Recorded Sound

Home > Support: Library > Speaker Design > Recorded Sound

All for the sake of entertainment

By Roy Johnson, loudspeaker designer, Green Mountain Audio, Inc.

The music and soundtracks we take for granted grew from the efforts of many individuals, as did the development of the speaker. Here we document the important breakthroughs, unpredictable paths across years of experiments, and flashes of insight. We leave you with an understanding of what you can expect to hear from recordings, for it will not be reality.

Uncertain beginnings

In December 1877, Thomas A. Edison demonstrated his first recorder-player to the editors of Scientific American magazine, having conceived of how it should work just the previous summer. It had a cylinder wrapped in a layer of very thin tin foil, upon which sound pressures forced a stylus into and out of the foil while the cylinder rapidly spun. The cylinder also moved slowly sideways, so the stylus produced an indented spiral path around it.

His 'phonograph' was cranked by hand at 2+ spins per second. One stylus and diaphragm assembly recorded, and another on the backside played back the two minutes of indentations. One was expected remove and store the foil, then put on another piece for a new recording.

Many inventions had already made Edison wealthy, yet he remained largely unknown to the public. Ten years earlier, he had invented an electric vote recorder, followed by his 'universal stock ticker.'

Just eight months before introducing his phonograph, Edison had patented a telephone transmitter ('microphone') that supplied much more signal to the telephone lines than what Alexander Graham Bell had accomplished. Edison had sold 150 of them to Western Union, which then formed the American Speaking Telephone Company to compete against the two-year-old Bell Telephone Company.

Western Union was already the famous telegraph giant, having the enormous advantage over Bell of 250,000 miles of wire hanging from 100,000 miles of poles. Bell had by then only installed 3,000 phones and still promptly sued Western Union, unfazed by its size. Such was the competitive atmosphere surrounding all the new communication inventions, something neither Edison nor others would escape in what would become the phonograph industry.

In the spring of 1878, a few months after his phonograph's introduction, Edison was hailed by a New York paper as "The Wizard of Menlo Park," but only because his medicinal preparation for pain relief had just been described in a magazine. One week later, Edison announced his hearing aid device. Later that year, he declared the problem of the electric light bulb solved and that his electric power system for cities would require a few more weeks' work. Gas company stocks plunged instantly, yet his research and development actually required another four years.

That press coverage finally brought Edison admiration worldwide, along with investors and international distribution. American investors paid him a handsome sum to allow them to form the Edison Speaking Phonograph Company to manufacture and sell his new device. For them, Edison quickly produced a simpler and improved version, which still did not work very well- but they all believed it would be bought for dictation only, taking its rightful place alongside the recently-perfected typewriter. In the same furious year of light bulbs and electric power, the first phonograph company began in the shadow of Edison's and J.P. Morgan's new Edison Electric Light Company, which soon became General Electric.

Edison's phonograph proved to be a hit outside the office, but only for its novelty of making funny noises, voices that squawked and how it distorted all musical instruments. That first model was very expensive and soon a smaller 'parlor model' was introduced. Still capable of both recording and playback, it sold for only $15, and encouraged founding of the first music-recording companies.

After 1882, Edison's work on electricity was out of the way, so he and his best friend Ezra Gilliland began working on an improved version of the phonograph. The existing model was still not selling well, competition was looming, and the only ones making money from it were traveling from town to town, charging admission to hear it make those amusing noises. As a point of philosophical consideration, up until then only 'natural sound' had been heard on the planet. Perhaps it is a good and proper thing that people first heard very low fidelity.

Competition arises

In 1881, the Volta Laboratory Association was formed by Alexander Graham Bell to quietly create an improved version of Edison's phonograph, again for dictation. Their final 'graphophone' was patented in 1886 and investors formed the American Graphophone Company to manufacture and distribute it.

Their graphophone indented its grooves into a revolving cylinder, as did Edison's. Unlike Edison's tin foil, its cylinder was made of hard wax. These cylinders produced less surface noise, had deeper grooves which did not skip, had better high frequency sound, and were less delicate than Edison's.

