The Birth of the
Recording Industry

For all of human history before 1857, sound was irreversibly transient. A voice spoken, a note played, a thunderclap — each existed only for the moment of its creation and then was gone forever. The idea that sound could be captured, stored, and replayed would have seemed as fantastical as bottling lightning.

The science of acoustics was well understood by the mid-nineteenth century. Researchers knew that sound was vibration — that a voice set the air in motion, that these waves could be observed in the movement of a taut membrane. What no one had yet thought to ask was whether those movements could be made permanent.

The answer, when it came, arrived not from a musician or an entertainer, but from a Parisian typographer and amateur inventor with an interest in shorthand and a passion for capturing the ephemeral.

Édouard-Léon Scott de Martinville was not trying to play sound back. That possibility had not occurred to him. What he wanted was a machine that could transcribe the spoken word automatically — a mechanical stenographer.

His phonautograph, patented on March 25, 1857, used a horn to focus sound onto a taut membrane. Attached to the membrane was a stiff bristle, which rested against a cylinder coated in lamp-black soot. As the cylinder rotated and the sound vibrated the membrane, the bristle traced a wavy line through the soot — a visual record of the sound wave itself.

Scott called his tracings phonautograms. He could see sound for the first time. But he could not hear his recordings. The idea of playing them back does not appear in any of his writings. The phonautograph was a tool of scientific observation, not reproduction.

It was not until 2008 — 151 years later — that researchers at the Lawrence Berkeley National Laboratory used optical scanning technology to play back Scott’s phonautograms. The ghostly sound that emerged, a wavering fragment of a folk song recorded in April 1860, is the oldest known recording of a human voice.

Thomas Edison came to sound recording sideways. He was working on a machine to transcribe telegraph messages automatically when he noticed that the indentations left by the stylus on paper could produce a faint sound when played back at speed. The observation was the spark.

In the summer of 1877, Edison sketched a design for what he called a phonograph — Greek for “sound writer.” His machinist, John Kruesi, built it in thirty hours. Edison wrapped a sheet of tin foil around a grooved metal cylinder, held a diaphragm-and-needle assembly against it, and turned the hand crank. He spoke the words “Mary had a little lamb” into the horn.

Then he repositioned the needle at the start of the groove and turned the crank again.

The machine spoke back to him in a faint, scratchy approximation of his own voice. Edison later said he was frightened by it. After a lifetime of invention, this was the one that genuinely surprised him.

The phonograph was demonstrated publicly in November 1877 and caused an immediate sensation. Scientific American described the scene: “the machine inquired as to our health, asked how we liked the phonograph, informed us that it was very well, and bade us a cordial good night.” Within weeks, Edison was demonstrating the device to President Rutherford Hayes at the White House at 11 p.m., the President so delighted that he called his wife to come and listen.

The tin foil phonograph was a marvel but a fragile one. The foil deteriorated after only a few playings, and the sound quality was poor. Edison, absorbed by his work on electric lighting, set the phonograph aside for nearly a decade.

Into the gap stepped Charles Tainter and Chichester Bell, who patented the graphophone in 1885. Their crucial innovation was the wax cylinder: harder, smoother, and far more durable than tin foil. The sound quality was noticeably better. They approached Edison about combining their improvements with his patent, and he refused — then raced to produce his own wax cylinder machine.

The result was a brief patent war that produced unexpected benefits for consumers. Competition forced rapid improvements. By the early 1890s, wax cylinder phonographs were being sold commercially, with pre-recorded cylinders available featuring brass bands, comic monologues, and operatic arias sung directly into the recording horn.

The problem was duplication. Every cylinder had to be recorded individually. An artist would sit before a bank of phonograph horns and perform the same song dozens of times a day, making ten or twenty recordings per take. It was exhausting, inconsistent, and fundamentally incapable of scaling into a mass-market industry.

Emile Berliner solved the duplication problem. A German immigrant working in Washington D.C., Berliner patented the gramophone in 1887 and introduced a decisive innovation: the flat disc record pressed from a hard rubber master.

The disc format meant that a single master recording could produce thousands of identical copies by stamping. The economics of recorded music changed overnight. What had required a performer to record the same song fifty times could now be pressed in an afternoon. The recording industry — as a true industry — became possible.

Berliner also understood promotion. He sought out the best performers of the day to record for his label, understanding that the value of his format depended on the quality of the music pressed onto it. In 1894, he established the Berliner Gramophone Company, the first enterprise dedicated specifically to the sale of recorded music.

The recording industry existed by 1900, but it had not yet proved itself as a medium for serious art. Records were still widely regarded as novelties — convenient for comedy sketches and marching bands, but unsuitable for the full emotional power of great music.

Enrico Caruso changed that. The Italian tenor made his first recordings in Milan in April 1902, singing into a large horn in a hotel room. The results were extraordinary. For the first time, listeners heard a voice of genuine greatness reproduced with enough fidelity to feel its power.

Caruso’s records sold in extraordinary quantities. They demonstrated that recorded sound could carry real artistic weight — that buying a record was not merely a purchase of novelty, but an acquisition of culture. The market for serious recorded music, which had barely existed, expanded rapidly.

Victor Talking Machine Company licensed Berliner’s patents and launched its Red Seal Records label in 1903, featuring Caruso and other major artists at premium prices. These were not cheap novelties. They were prestige objects, sold to people who would previously have attended live opera. The recording industry had become, for the first time, a cultural institution.

The years between 1857 and 1925 — from Scott’s sooty cylinder to the electrical microphone — represent one of the most compressed periods of technological transformation in history. In less than seventy years, humanity went from having no means of preserving sound whatsoever to a mature global industry capable of distributing the voices of the greatest performers of the age into millions of homes.

The implications extended far beyond music. The same technologies that preserved opera arias also preserved speeches, poetry readings, comedy, drama, and eventually the spoken word in all its forms. The audiobook — the medium at the heart of what PhonoBooks exists to serve — is the direct descendant of those first wax cylinder recordings of the 1890s, when performers first sat before a horn and read aloud for a machine.

The public domain recordings available today on LibriVox and the Internet Archive — thousands of hours of classic literature read aloud by volunteers — carry forward a tradition that began in a Paris garret in 1857, when a typographer pressed a bristle to a soot-covered cylinder and asked a simple question: what if we could keep the sound?