Cylinder Recordings: Preserving Our Acoustic History
The wax cylinder audio recordings of the late 19th and early 20th century are an important resource for understanding the past and preserving cultural heritage. As the first commercially available audio recording and playback system, the phonograph with wax cylinders was the reigning technology from 1889 until 1910, and remained in limited use through the 1930s. Many of these early recordings have been sitting dormant for the past century. Since the early 1990s librarians and researchers have made a concerted effort to unearth and restore these historic assets. This effort involves assembling and preserving collections of wax cylinders, as well as the phonographs that play them, and making archival digital recordings of their content.
The value of these recordings lies in their potential to resurrect lost histories and forgotten cultural and artistic practices: from inspirational stories of human triumph such as Shackleton’s recounting of his ill-fated expedition to Antarctica, to the darker moments of our past, when “coon songs” and “Irish character songs” exhibited both ignorance and prejudice toward racial and ethnic groups. Until recently, what remains of our earliest audio recordings has been accessible to a select few. This is now rapidly changing, as major retrieval and preservation efforts are underway at libraries and research centers around the world. In one of the most ambitious projects, UCSB’s Cylinder Audio Archive, a large collection of wax and acetate cylinders are being preserved. In response to the fragility of these cylinders and the finite number of antique machines that can play them, UCSB is in the midst of a mass digitization undertaking that makes these recordings available to the public.
To understand the technological innovation of the wax cylinder and why preservation and access is important, we must first understand how it joined and changed its own historical moment. The history of wax cylinders begins in France in the 1850s. There Édouard-Léon Scott de Martinville pioneered a device that could turn sound into a visual representation. He called it the phonautograph, a device which influenced, and mechanically resembled Thomas Edison’s future inventions. It operated by using a gas lamp to leave soot on a piece of paper that was wrapped around a horizontal cylinder. As the cylinder rotated, a stylus attached to a small diaphragm bounced up and down creating a visual imprint of the person speaking; sound had been made visible. While there was no way of playing back the recording that had been made, this progress in translating sound to a physical material was critical to the eventual success of Edison’s phonograph.
In 1877 Thomas Alva Edison and his colleagues designed a machine that would be able to both record and reproduce sound. They named it the phonograph, a combination of the Greek words for “sound” and “writing.” Its first version used tin foil wrapped around a large cylinder. A stylus would then engrave the sound wave while the cylinder spun via hand crank. Sheets of tin foil, however, had multiple drawbacks as a recording medium. Tin sheets are fragile and can only be played back a few times. Moreover, once the sheet is removed from the cylinder it is impossible to replace it, much less get the grooves to line up so that it could be played again.
A major breakthrough in the evolution of this sound writing machine was the transition from tin to wax. Jonas Aylsworth, Edison’s chief scientist in charge of developing materials, took up the challenge to create something that was soft and malleable enough that it could be etched into, and durable enough to withstand multiple playbacks. The wax cylinders designed by Aylsworth were created to be reusable, by simply shaving off the top layer, producing a blank cylinder. Edison expected that the phonograph’s primary use would be as a business dictation machine. He did not envision it as the mass market consumer product it would become fifteen years later. Thus the potential of this machine as a device for popular entertainment was initially overlooked.
The phonograph with wax cylinders was the first commercially available audio recording and playback system. They were first demonstrated and made available to the public in 1888. To build them, Edison used non-specialized laborers who assembled pre-manufactured parts in the American system of manufacturing. Before the advent of this system, highly skilled laborers, who were expensive to employ, would have built each phonograph individually. This new system lowered the labor cost of manufacturing each phonograph as well as increased the speed of their production. This is one of the factors that allowed the price of the phonograph to drop significantly in a short period of time. In 1896 Edison introduced the Home Phonograph with a retail price of $40. By 1899 they were selling the Gem Phonograph for just $7.50 (de Graaf 106).
