, and by the late 1970s FM had overtaken AM in national listening, leaving news, sports, and talk to the older band.
The High-Fidelity Movement
While broadcasting evolved, a parallel movement sought to improve the reproduction of recorded sound in the home. The term "high fidelity," often shortened to "hi-fi," came into common use after the war to describe equipment that reproduced sound accurately across a wide frequency range and with low distortion. The movement drew on wartime advances in electronics and acoustics, and it appealed to a growing community of enthusiasts who treated faithful sound reproduction as both a technical pursuit and a serious hobby. High fidelity depended on improvement across the entire chain of reproduction, and the weakest link governed the result.
Amplifiers and Loudspeakers
Amplification set the first limit. In 1947 the British engineer D. T. N. Williamson published in Wireless World a power amplifier design that became the template for a generation. It combined triode-connected KT66 output tubes, a carefully specified wide-bandwidth output transformer, and a generous loop of negative feedback to reach distortion on the order of a tenth of a percent across the audible band, a figure earlier consumer equipment did not approach. Amateurs built the circuit from the published schematics, and manufacturers on both sides of the Atlantic adapted it.
Loudspeakers set the second limit, and for a time the accepted answer was size, since extended bass appeared to demand a large cabinet. In 1954 Acoustic Research, founded by Edgar Villchur and Henry Kloss, introduced the AR-1 and with it the acoustic-suspension principle. Sealing the driver in a small airtight box makes the trapped air, rather than the driver's own mechanical suspension, supply most of the restoring force. Air is a far more linear spring than a stiff surround, so bass distortion falls, and the designer wins deep response from a cabinet that fits on a shelf. Efficiency suffers, but amplifier power had by then become cheap enough to pay the price. The bookshelf loudspeaker followed from this trade, and with it a system suited to an ordinary room.
Phonograph pickups, tonearms, turntables, and tuners received comparable attention. The Audio Engineering Society, founded in 1948, gave the field a forum in which measurement and listening could be argued out together, and published specifications for distortion, frequency response, and noise became the common currency of the industry.
Magnetic Tape Recording
Magnetic tape recording contributed as much as any single technology. German engineers had refined the Magnetophon during the war, combining plastic-base tape with high-frequency alternating-current bias, a technique that linearizes the magnetic transfer and lifts fidelity far above what disc recording could offer. The American signal officer Jack Mullin brought machines and tape back from Germany in 1945 and demonstrated them in the United States. Bing Crosby, who wanted to record his radio program rather than perform it live twice, backed the young Ampex company, and the first production Ampex Model 200A recorders were delivered in 1948.
Tape changed practice as much as it changed quality. It offered a wide, flat frequency response and a low noise floor, it ran for half an hour without a break, and, decisively, it could be cut and spliced. Studios adopted it for recording and mastering, so the discs pressed for the public were made from far better originals. Consumer tape machines then extended the same capability into the home. Better sources and better home equipment together raised the standard of reproduced sound.
The Long-Playing Record
The phonograph record underwent a fundamental change at the end of the 1940s that proved essential to the high-fidelity movement. The dominant format, the shellac disc turning at 78 revolutions per minute, held only four or five minutes of music per side and suffered from high surface noise, because the abrasive filler that made shellac pressable also made it hiss. In June 1948 Columbia Records introduced the long-playing record, developed by a team working under Peter Goldmark: a vinyl disc turning at 33 1/3 revolutions per minute and pressed in both ten-inch and twelve-inch sizes.
The long-playing record, universally called the LP, transformed recorded music. For the first time an entire symphonic movement or an extended program could be heard without interruption, and the album emerged as a coherent artistic unit rather than a bound set of short discs. The quieter surface and wider dynamic range of vinyl suited the demands of high-fidelity reproduction far better than the shellac disc it replaced.
Microgroove and Record Equalization
Two changes made the LP possible. The first was the microgroove. Columbia cut roughly two hundred to three hundred grooves per inch where a 78 carried fewer than a hundred, and tracked them with a stylus of about one thousandth of an inch tip radius in place of the three thousandths used before. A narrower groove and a slower speed together yielded more than twenty minutes of music per side on the twelve-inch disc. The second change was the material. Polyvinyl chloride is quiet, tough, and free of the abrasive filler that gave shellac its characteristic hiss.
Cutting a disc also requires equalization. Bass recorded at constant amplitude would cut a groove so wide that adjacent turns would collide, and treble recorded at constant amplitude would disappear into surface noise, so the cutting lathe attenuates low frequencies and boosts high ones, and the playback preamplifier applies the exact inverse. Every label had once used its own curve, and a serious listener needed a preamplifier fitted with a bank of selector switches. In 1954 the Recording Industry Association of America standardized a single characteristic. The RIAA curve ended the confusion, simplified equipment, and remains the standard for vinyl playback today.
