Waxing poetic: Digitizing cylinder recordings

The National Anthropological Archives has been hard at work over the past two years digitizing our entire collection of sound recordings as part of a push to provide access to endangered language and culture. As the Arcadia Fund-supported project wraps up, we’re digging deep within the collection and coming up with some particularly esoteric recording formats. One of these formats is the wax cylinder, one of the earliest media for recording sound. Anthropologists such as John Peabody Harrington, Frances Densmore, and Jesse Walter Fewkes used this technology to make field recordings of American Indian language and music, lugging phonographs around the country to record speakers and performers. The NAA holds a collection of 13 wax cylinders of American Indian songs and dances, dating from the early 20th century, whose provenance is unknown. Since we don’t have the equipment to play these recordings safely, all our information comes from the labels scribbled on the containers—making for a mystery that can hopefully be solved through digitization.

The NAA’s archival holdings include a collection of 13 mysterious wax cylinders.

A few factors make this format particularly complicated to digitize. For one thing, few archives have cylinder phonographs lying around to even play the cylinders. The phonograph disc (which would evolve into modern-day vinyl records) won the market as early as the 1910s, and cylinders fell quickly out of favor—as a comparison, imagine trying to find a Betamax player in 2080. Another set of difficulties comes with the physical qualities of the format. Wax is soft enough that a few seconds of heat from your fingers will expand the surface and damage the etched sound recordings. Even playback on a period machine can damage the cylinders; many older phonographs had particularly heavy styluses, which allows for louder playback but causes degradation and even gouging in the grooves.

This wax cylinder (housed in an original Edison casing) had cracks and gouges, a sign of the fragility of the format.

These obstacles to access and preservation may be significant, but these recordings looked to be an invaluable resource to anthropologists and others who study languages and culture. Luckily, we were able to turn to an expert: David Giovannoni. David is a specialist in early audio media who owns an Archeophone, a modern machine with a gentle enough playback system that—when operated by a professional!—can safely play and record these wax cylinders. As he mounted the first cylinder, we leaned forward in anticipation as the stylus lowered and music crackled to life. We were listening to turn-of-the-century sound on the era’s own media—it doesn’t get much better than this for an archivist.

The Archeophone will play and transfer any type of cylinder recording, making it invaluable for archives specializing in early sound.

The cylinders turned out to be a selection of mostly love songs and lullabies. Though their origins remain undetermined, announcements at the beginning of the recordings along with the handwritten labels lead us to think they come from Plains and Southwest tribes. This information makes for a start in describing the cylinders, a process that will continue with the delivery of archival-quality sound files. These files will eventually be made available to visitors to the NAA, allowing for research and (hopefully) further identification by the public—all in a day’s work in the pursuit of knowledge through access!

Click here to download and listen to a brief sample of one of the transfers we made, introduced as “Indian Love Song.”

Annie Schweikert, Media Archivist
Human Studies Film Archives