The cylinders themselves progressed through a series of mediums and production methods as they evolved toward popular use. In 1881, Edison had all but abandoned active development of the phonograph. It was still at the stage of using tin foil as the recording medium. This left a vacuum that Alexander Graham Bell and the Volta Laboratory attempted to fill, and wax was introduced as the recording medium. Called Bell-Tainter Cylinders, these “were the first removable and portable records used in commerce” (Wile 163). But the Bell-Tainter Cylinders proved too fragile and vulnerable to heat to succeed on a commercial level.
In 1887, after a decade devoted to developing the electric light, Thomas Edison resumed his work on the phonograph. The earliest blanks that his company produced were the “white wax” type. After experiments with different materials to make a core for support proved unsatisfactory, they went with solid wax for the blank. But these white wax cylinders proved impractical because of their softness. The subsequent “brown wax” cylinders were not made from wax, however they continued to be called such because of their moldability and waxy feel. Technically the materials were metallic soaps, created by combining stearic acid, caustic soda, and acetate of aluminum.
Ultimately celluloid, the first synthetic plastic, became the material of choice for commercially produced audio cylinders. Unlike its predecessors, the celluloid blank could not be recorded onto using a Home Phonograph. Instead, it was molded at the factory and contained commercial content. Celluloid had several advantages over wax when it was introduced. Physically sturdier than cylinders made from wax, it stood up better to repeated playing, and was unlikely to break when being handled. It was also much less susceptible to mold. As the celluloid aged, however, some of its faults and drawbacks become clearer. Celluloid shrinks when it loses camphor to the atmosphere. Edison Blue Amberol cylinders, for instance, had flanges that covered the edges of the core so when it shrinks lengthwise, the stress causes it to crack. If the core (often made from cardboard or plaster) absorbs too much moisture and swells, the cylinder can distort and become out of round.
While disc records were starting to become popular by this time, one advantage of using a cylindrical shaped carrier rather than a disc is that it allows the linear speed of the recording to stay constant. With discs, the linear speed decreases as you move toward the center giving you less space to record information, resulting in loss of quality. Both disc and cylinder recordings use a stylus to etch a groove into the recording medium. The difference between these grooves is that cylinders have vertical grooves (peaks and valleys) while discs have lateral grooves (side to side).
In 1888, ten years after the phonograph was first invented, Thomas Edison wrote an article in The North American Review, a literary magazine that continues to publish to this day. His article, “The Perfected Phonograph,” provides us with valuable insights into how the phonograph was used in his lifetime and, more importantly, where and how he thought it would be used in the future. Edison explained that a key feature of this latest model was a “turning tool” that allowed the wax cylinder to be, for all intents and purposes, erased. He claimed that each of these cylinders would last for “fifteen or twenty successive records before it is worn out” (Edison 645). It is clear that he envisioned his phonograph as an essential office tool for repeated transcription, but not as a tool for permanent storage. With this use in mind, he was extremely proud of this model’s ability to do what we now refer to as “pause and rewind.” Edison did realize that some people would probably want to save these cylinders as a permanent record, but because the technology was so new, he believed that the public needed to be explicitly told this. Edison advised that in order to keep each recording permanent, “the wax blank must not be talked upon again” (645).
As the article continued, Edison named a list of ten things that the phonograph is perfectly suited for. True to his original beliefs, Edison confirmed its primary use for office dictation. Edison included the reproduction of music as one of the uses in his list, but this rated as less important than the “teaching of elocution” (646). While Edison was unable to see the global importance of the phonograph for the mass distribution of music, he clearly understood its importance in history. He prophesied that this system would be utilized by world leaders and for the recording of historic events. He also predicted that future generations would eventually listen to perfectly clear audio recordings from the past, on a machine that would be in every household.
Projects around the world like the University California, Santa Barbara’s Cylinder Audio Archive are making this possible. UCSB’s collection of wax cylinders alone contains over 15,000 recordings, including both mass distributed popular music and personal recordings of both song and spoken word. Since 2002, the archive has been actively digitizing these cylinders and posting the audio files on the World Wide Web for the public to listen to and explore. If Edison could see where these cylinders have ended up, and how they are being preserved and transitioned to new technology with a truly global reach, no doubt he would be impressed.