The Battle of the Speeds
The Radio Corporation of America answered Columbia's LP in 1949 with a competing format: a seven-inch disc turning at 45 revolutions per minute, with a large center hole and a changer mechanism that dropped the next record in seconds, designed above all for single songs. The two companies promoted incompatible systems, and consumers faced a confusing choice among three speeds, the older 78 together with the new 33 1/3 and 45. This rivalry, often called the battle of the speeds, ended not in victory but in a practical division of the market. The twelve-inch LP became the standard for albums, the seven-inch 45 became the standard for popular singles, and manufacturers simply built players that handled every speed. The 78 faded through the 1950s.
Stereophonic Sound
The next major advance was stereophonic sound, which reproduces a recording through two channels and two loudspeakers to recreate a sense of spatial width and the placement of instruments across a sound stage. The principle was not new. Alan Blumlein of EMI patented a comprehensive stereophonic system in 1931, including the idea of carrying two channels in a single groove, and Bell Telephone Laboratories transmitted a live orchestral performance in stereo in 1933. Neither effort reached the public, and the technique waited a quarter century for a consumer medium.
The enabling development was the Westrex 45/45 system, devised in 1957 at the Westrex division of Western Electric and adopted as the industry standard. Its two groove walls stand at right angles to each other, each inclined 45 degrees from the vertical, and each carries one channel independently. The geometry has an elegant consequence. When both channels move in phase, the stylus is driven laterally, exactly as a monaural record drives it, so monaural recordings play on stereo equipment and stereo discs remain compatible with monaural pickups. Out-of-phase content produces vertical motion instead. Lateral and vertical motion thus encode the sum and the difference of the two channels, the same decomposition the FM stereo standard would adopt four years later, and for the same reason: compatibility with the equipment already in listeners' homes. Mass-produced stereophonic discs reached the market in 1958, and stereo soon became the standard for recorded music.
The Audio-Component Era
The pursuit of high fidelity and stereo gave rise to a distinctive consumer category, the audio-component system. Rather than purchasing a single integrated console from one manufacturer, enthusiasts assembled systems from separately chosen parts: a turntable with its own tonearm and cartridge, an FM tuner, a preamplifier and power amplifier or a receiver combining tuner, preamplifier, and power amplifier on one chassis, and a pair of loudspeakers. The parts interconnected through conventions that no authority had formally decreed but that everyone observed, chiefly the RCA phono connector and a common understanding of line level. That informal compatibility was what made mixing brands possible at all.
The modular approach let a listener choose each element on its merits and replace one at a time as budget or taste changed. A substantial industry of specialist manufacturers grew up to serve the market, along with dedicated magazines, review columns, and retailers who let customers audition equipment in listening rooms. American, British, and, from the 1960s onward, Japanese manufacturers competed on measured performance as much as on reputation, and published figures for distortion, frequency response, and noise became ordinary marketing copy.
The component era reinforced the connection between FM broadcasting and high fidelity. A high-quality FM tuner became a standard element of a serious system, and the spread of FM stereo gave component owners a source of high-fidelity music over the air to complement their records and tapes. The transistor, which displaced the vacuum tube through much of consumer audio during the 1960s, made equipment cooler, more reliable, and more compact, and it carried high-fidelity components out of the enthusiast's workshop and into the general market.
Summary
The history of FM and high fidelity records the convergence of two efforts to bring accurate sound to the listener. Edwin Howard Armstrong invented wideband frequency modulation in the early 1930s, achieving a dramatic reduction of noise by varying the frequency rather than the amplitude of the carrier and by trading transmitted bandwidth for quieting. His system met sustained commercial and regulatory resistance, including the disruptive reallocation of the FM band to 88 to 108 megahertz in 1945 and years of patent litigation that ended only after his death in 1954. Frequency modulation nonetheless proved superior for music, and its advantage became decisive with the compatible pilot-tone stereo standard adopted in 1961, which carried two channels while preserving service to monaural receivers at a cost paid only in signal-to-noise ratio.
In parallel, the high-fidelity movement improved every link in the chain of reproduction, from the Williamson amplifier and the acoustic-suspension loudspeaker to magnetic tape. The long-playing record introduced by Columbia in 1948 provided a quieter, longer-playing medium suited to faithful reproduction, RIAA equalization ended the confusion of competing playback curves in 1954, and the battle of the speeds settled the LP and the 45 rpm single into their lasting roles. The Westrex 45/45 system brought stereophonic discs to market in 1958, and the audio-component system emerged as a consumer category in which listeners assembled equipment from separately chosen parts. The transistor broadened that market in the 1960s. Taken together, frequency modulation and high fidelity established the technical and cultural foundations of modern home audio, and the sum-and-difference encoding both adopted for the sake of compatibility remains a standard technique in broadcasting and recording.