The invention of the cylinder-based phonograph (and of recorded-sound itself) is ascribed to Edison in 1877 and generally this is a fair attribution. Although the French physicist Charles Cros presented a similar idea to the Academie des Sciences in Paris the same year, Cros never made of his idea a practical reality. In turning ideas into practical inventions and then promoting and selling them, Edison excelled. It is to Edison we owe the word phonograph.
On this page we give: a brief introduction to cylinder records - far less familiar to many of us than disc records; advice on how best to deal with cylinder needle-drop recordings in Stereo Lab; and advice on preamplifiers and other equipment to obtain needle-drops of cylinder records.
Finally an appendix concerning the handling and care of cylinder records is included.
Oddly enough, Edison appears to have attached little importance to the phonograph at first. However, his competitive instincts led him to return to it ten years later, when Emile Berliner obtained patents for several improvements to sound-recording using a flat disc as the medium.
The cylinder phonograph only matured into a series of stable products in the late 1890s during which a range of spring-powered phonographs was introduced. It is curious that, despite the fact that the image of clockwork motor wound by a crank-handle is iconic of the phonograph, clockwork motors were not introduced until the late nineties. Until that time, all sorts of alternative propulsion systems were tried: a treadle machine; a machine powered by a water motor; and various electric machines driven by Grenet-cell batteries or by "electric light current".
All these other machines fell by the wayside. The public's imagination was captured - and commercial success gained - with the introduction of the first Spring Motor Phonograph (1896) which evolved into a machine called the Triumph. Later Edison phonograph machines were: the Home (late 1896); the Standard (1898); and the Gem (1899), the latter an economy machine which became very popular.
Types of cylinder record
There are three main types of Edison cylinder records: brown wax, moulded wax and celluloid.
Brown wax cylinders are made of a metal soap composed of stearic acid and aluminium powder. They originate from 1888 to 1902 and were directly recorded, or mechanically duplicated. These cylinder records are fragile and most are considered rare.
Moulded (Gold Moulded) and [black] Amberol cylinders originate in the period from 1902 to 1912. Once again, these are fabricated from a metal soap but with added lamp-black (carbon) which is the ingredient which makes them black like a disc record. These discs were mass-produced via moulding process. Gold Moulded records play for 2 minutes and Amberol records for 4 minutes. Moulded cylinders are as fragile as brown wax cylinders, but they are far more common because production volumes were so much greater.
Blue Amberol ("Everlasting" or "Indestructible") cylinders are made with a plaster (or sometimes cardboard) core with celluloid (in Blue Amberols, nitrocellulose) plastic playback surface layer. These records originate from 1912 to 1929. They were mass-produced via moulding process. Not subject to attack by mildew, like the "wax" records, Blue Amberols may suffer from absorbed moisture into the core material which causes the interior to swell and distort.
Other notable record types (by manufacturer) are:
Pathé and Lioret in Paris. The Pathé brothers produced very many standard cylinders and also bespoke 3½" x 4" Salon cylinders. Here is a link to a page which recounts the story of the Pathé brothers. More details of their cylinder discs is given below. For a short time between 1903 and 1905, Pathé made the largest ever cylinder record called the Céleste which was 5 inches in diameter (127mm) and 9 inches (230mm) long.
Henri Lioret was a clockmaker and inventor and made cylinder records originally for talking dolls. His cylinders were unique in their construction, comprising a brass tube with spokes covered by a molded celluloid sleeve carrying the grooves. Highly innovative, Lioret was the first to use a durable celluloid (early plastic, like the Blue Amberols) for cylinders and was the first to manufacture cylinder records by moulding. As early as the 1890s he was recording and manufacturing several sizes of musical cylinders to be played on his clockwork phonographs.
How it works
Just like a disc phonograph record, cylinder records are a physical memory storage device. Edison’s phonograph recorded the modulations as variable depressions in the cylindrical medium, a method called vertical-modulation or hill-and-dale modulation. Berliner chose lateral modulation for his discs in which the cutter - and subsequently the stylus - “wiggle” from side to side in sympathy with the sound. This choice of vertical-modulation was later to become a hindrance to Edison and ultimately led to the truimph of the disc medium over that of the original cylinder as we'll see.