UCSB has innovated a number of ways to fund their digitization project. The “Adopt a Cylinder” program was implemented both to secure funding and to engage the public with the collection. There are over 3,000 cylinders in the collection that have (as yet), not been digitized. For $60 a user can pick out one or more titles from the library’s catalog to be digitized and posted to the university’s website. The software UCSB uses for these requests is Aeon, developed by Atlas Systems, and was designed specifically for archives and special collections. These requests tend to take priority over whatever is next in the queue to be digitized, as the people who adopt cylinders are likely to notice a large lag-time between their request and its posting to the Internet. Even more important for public engagement, the university has made these recordings free to use for non-commercial applications under the Creative Commons Attribution-Noncommercial 2.5 License. However, commercial entities can license sound from the collection for varying fees, depending on quantity and use.
As part of my research, I visited Brendan Coates, the project’s Audio Digitization Technician, in their new facility at the UCSB Library’s Special Research Collections. Coates has been working on this project for the past two years, and his job includes the development of the workflow digitization. Coates was challenged to design a workflow that provided maximum efficiency, as well as being transparent and easy to follow. Step-by-step instructions have been posted on a wiki created specifically for this project. The wiki, a website that allows project workers to add and edit content, includes both written directions and screen-cap illustrations.
Mr. Coates gave me a demonstration of their digitization process. The machine that makes this endeavor physically possible is called the Archeophone, designed by the French hobbyist Henri Chamoux in 1998. Mr. Coates explained that one of the challenges that wax cylinders present is their complete lack of interoperability. For example, if a four-minute player is inadvertently used for a two-minute cylinder, it will carve a trough through the recording, permanently damaging it. Through a major feat of engineering, however, Chamoux has made a machine that can play any and all cylinder formats, by fitting it with various mandrels and styluses.
Before the cylinder can be played on the Archeophone however, it must be cleaned. UCSB uses a solution devised by the Library of Congress that contains ammonia, a strong detergent, and an anti-fungal solution. The cylinder is then placed on the appropriate size mandrel and fitted into the machine. The mandrel with the cylinder on it is then centered, using a set of screws to move it around the various axises. This step is critical, as it allows the pressure from the stylus to remain consistent throughout the entire recording and often takes up to thirty minutes. When asked if continual practice speeds up the duration of time it takes to center the mandrel, Mr. Coates responded that while he cannot complete the task any faster than when he first started, he has gained more patience by experiencing first hand, how critical this step is to making a quality recording.
Mr. Coates also explained that two recordings are made at the same time. The archival version (“A”) is captured and stored in RAW format.The broadcast version (“B”), is the recording that is available to the public and has been digitally enhanced to make it more intelligible. Additionally, it is is stored as an MP3 for easier accessibility. These tracks are recorded onto the computer at the same time, but their paths diverge slightly on their way there. The audio signal passes from the Archeophone through a Souvenir EQS MK12, a remastering phono preamp. From there, the signal branches out to create separate archival and broadcast recordings. The raw archival signal then travels through the Prism ADA-8XR, which converts the signal from analog to digital. Then on to the Lynx AES 16 src PCI card, and finally into WaveLab software on the computer. The signal that becomes the restoration copy available on the website takes a slightly different path. When the signal leaves the Souvenir, it is converted from analog to digital by the Sonifex RB-ADDA. Then, it is converted to the Cedar DCX, CRX, DHX, BRX (declicker, decrackler, dehisser, debuzzer respectively). It goes through the Lynx and ends up in WaveLab with an auto-naming file system. The dehisser is used sparingly because it often masks distortion in the original recording.