In 1886, before taking their device to market, American Graphophone proposed a merger with Edison, who refused and promptly bought back the assets of the floundering Edison Speaking Phonograph Company, re-organizing it as the Edison Phonograph Company.

Edison's original machine was still seen as "not only a clumsy piece of mechanism, frequently getting out of adjustment, but more especially because of the fact that the surface upon which the record was made was pliable, and likely to be obliterated by a mere accidental pressure upon it." (Harper's Weekly, July 1, 1886). He finished work on his 'Improved' Phonograph that same year in a flurry of activity.

Always the master of promotion, Edison arranged for the staging of a pre-introduction photograph. It showed him apparently lost in deep thought during a three-day marathon effort of final improvement. That image had been calculated to catch America's attention of this 'hard-working genius' -- and it did, with a color painting of it used for decades in his advertisements.

Edison's 'Improved' Phonograph bore many similarities to American's graphophone. Even American's wax cylinders fit onto his machine! For years, there had been wide speculation about leaks from the U.S. Patent Office, but in this instance, American had already placed ads and sold some machines. Thus, it is likely Edison already knew much about their graphophone. Still in that same year, Edison Records, Columbia, and other companies agreed to standards that also made pre-recorded wax cylinders possible for the first time.

Edison immediately sold (only) the U.S. sales rights for his 'Improved' Phonograph to Ezra Gilliland, who by then had introduced Edison to the woman who would become his second wife. Edison sold his patents and the rights to make his machines to his own newly-formed Edison Phonograph Works. He moved on to other projects, including designing miniature phonographs to make dolls speak. Edison's machine and American's Graphophone were still expensive contraptions, but unlike Edison's, American's version worked well enough to be widely adopted by the official reporters of the U.S. Congress; stenographer companies, and law offices around Washington, D.C. (Electrical World, July 14, 1888).

In 1887, Emil Berliner of Germany, by now an American citizen, patented his method for cutting lateral-motion grooves on a cylinder. Edison's and American's grooves were cut vertically, and Berliner's wiggled sideways. The significance of this would become apparent in a few years.

Mergers, betrayals, profit, and piracy

Seeing the widespread acceptance of American's graphophone in the U.S. capital, many people were determined to make easy money from these new recording/playback machines. In 1888, the North American Phonograph Company was established to lease phonographs and graphophones for dictation.

It then immediately invested heavily in American Graphophone, Edison's competitor, in return for the outright sales of that machine in the U.S. except in Virginia, Delaware, and the District of Columbia. During that same year, it also managed to purchase the U.S. sales rights to Edison's machine from Ezra Gilliland. Ezra profited greatly from this secret deal and so Edison sued his best friend of 20 years-- some say he never spoke to him again.

After Edison learned of North American's deal, he promptly sold them all of his phonograph patents and his Edison Phonograph Company, and then created the Edison Phonograph Works to manufacture the phonographs and cylinders for American and others. It published the first record catalog in 1890, listing both Edison and Columbia cylinder titles.

In 1889, North American and its American Graphophone division were purchased by Columbia, which had orginally been founded to sell Edison's 'Improved' Phonograph in Washington, D.C. The Columbia company discovered it made a lot more money selling music cylinders and had begun its own recording label. Copyrights and paying performers became issues.

In the meantime, Emil Berliner (left) had moved his lateral-groove method over to flat discs, because they were very easy to mass produce, not true for cylinders. Berliner introduced his 'gramophone' in 1891 (below right). It had a horn attached to its diaphragm, which was a first and also eliminated the ear tubes used on cylinder machines. Simple, hand-cranked, inexpensive, it offered only playback from its 7" discs, lasting the same two minutes as cylinders, spinning at ~70rpm.

At first, Berliner's sales were limited because his discs were molded in celluloid, which wore quickly and did not sound very good, especially on their inner grooves. Edison, Columbia, and others were still selling music cylinders, and Columbia was disturbed that its Edison 'Improved' Phonograph still did not work well, even as a dictation machine. But neither did its own graphophone. Foot-powered like a sewing machine, its speed varied annoyingly and the motorized one required office workers to frequently refill its batteries with noxious chemicals.