The standard cylinder size was decided upon by Edison Records, Columbia Phonograph, and other companies in the late 1880s. The cylinders are about 4½ inches (115mm) long, 2¼ inches (57mm) in diameter. Hollow, with a tapered internal dimension, Edison cylinders fit closely onto the tapered machine mandrel* thereby preventing slipping. With the introduction of Edison' mass-produced Gold Moulded cylinders in 1902, record and playback speed was standardized at 160 revolutions per minute (RPM) of the cylinder. Up until that time various speeds were employed.
*Mandrel - a word borrowed directly from the metal worker's lathe where it refers to the shaft to which work is fixed while being turned.
With a standardised groove pitch of 100 lines per inch, these "standard" cylinders play for about 2 minutes. Later in 1908, better to compete with disc records which played for longer, the groove pitch on the cylinder records was increased to 200 lines per inch to play for around 4 minutes. These are the discs which are called Amberols. Some phonograph players allow the lead-screw to turn at different speeds to accomodate two-minute and four-minute records. These are called combination phonographs.
The principal working parts of the phonograph (apart from the motor) are illustrated right¹. Of these be sure to identify: the lead-screw (6, red), tapered mandrel (1, yellow) which runs on the main-shaft (6, red), reproducer carriage (14, green), the reproducer (15, blue), back rod (4 brown) and its sleeve (34, light-brown), half nut(s) (28) and spring (3 light-blue), lift lever (18, purple); and swing arm (11, orange). This last part temporarily swings away from supporting the main-shaft so as to provide access to the mandrel when loading and un-loading cylinders.
In the Edison machine, the weight of the moving carriage (analogous with the tonearm in a disc player) is not supported by the medium, instead it is supported by the straight-edge bearing (13, cyan) upon which it runs.
In operation, the Edison machine more closely resembles the lathe than a disc player. The tonearm (or, in this case, the reproducer carriage 14) is not dragged across the record via its engagement with the groove as in a disc phonograph: during recording and playback, the reproducer (15, blue) is piloted across the disc via the carriage driven by the half-nut (28) riding in the threaded lead-screw (6, red).
Edison originally saw the application of the phonograph as a machine to aid business: as a means of leaving messages, taking dictation and so forth. A crucial part of Edison's propositon was that the wax cylinder6 phonograph could record as well as reproduce; a great advantage (in Edison's mind anyway) over the disc record which only allowed for reproduction. Moreover, the relatively soft wax-cylinder medium could be "shaved" of its modulation and erased for multiple use and all early machines incorporated a knife blade specifically to perform this erase function.
In use
For reproduction use (which, of course, became the principal application of the phonograph), the main-shaft is set in motion by removing some kind of brake upon the clockwork motor which is coupled to the main-shaft via a belt. As the mandrel turns, the lead-screw is set in motion by a set of reduction gears driven from the main-shaft.
The reproducer carriage is held clear of the straight-edge by a pin or lever when not playing. The pin (or lever) is withdrawn to start playing and the carriage lowered at the left-hand end of the record until the carriage mass rides on the straight-edge bearing and the half-nut engages with the lead-screw.
The reproducer stylus (a glass or gem-stone ball) engages with the shallow grooves of the cylinder record. The playing downforce in the cylinder player is provided by the hinged lower plate of the reproducer (and its fishtail weight, indicated D in the illustration right²). The reproducer arm (E) is hinged (G) and a link (P) communicates the vertical motion of the stylus to a sensitive diaphragm made of mica or ribbed aluminium or copper (N), clamped at its outside edge between two soft rubber gaskets (M and O). The vibrations of this diaphragm are amplified by the horn which is normally fitted to this type of phonograph.