Before the tracks are recorded, the noise-reduction settings for the broadcast copy are decided upon. The decision of how much noise-reduction and cleanup to apply is given to the audio technician. This is a very subjective task, but has historically followed different trends. The current sentiment is that it is best to do as little augmentation to the track as possible, while still being listenable. Users listening to these tracks tend to care more about their historical significance and authenticity than their pure aesthetic value.
The broadcast version is what is posted to the website for download and streaming. While users cannot listen to the archival copy directly from the website, they can easily request access to them. Mr. Coates remarked that he is impressed that more and more people are interested in the “authentic” original, unrestored versions of these materials. Bearing in mind this ideal of authenticity, it is UCSB’s policy to digitize and post all of their cylinders, even if the quality of the recording is sub-par. The audio technician can make the decision to not load a track to the website if the content is inaudible and cannot be fully played through, or if the cylinder is simply too damaged. Usually this occurs when a transfer is not technically possible, but that rarely happens.
I also corresponded with David Seubert, the head of the UCSB Cylinder Audio Archive who has been with the project since its inception in 2002. According to Seubert, the collection now contains more international content, but the major framework and objectives for the project have changed very little over the years, and have been primarily technical in nature. This includes switching to 96/24 bit from 44.1/2 bit for digitizing. The department has added “memcache, a code repository, and a development server to make things more robust.” They have also switched from m3u streams to mp3 podcasts. The biggest improvement happened a year ago, when the website was rewritten from the ground up.
UCSB’s new library has a state of the art archival repository that Mr. Coates complimented highly, singing the praises of the project manager and the entire construction team. The building planners and designers worked with the archive staff to understand exactly what is required as well as desirable for such a facility. Their physical move into this new custom-built facility took about eight weeks, starting in December of 2015. The staff and a local company moved all of the furniture and equipment. BiblioTech, a library support services company, coordinated the collection move. In a follow-up interview, when asked about the challenges of the move, Mr. Coates responded “I was actually the first denizen of the new building and had it all to myself and my staff for about two weeks, which was really fun.”
Prior to the move, the cylinders were kept off-campus in a storage facility with a standard, but less than optimal, commercial-grade HVAC system. Despite this, no damage was ever found that could be attributed to the building or the HVAC. Fortunately, the new facility has state-of-the-art climate control, specifically designed for archives, including a cold-storage room where more sensitive materials such as magnetic tapes are stored. A large earthquake in the old facility would have been catastrophic for the collection, as the cylinders were kept on traditional shelving units. Their fragility means that a drop from even a modest height, could shatter the cylinders. They are now kept in locking drawers in large cabinets that are designed to handle seismic activity.
Prior to the digitization project, the cylinders were stored in their original cylindrical cardboard containers. As part of the project, UCSB worked with Hollinger Metal-Edge to design archival-grade containers. The final product was a box with a separate lid, padding on both the top and bottom, and a foam core in the middle to minimize movement. This product is now available for anyone to purchase. When rehousing the cylinder, the original lid is saved because it often contains a large amount of useful written information. The body of the container is usually not saved. The collection also contains a few of the Pathé Concert Cylinders, which have a much larger 5 1/4” diameter. These are stored in their original containers, as there is currently no suitable archival container in that specific size. Mr. Coates also stressed that it is important to remove any cotton padding from the inside of the original container, including the lid, since it can become a breeding ground for mold.
Despite great care and precaution, cylinders do occasionally break. Currently, there is only one person in the country who is able to mend broken wax cylinders, and he has a two-year backlog of cylinders to repair. His services are most likely to be sought if one of the Brown Wax cylinders breaks. Brown Wax was the medium available to consumers in blank form to be recorded onto, therefore unique recordings from a home or ethnographic trips tend to be stored on them. Many of the cylinders in the collection were mass-produced commercial recordings. In this situation it is faster, easier, and cheaper to obtain another recording, especially when they break, as it is often easier and faster to find a duplicate cylinder as opposed to repairing it.