In 1894, Columbia allowed the North American Gramophone division to declare bankruptcy. The complex financial manipulations that brought this large company down made headlines news in the New York Times. The article goes on to explain how Edison was able to suspiciously re-acquire all his patents and all his sales rights to his machine for pennies on the dollar.

That same year, Berliner introduced longer-lasting, hard-rubber flat discs (left). Sales of his players and music took off so quickly that many cite this as the true beginning of the record industry. Two years later, Berliner created discs made with shellac because they sounded much better than the hard rubber ones. Breakable and still wearing quickly from the early 9oz. steel-needle pressures, shellac would remain the common material used for records until after WWII, when vinyl was introduced.

Also in 1894, Columbia introduced a home version of their graphophone for playback of music cylinders through a rubber tube. While the Berliner disc machine had its horn, it was still hand-cranked, while the Columbia model was spring driven. It spurred more sales of cylinder recordings and the founding of other recording labels.

By 1897, lawsuits and patent manipulations had given Columbia the rights to produce its own Edison-style cylinder players besides their graphophone version. Columbia took this opportunity to introduce the first inexpensive cylinder player (with a horn!) at a retail price of $10, right. In 1898, Edison introduced his own inexpensive spring-motor Gem model, lower right, which also featured a horn-- no more eartubes. Columbia followed with its own inexpensive, spring-motor FLAT disc player, the Columbia AH, lower left, to compete with Berliner, and began offering flat-disc recordings. 20 years after Edison's invention, consumers finally had choices of affordable and reliable music-playback machines and software.

In 1898, the first pirated Berliner flat discs arose, as they were so easily copied. Berliner successfully closed down one U.S. pirate after another in that same year, but when he then suspended record shipments to another blatant infringer, they sued him for not following the terms of his supply contract, and won!

Berliner's company was shut down. Unable to overturn the court's ruling, Berliner sold all of his patents to Eldridge Johnson, the manufacturer of Berliner's spring-motor disc players, and who also had recently invented a much better spring motor that made speed fluctuation a thing of the past.

Berliner had also just arranged for the trademark use of Nipper listening to his machine, originally painted with an Edison machine (which Edison didn't buy). Berliner passed that famous image over to Johnson who came to use it after re-naming his own company the Victor Talking Machine Company in 1901. According to some, the Victor name was in honor of Johnson winning his own patent lawsuit. Berliner's UK company remained, joining with EMI in 1931. His German division also remained, which had been called Deutsche Gramophone almost from its beginning.

Victor and Columbia would go on to compete in the manufacture of flat-disc players and recordings for many years. While Victor never made music cylinders, Columbia gave up on cylinders completely in 1912. Only Edison was left to promote his technology. His machines and recordings continued to undergo improvement, remaining highly competitive in sound, if not sales, with disc players until 1925. There was great pride in owning an Edison machine because of its beauty and workmanship.

Columbia's graphophone for dictation would undergo many improvements by the 1920's, along with Edison's version. Columbia eventually spun off their recorder as the Dictaphone company, competing with Edison's business machine for many decades. Columbia stopped manufacturing flat-disc players in the 1920s and later became a division of the Columbia Broadcast System (CBS). Sony purchased its record division in the 1990's.

Birth and growth of music recordings

Before electricity, the acoustic power of natural sound had to be gathered in by a horn to scratch a needle onto the cylinder or disc. Performers crowded close to the recording horn, and any glitch meant they had to begin again.

Early recordings would squawk or blast so much it was hard to tell a man's voice from a woman's. Drums could not be recorded because they made the stylus skip. Violins were inaudible, until Stroh introduced a violin with a horn to amplify the strings. Surface noise was intrusive, and the wax cylinders and celluloid discs quickly wore out. There was neither bass nor treble.

