CAPS 2026 APN

A typographer by trade, Scott learned stenography and became deeply interested in the possibility of recording speech mechanically. This research led him to invent the phonautograph, a machine capable of recording sound vibrations visually.

The phonautograph consists of a funnel into which one speaks, positioned in front of a rotating drum covered with paper blackened by smoke. At the end of the horn is a diaphragm that vibrates in response to sound. Attached to the center of this diaphragm is a wild boar’s bristle, which traces the vibrations onto the smoked paper. The recorded traces are fixed by immersing the paper in water mixed with egg white.

Scott patented the phonautograph on March 25, 1857—exactly 20 years before Edison’s phonograph— under the name “speech writing by itself.”

My own third significant encounter with the phonautograph occurred in 1989 at the Musée de la Civilisation during the exhibition “From Cylinders (Edison’s) to Laser.” For this exhibition, I loaned several artifacts from my own institution, Le Musée Edison du Phonographe.

*Scott's original 1857 patent drawing*

That museum was closed in 2017 following a stroke that left the right side of my body paralyzed. For the 1989 exhibition, the museum borrowed a phonautograph from Utrecht in Holland, as at that time no phonautograph was known to exist in Quebec.

In September 1994, I wrote to Utrecht to ask for phonautograms, as my goal was to make them speak by one method or another. My first real contact with the phonautograph dated back to the early 1980s, when Allen Koenigsberg, a collector in New York and owner of one of only four known examples at the time, sent me a copy of Scott’s 1857 patent and its 1859 addendum. He asked me to translate these documents into English, particularly the sections describing the installation of the diaphragm and stylus, which I did.

The Utrecht museum replied that they regretted they had no phonautograms. In November 1995, I wrote again to the curator, pointing out that their Phonographic Bulletin of April 1977 mentioned phonautograms in its references. I received no reply. In July 1997, using the Internet, I sent an email containing the text of my unanswered letter.

The curator replied that he had since changed positions within the university, explaining why the letter had probably been lost. Once again, the answer was: no phonautograms.

*Catalogue record, Musée de la Civilisation, Quebec City*

That same summer of 1997, while visiting the reserves of the Musée de la Civilisation with Mme Toupin, I made an unexpected discovery: a “Duhamel vibroscope,” corresponding to the rear section and drum of a phonautograph. A few minutes later, we also located the metallic horn, shaped as a paraboloid of revolution, though still lacking the recording head. By consulting the list of scientific instruments purchased in Paris in 1858 by Abbé Hamel for Laval University, we were able to identify the origin of this instrument.

I had originally gone to the reserves searching for a curious hand-driven Berliner gramophone that I had seen in the 1970s at the Petit Séminaire in Quebec, and again in the early 1980s at the seminary’s museum.

*Correspondence with Teylers Museum, Haarlem, Holland*

In September 1997, I contacted the Utrecht curator once more, asking if he might lend me the recording head so I could make a copy. He refused, but promised to send sketches. He also informed me that the Teylers Museum in Haarlem, Holland, possessed a phonautograph.

I immediately wrote to request a loan. In October, they declined but promised to provide information.

On September 26, I wrote to the Musée de la Civilisation to explain my project of making phonautograms audible. To do this, I needed access to a complete phonautograph, which I had been unable to locate elsewhere despite years of effort. I emphasized that this would be a world premiere and could not go unnoticed.

*David Giovannoni*

In April 1998, I again wrote to Mme Toupin requesting a loan of their phonautograph so I could make a complete copy. I received no response.

In November 2000, Mme Toupin and I returned to the reserves to continue our search for the missing recording head. Our plan was simple: I began from the location where I had found the vibroscope, and she started from where the horn had been discovered. As we opened boxes and drawers and gradually closed the distance between us, she finally found the recording head.

That same month, I submitted a new request—this time for a long-term loan of the phonautograph for my museum in Sainte-Anne-de-Beaupré. This request was accepted in December, and in January 2001 the phonautograph arrived in Sainte-Anne. After seven years of continuous effort, I was finally able to produce phonautograms, having been unable to locate any original examples.

*Original phonautograph*

Between 2001 and 2007, I produced several phonautograms. During this period, I contacted a university professor in the United States—the same individual David Giovannoni later contacted in 2008 to make the famous 1860 phonautogram “Au clair de la lune” audible, now accessible online through Wikipedia.

Unfortunately, because I was unable to properly fix the smoked paper and because the professor refused to work with photocopies, I had to abandon the project— though I was, quite literally, only a wild boar’s hair away from success.

In 2008, David Giovannoni traveled to Paris with a high- precision scanner to digitize original phonautograms. That same year, I began building my first phonautograph.

Unable to reproduce the version with the metal paraboloid horn, I instead chose to replicate the model shown in Scott’s patent, featuring a wooden ellipsoidal horn. In April, I experienced what I call my “chemin de Damas”—a sudden understanding of how lateral recording on paper was possible even though the diaphragm vibrated perpendicularly to it. I was driving back from Wayne, New Jersey, still reflecting on the problem, when the solution became clear: the diaphragm is not parallel to the axis of the drum, but set at a slight angle.

The timing was ideal. I delivered my first machine to Mr. Eric von Grimmenstein, president of the Dictaphone Museum in Indianapolis, Indiana. In October 2014, I delivered my second machine to David Giovannoni.

Construction of the octagonal horn

In early 2016, I began work on a third machine for the Phono Museum in Paris, to commemorate the 200th anniversary of Léon Scott de Martinville’s birth on April 25, 1817.

On that occasion, Scott’s grandson, Laurent, was present, and I had the opportunity to speak with him. He told me that a phonautograph had once been kept in his family’s holiday home but had been stolen. He could not say which type it was. It was almost certainly a model with a metal horn, as no original wooden-horn phonautograph like the ones I build has ever been found. However, it may also have been an experimental version previously undocumented.

The machine itself is large, mounted on a wooden base measuring 22 by 31.5 inches. To fabricate the octagonal-profile horn, the first step is to create eight curved “petals.” These are made from special bendable mahogany planks, each consisting of two layers glued together with contact cement on a custom mold. Once removed, they permanently retain their shape.

The eight petals are then arranged in order to preserve the orientation of their surfaces. Each petal is cut so they can be joined two by two, with their edges cut at an angle of 67.5 degrees, and then assembled four by four.

At this stage, the horn exists in two halves. Before joining them, plaster of Paris is applied to form an ellipsoid of revolution. To achieve this, I built a special rotating device that allows the plaster to be shaped while still wet. I use a prepared spackling material that appears pink when wet and turns white when dry.

All metal parts were already available, as I had manufactured them in 2008 in anticipation of building four possible machines. The components supporting the horn and the drum were copied directly from the original parts of the Quebec phonautograph and laser-cut from a metal plate one centimeter thick.

The phonautograph reminds us that the recording of sound began not with playback, but with the desire to make the voice visible and permanent. My own work with this instrument reflects that same curiosity and perseverance. Through reconstruction and experimentation, the phonautograph continues to speak—not by sound alone, but through the traces it leaves behind.

*Replica of Scott's Phonautograph built in 2014 by Jean-Paul Agnard*