UCSB Library’s cataloging department is responsible for the creation of the MARC record for the cylinders. This information comes primarily from the cylinder’s container and end engraving. However, if these items are missing (or the content is in question), an audio technician will use the Archeophone to let the cataloger listen to the recording. The cataloger places the item’s barcode on the Hollinger box, since there is no space where a label can be affixed. This does mean that if a cylinder were to become separated from its container, there is no quick way to reunite them. Fortunately, this scenario has never occurred.
One problem with the current cataloging system, however, is the lack of a field for technical metadata. It would be useful to have a record of the system and settings used in the digitization. Typically, cylinders are captured at 160 revolutions per minute, but not always. Sometimes, trial and error is required to find the proper playback speed. As of this writing, there is nowhere in the MARC record where one is able to look up what machine was used to play it, what the final rpm was, or the software involved in its capture. Looking at other standards does not provide any inspiration as to how to handle this issue. PBcore, the metadata schema specifically designed for sound and moving images, does have a field for “media materials” for film and video, but these options do not exist for audio media and materials.
Other institutions around the world have also embarked on the task of digitizing their audio cylinders. In 2008, a collection of 26 audio wax cylinders was donated to the Museum of London. Between 1902 and 1917, Cromwell Wall made recordings of his family in their New Southgate home. Of the 26 cylinders, 24 contained discernable audio content that can now be heard on the museum’s website. Included in the collection is a recording made on Christmas Day in 1902, and is the earliest known sound recording made on Christmas. These recordings give us a window on the life and sounds of a middle-class family living in London at the start of the 20th century.
University College Cork in Ireland has 114 recordings in their Henebry/O’Neill Wax Cylinder Collection. In an attempt of preserve traditional Irish music, Rev. Dr. Richard Henebry used the phonograph to record artists throughout the Déise region in County Waterford. His goal was to create a national archive of traditional Irish music at UCC. This didn’t happen in his lifetime, but the small collection he left to the university provides a wealth of information to historians, musicians, and ethnographers. Like UCSB, the project in Cork used the Archeophone to digitized the cylinders, and posted the recordings for the public to listen to on their website.
The modern day discovery of cylinder recordings (once thought lost or destroyed), can provide us with valuable historic and cultural treasures. In 2002 a collection of 200 cylinder recordings was discovered in Russia. These recordings were made in the 1930s by businessman and classical music aficionado Julius H. Block and donated to the Phonogramm-Archiv in Berlin upon his death. It was assumed that they had been destroyed during World War II, but fortunately they were moved to Silesia (located in modern-day Poland). From there, they made their way it to the Institute of Russian Literature, also known as Pushkin House. There they sat for decades, holding in silence their great secrets. Then, while writing a book on Jascha Heifetz, a Russian scholar discovered the recordings and realized what they were and their significance.
The recordings were made by Block between 1890 and 1923 in Russia and Germany. The list of musical talents that he recorded includes Nikolai Figner, Paul Pabst, Anna Esipova, and Lavrovskaya. Additionally, the collection contains spoken words from Leo Tolstoy himself. It also contains songs from prominent performers of the day for whom we previously had no audio recordings. These recordings are an excellent resource for music scholars. A 1901 Berliner recording of Figner’s “I Remember the Evening” has been available to the public for many decades. Ten years prior, in 1891, Block also recorded Figner performing “I Remember the Evening,” in E major, while the Berliner version is sung in F. This has caused some to speculate if our prior understanding of his timbre is flawed. To lay people this seems like splitting hairs, but is quite valuable information in understanding music history.