Rubber ear tubes were used before horns, so only one person could listen, typically men standing in train stations, drugstores, and arcades. They voted with their nickels for comedy monologues and funny voices -- not music. The first jukebox appeared in 1890 and used four ear tubes attached to one cylinder player.

In the 1890's, the cylinder recordings of John Philips Sousa became tremendous sellers for Columbia and spurred even more interest in finding artists and songs to record. In 1901, Edison's improved Gold-Moulded Cylinders debuted for fifty cents each and featured music from Edison's own recording studios. These were the first cylinders to be successfully mass produced. The price later dropped to 35 cents each.

Ten-inch flat discs were also introduced at that time, followed by 12" discs. Along with Edison's new cylinders, those could hold four minutes or more of music. In 1902, Victor began importing 'Red Label' 10" discs of famous European opera singers, such as Caruso, and gained endorsement for their machines by many leading opera singers from Europe. Victor would go on to become the largest seller of classical music by the 1920's, while Columbia and Brunswick produced many of the 'popular music' records.

New materials and technology continued to improve the sound of both disc and cylinder, and those only continued to increase the popularity of listening to music. Edison improved the sound of his cylinders once again in 1912 and introduced a diamond stylus, which made his cylinders and players quieter and better-sounding than Victor's flat-disc machines. He finally began producing flat discs later in the decade, Diamond Discs which would only play on his own disc machines.

By 1920 there were thousands of acoustically-recorded music titles on disc and cylinder in the U.S. One change that pre-recorded music had been making to American life was the gradual disappearance of musical instruments from the home. The introduction of radio had been cutting into record sales since before 1920, but the popularity of the new jazz music after WWI had offset much of that decline.

In the 1920's, a long-lived record title might sell 200-300,000 copies in the U.S. Popular 'race records' (by African-American artists), recordings of Irish accents, country bumpkins, and drunken monologues would sell a profitable 20-30,000 copies. During 1927, some reports say that nearly 1 million machines and over 40 million records were produced in the U.S. In 1929, the stock market crashed, with one casualty being the gramophone business. Why buy machines and records when one could hear the music for free on radio? In 1932, apparently only 40,000 players were produced and no more than 6 million records were sold in the U.S.

The four-minute playing times of flat discs and cylinders set the maximum length of popular songs until the introduction of the LP in 1948, and apparently this length still suits our modern attention spans. Berliner's 1887 principles for cutting lateral-motion grooves on flat discs remained effective until the release of the stereo LP in the late 1950's. It combined Berliner's lateral-cut with Edison's vertical-cut to obtain its two channels of sound.

Acoustic to electric

For many years, a master recording could produce only a limited number of copies. Extra masters had be made one at a time by the artists recording the song again and again, for an entire day. If a song proved tremendously popular, different artists might be found to quickly record new masters, with the names on the label left unchanged.

Then came 1925. An all-electric recording process for flat discs had been invented by Bell Labs and was introduced by their Western Electric division to radio broadcasters and record companies. A microphone's signal was amplified and sent to the electric-powered cutting head on a motor-powered turntable (still revolving at 78 rpm). Artists now had their own microphones and could sing instead of shout. Band members were positioned more comfortably, and multiple disc masters could be cut simultaneously from one session, ending callbacks for the band.

These new electric 78s were first produced by Columbia and Victor, and featured more bass and treble with much less resonance and distortion. Background noise had seemingly vanished and these sonic differences were easily audible on acoustic playback machines. State-of-the-art Victor Orthophonic Victrola acoustic players for 78s were simultaneously introduced in 1925. Their finest 1925 Credenza model cost at least $10,000 in 2008 dollars. All this marked the beginning of the end for Edison's cylinder recordings.

These new Orthophonic Victrola players produced better sound even from the older acoustic 78s, and had come from Victor engineers applying Bell Lab's recent research in horns from theater speakers, which in turn had come from the signal-transmission mathematics developed by the U.S. Navy in WWI. In 1926, Victor contracted with RCA for electronics inside their cabinets, including a radio, then merged with RCA in 1929 to become RCA-Victor. Only electric players were made afterward (except for portables) and Edison produced his last cylinder recordings by 1928.