Early wax cylinder audio recordings also play an important role in ethnography, linguistics, and folklore. This has been recognized since the early days of the technology. Jesse Walter Fewkes made the earliest known recordings of Native American songs and speech. In 1890 as a practice run for his longer expedition to study the Zuni people, Fewkes recorded 35 cylinders with the Passamaquoddy in Pennsylvania, 31 of which now reside at the Library of Congress and are included in the National Recording Registry. It was well recognised by this point that the languages of indigenous people were on the verge of extinction. Fewkes was confident in the phonograph’s potential as an “aid of which the languages of our aborigines can be permanently perpetuated” (257). He captured both sung songs and spoken stories in their original language. To verify the accuracy and utility of this tool he played some of these recordings to Passamaquoddy who were not present at their recording. They were able to both understand the words spoken and provide an English translation. While these recordings were far from perfect, Fewkes was positive that he had “obtained enough to demonstrate the value of the instrument, in the preservation and study of aboriginal folk-lore” (280).
Across the Atlantic just a few years later Percy Grainger, renowned composer and pianist in the first half of the 20th century, also made great contributions to the field of ethnographic recording. In 1908 after his expeditions to record folk songs in North Lincolnshire, England, he published an article in the Journal of the Folk-Song Society entitled “Collecting with the Phonograph.” He aims to share practical knowledge gained on his expeditions in hopes of encouraging others to go out and make recordings themselves. He is enamored with this new technology and sees its usefulness in preserving “all details of performance for study by future generations” (150). Before the phonograph, handwritten notation was the only tool available to convey to posterity what one had heard.
One technical challenge in working with cylinders recorded on field expeditions, as well as commercially-produced cylinders, is determining the speed at which they were recorded. While the recording speeds of commercially-produced cylinders were not 100 percent uniform, 160 rpm was common enough to make it a good starting spot when attempting to play them. But the recording speed for cylinders recorded on expeditions often varied between recordings (even those made by the same ethnographer). This happened primarily because it enabled longer recordings to fit onto each cylinder. It was distracting and detrimental to stop a performance in the middle in order to swap in a blank cylinder. For ethnographers, the reduced audio quality that results from a lower speed was worth the trade off for a longer recording time. To further complicate things, some were recorded at multiple speeds. When an ethnographer noticed that a performance would not fit onto the cylinder, they could slow it down at any point in the recording.
Another challenge, which Adrian Poole notes in his article “Determining Playback Speeds of Early Ethnographic Wax Cylinder Recordings: Technical and Methodological Issues,” is matching the pitch of the sound traditionally made at the beginning of the recording. In the early 1900s it was common practice to sound a reference horn or tone at the start of the cylinder. However, these are frequently inaudible due to physical damage to the ends of the cylinders. Additionally, the sounding note was not standardized across regions and sometimes not even standard within a single region. When considering these challenges today, or passing judgment on past ethnographers, it is important to remember that for many of these people, audio recordings were viewed as a supplement or aid to the notes and transcriptions written in the field.
As scientists and archivists around the world continue to develop novel techniques for non-destructive playback, audio recordings made on wax and celluloid cylinders over 100 years ago are breathing new life into our appreciation of the earliest recorded sounds. A century ago, the technological world was dramatically changing in ways that would disrupt life in the home and office, and forever change the way we entertain ourselves. Ethnographers, linguists, musicologists, and ordinary people captured this rapidly changing world on wax and celluloid cylinders. But without the work of the archivists at the forefront of the recent movement to preserve and digitize these cylinders, these recordings and their auditory gems would have remained silent in drawers and cupboards. Preserved, digitized and made accessible, wax cylinders bring speakers alive with a multifaceted presence that the written word cannot convey. Suddenly the dead talk, music from an earlier era fills the room, and extinct languages come to life. These cylinders provide a temporal bridge across which lost dimensions of history, art, language, and culture travel in the form of sound.
Sources
Barnett, Kyle S. “Furniture Music: The Phonograph as Furniture, 1900–1930.” Journal of Popular Music Studies 18.3 (2006): 301-324. Web. 28 Feb. 2016.
Boström, Mathias. “The Phonogram Archive of the Stockholm Ethnographic Museum (1909-1930): Another Chapter in the History of Ethnographic Cylinder Recordings.” Fontes Artis Musicae 50.1 (2003): 22–35. Web. 3 Mar. 2016.