The same all-electric technology brought sound to motion pictures. After the first successful sound films in 1929, U.S. theaters filled as never before despite the economic conditions at the time. Because film sound remained a much bigger market than recorded music for the next 20 years, almost all audio-technology breakthroughs were driven by film-sound needs. Record players were electrified but remained spinning at 78rpm.

It is interesting to note that the 70 to 78rpm disc speeds came first from observations that this was a comfortable rotational speed for hand-cranking, which produced a body rhythm similar to the cadence of marching and to the human pulse rate. It was also seen as most comfortable for the hand-cranking of farm machinery. Later, the 78 speed was easily geared to standard electric motors -- so there was no reason to change.

Consumption, marketing, and distribution

After WWII, the U.S. economy boomed. Consumer demand drove record sales. Singles had always been on a 10" or 12" 78rpm record and an entire symphony would occupy nine or 10 78rpm records, making an 'album.' However, all that changed in 1948 when Columbia Records first introduced the 33 1/3rpm long-playing microgroove record on the latest vinyl compound.

The original need for that 33.33 speed had come from experiments for film-sound in the 1930s, because it allowed a 20" disc to last the 11 minute length of a 35mm film reel. Some 12" music 33s and players had been sold in the 1930's but response had been limited by Depression conditions. The new microgroove technology allowed this speed to be the best choice to produce a record with a long playing time and high fidelity, and also be geared to standard electric motors.

To replace the bulky 78 record and player and compete for music sales, RCA introduced its own 45rpm records and player in 1949. By the end of that year, RCA's press releases claimed they were producing over a million 45s per month and could not keep up with demand! Soon licensed to others, 45s were smaller, far lighter, almost unbreakable, and less expensive than 78s. Some observed that 78 minus 33 gives 45, but this was only a coincidence, as evidenced by RCA's exhaustive 45rpm research papers for their new product.

Magnetic tape recording for studios took off almost overnight in 1949, although one Ampex mono tape recorder cost as much as a new house at the time. The convenience of tape and its ability to be edited forced the direct-to-disc recording process to be abandoned. This disappointed many as the direct-to-disc process did sound more 'alive' than a tape recording put onto disc (and still does). By 1954, two- and three-track recorders were in studios, and the first stereo recordings were released on 7" reel tapes.

In 1955, Frank Sinatra created what many call the first 12" 'concept LP,' In the Wee Small Hours, where all songs were composed around a mood of late-night loneliness and lost love. With Elvis Presley's appeal beginning in 1956, by 1957 there existed an actual 'youth market' for 45 and 78 singles, where none had existed before. The last 78s were produced in the early 1960's, although many promotional ones appeared in the 1990s (play those with a regular stylus, not a '78 stylus').

Throughout the 1960's, record companies scouted and signed new artists to meet the youth demand, often having no idea what might sell. Some artists were allowed to create whatever they wanted, others were guided by their assigned 'producers.' The A and R department (Artists and Repertoire) of a record label found the talent and helped guide their sounds, while talented recording engineers created new and unique sounds. Phil Spector and Quincy Jones became two of the most sought-after producers. Spector was famous for his 'Wall of Sound' mixing technique and Jones first was known as a great arranger.

The studio-system for artist development and management persisted until the late 1980's when recording equipment finally became affordable for artists to build their own studios. Many artists then left the big labels, signing with independents to obtain more percentage of their sales, especially after seeing Nirvana's success.

In the 70s, pre-recorded stereo cassette tapes with the Dolby noise-reduction system reduced the sales of records, eventually outselling them by the end of the decade. The CD's introduction in 1983 had an almost immediate effect on the sales of turntables in the U.S., especially premium models. For many reasons, though, the U.S. record industry was selling fewer recordings -- some say because the music was not very good. While we tend to agree, we suspect that the continued growth of cable TV and VCRs, the widespread appearance of decent-sounding car stereos, the emergence of MTV, and the advent of video games all had a great effect on USA music sales. Many claimed that the enormous sales of Michael Jackson's Thiller album saved Columbia Records from bankruptcy.

Beginning about 1990, several important changes came to the U.S. music distribution business: Large, regional record distributors were established, with the record labels signing contracts to sell only to them. The first result was only large store chains could afford the minimum purchases, which effectively shut out independent record stores and also allowed prices to rise. A few years later, these regional distributors stopped supplying free promo CDs to radio stations and most reviewers.

Today, U.S. stations must buy the song or entire CD, which keeps many new artists from being heard on the radio, because a station can only afford to buy the releases of established artists. Many station purchases are now downloaded, having less fidelity than the CD. MP3 downloads are tremendously popular for consumers, despite their limited fidelity. Fidelity is making a comeback with HDCD, XRCD, SACD, and Blu-Ray HD audio. Sales of LPs and turntables are increasing, for sonic and aesthetic reasons.

What can be recorded

A group and producer first decide how much music will be played together, and which instruments and voices will be overdubbed later. Most groups want to play together, to be able to see, hear, and react. Their engineer must then separate them enough so each does not bleed into another's microphone, but not so much they cannot see and hear each other. Regardless, each microphone still must be placed close to its instrument or voice, for sufficient isolation of the sounds.

Microphones placed close never hear what we hear -- 'the natural sound' -- disappears. That single fact leads to the choices of microphone and all processing of signals. Microphones each sound different -- there is no perfect microphone, only one best for its application and that artist. Some mics accentuate bass or treble, others soften a hard-edged voice. Some give a thin voice more power and aggression. Certain ones sweeten the sound of guitars or strings. What microphone works best comes from the engineer's experience.

The microphone's selection is also tied to the sound of a microphone preamplifier, because every mic preamp injects its own personality. The multi-channel mixer then has its sound, and if the mixing is done, perhaps in the computer (mixed "in the box"), then the computer's acoustic-effects software has its own sound.

Many engineers pull the signals out of the computer, run them through vintage audio processors, then back into the computer to get 'their sound.' Signals are dynamically manipulated by audio processors. Large signals are compressed, peaks limited, and soft signals are made louder or 'gated' (cut off completely). Voices can be 'de-essed' and double-tracked for a 'chorus' effect. Among other effects, echoes (reverb) are added, and the left-right placements are chosen - even for the reverb, which could itself be panned to an opposite channel, compressed, delayed, phase-shifted, and tone-equalized.

What we should hear

So, recordings were first demanded because they were entertaining, not because they sounded like the real thing. To this day, our best recording and playback technologies cannot capture and playback reality. They can only simulate realities, each chosen to suit a song or soundtrack. All we can then expect from a recording is entertainment, not reality, so what we need from our systems is fidelity, a faithful recreation of the recording. Thus, while no recording can be used to judge a system for reality, any recording can be used to judge the system for the aspects of fidelity such as clarity and musicality, which are so important to your enjoyment.

For any two channel recording, consider it was mixed for the two ears of someone seated in the middle. It was mixed also for playback in room without a great deal of echo. Under these conditions, you most easily hear the background acoustics, real or artificial, which inspired the musicians and improved their sound. Sitting in the middle, you most easily hear all their musicianship and artistry.

With the proper speaker setup, the end result can be a stereo image heard beyond the bounds of the two speakers. Each instrument and voice will lie somewhere along a line from left to right, with no 'height,' because microphones do not know 'height.' The instruments and voices will appear to be different distances away from you, from front-to-rear, which is called depth. In surround sound, that line of 'depth' will wrap all around you. When the speakers are in the wrong places, you hear everyone jumbled together. You miss much of the music and complex soundtracks are more difficult to enjoy.

• Recorded Sound - The Ideal Soundfield - Time and Phase Coherence - Math and Physics - Drivers - Adjustability - Cabinets - Ports - Circuits - Measurements

• Home > Support: Library > Speaker Design