Breckbill, David. “The Dawn of Recording: The Julius Block Cylinders.” ARSC Journal 41.2 (2010): 323-8. ProQuest. Web. 21 Mar. 2016.
Clayton, Martin. “Ethnographic wax cylinders at the British Library National Sound Archive: A brief history and description of the collection.” British Journal of Ethnomusicology 5.1 (1996): 67-92. Web. 17 Mar. 2016.
Coates, Brendan. Personal interview. 1 Mar. 2016. Follow-up interview 15 Mar. 2016.
De Graaf, Tjeerd. “Endangered languages and the use of sound archives and fieldwork data for their documentation and revitalization: Voices from Tundra and Taiga.” IJAPS 7.1 (2011): 27-46. Web. 3 Mar. 2016.
Edison, Thomas A. “The Perfected Phonograph”. The North American Review 146.379 (1888): 641–650. Web. 10 Mar. 2016.
—. “The Phonograph and its Future.” Scientific American 5 (1878): 1973-1974. 10 Mar. 2016. 11 Mar. 1016.
Fewkes, J. Walter. “A Contribution to Passamaquoddy Folk-lore”. The Journal of American Folklore 3.11 (1890): 257–280. Web. 21 Mar. 2016.
Hoffmann, Frank W, and Howard Ferstler. Encyclopedia of Recorded Sound. New York, NY: Routledge, 2005. Print.
Iwai, Toshiaki, et al. “Reproduction of sound from old wax phonograph cylinders using the laser-beam reflection method.” Applied optics 25.5 (1986): 597-604. Web. 2 Mar. 2016.
Laing, Dave. “A Voice Without a Face: Popular Music and the Phonograph in the 1890s”. Popular Music 10.1 (1991): 1–9. Web. 2 Mar. 2016.
McLuhan, Marshall. Understanding Media: The Extensions of Man. 1964. Print.
Nakamura, Takashi, et al. “Optical Reproduction of Sounds from Old Phonographic Wax Cylinders.” Fifth International Topical Meeting on Education and Training in Optics. International Society for Optics and Photonics, 1997. Web. 2 Mar. 2016.
Penn, William, & Martha Hanson. “The Syracuse University Library Radius Project: Development of a non-destructive playback system for cylinder recordings.” First Monday [Online], 8.5 (2003). Web. 27 Feb. 2016.
Picker, John M. “The Victorian aura of the recorded voice.” New Literary History 32.3 (2001): 769-786. Web. 18 Mar. 2016.
Poole, Adrian. “Determining Playback Speeds of Early Ethnographic Wax Cylinder Recordings: Technical and Methodological Issues.” Ethnomusicology Forum. Vol. 24. No. 1. Routledge, 2015. Web. 2 Mar. 2016.
Porter, James. “Documentary Recordings in Ethnomusicology: Theoretical and Methodological Problems.” Association for Recorded Sound Collections—Journal 4.2 (1974). Web. 17 Mar. 2016.
Rehding, Alexander. “Wax cylinder revolutions.” The Musical Quarterly 88.1 (2005): 123-160.
Seubert, David. “Re: Cylinder archive questions.” Message to Crosby Buhl. 15. Mar. 2016. E-mail.
Stadler, Gustavus. “Never Heard Such a Thing Lynching and Phonographic Modernity.” Social Text 28.1 102 (2010): 87-105. Web. 20 Mar. 2016.
Thompson, Emily. “Machines, Music, and the Quest for Fidelity: Marketing the Edison Phonograph in America, 1877-1925”. The Musical Quarterly 79.1 (1995): 131–171. Web. 3 Mar. 2016.
Welch, Walter Leslie. Preservation and restoration of authenticity in sound recordings. Library trends, 1972. Web. 1 Mar. 2016.
Wilkinson, Alec. “A Voice from the Past: How a physicist resurrected the earliest recordings.” The New Yorker, 19 May 2014. Web. 28 Feb. 2016.