The demise of the cylinder
Edison's clockwork cylinder phonograph machines were rugged and reliable, and millions were sold. The year of 1903 was the peak for Edison's cylinders. After that, disc machines began to outsell cylinder machines. Edison's choice of vertical, gentle modulation of shallow grooves ensured that cylinder records didn't sound as loud as the flat discs made by Victor and Columbia with their deep, widely modulated grooves. Disc records also played up to three minutes of music on both sides which outstipped Edison's brief, two-minute cylinder records.
Never a man to resist a challenge, Edison fought back. An interim solution involved increasing the size of the cylinder so as to be able to increase the recorded velocity - and thereby the loudness - of the cylinder record. These cylinders were known as Concert records and were intended for public performances, but a special machine was required to play them (the Edison Concert Phonograph) so this was not a great success. And, as explained above, in 1908, new Amberol cylinders with a 200 lines-per-inch pitch were introduced to counter the duration criticism.
Even Edison's four minute celluloid records in a brilliant blue color, the "Everlasting" and "Indestructible" Blue Amberols were not introduced in time, and by 1912 the format war was well and truly lost to the phonograph disc. The last Edison cylinder phonograph known as the Amberola was manufactured in 1919.
The Legacy
Edison preferred his cylinder phonograph invention to the day he died (in 1931) over Berliner's discs. And, in many ways, the Edison system was ahead of its time. The use of a gem-stone stylus riding on a relatively soft, smooth medium anticipated the LP by many years. The surface noise (except due to damage) is substantially below that of Berliner's discs because the wax material did not require to be filled with powdered slate to support the weight of the tonearm. (In this, the criticism of low output is somewhat unfair because the noise of the medium is much lower too³.) And the cylinder has the great advantage that the angular speed is constant at all parts of the recording and doesn't decay away as the record turns, as it does on a disc.
But the real distinction Edison's machines had over the disc phonograph was the ability to record. And, in the case of the smaller clockwork machines, the ability to record in inaccessible places. In this, the clockwork phonograph anticipated the portable tape-recorder by 60 years and was the go-to technology for anthropologists and ethnomusicologists to record traditional ceremonies, songs and stories way into the 20th century.
It is in this last rôle that the cylinder phonograph left its greatest legacy. Even to this day, there are priceless collections of recordings made on cylinders of lost traditions, worlds and peoples.
Frances Theresa Densmore, American anthropologist and ethnographer, is known for her studies of Native American music and culture. Densmore spent over fifty years studying and preserving American Indian music and collected thousands of recordings on wax cylinders, many of which are now held in the American Library of Congress. This is a short video about her work.
Cylinders and Stereo Lab
Specialist Players
The first complication when archiving cylinder recordings is a machine on which to play them. Historical phonographs (in which there remains healthy interest) are not ideal as their old, heavy reproducers can wear or damage precious recordings. In addition, recording an acoustic player with a microphone at the horn mouth is unlikely to get the best results.
Several commercial instruments exist of which the most famous is the Archéophone due to Henri Chamoux of France. Ingeniously designed to play all formats of wax or celluloid cylinders, produced between 1888 and 1929, and even later, the Archéophone, allows the transcription of the cylinders using a lightweight tonearm and modern, electrodynamic phono cartridge.
Traction to the Archéophone's polypropylene mandrel is provided by a belt driven by an asynchronous electric motor (which is accurately calibrated in RPM) and the machine incorporates a linear tracking servo controlled tonearm. The Archéophone is used by the largest archives which preserve such media in the world including the American Library of Congress, Bibliothèque Nationale de France, the Edison National Historic Site, University College Dublin and the Canadian Museum of Civilization and many other libraries and archives. (You can briefly see the Archéophone in action in the Library of Congress in the Frances Densmore video above.) Here is a link to more technical data on the Archéophone. And here is a link to a very good (French language) video about the Archéophone.
French connection
The French connection with the cylinder phonograph goes back to the very earliest days of the format and its adoption by two brothers, Emile and Charles Pathé (Les Frères Pathé) who owned a successful bistro in Rue Fontaine in Paris where they entertained their customers with an early, imported Edison cylinder phonograph; an innovation which proved so mesmerising that many of their clients asked if they could buy one.
Unable to obtain the rights to manufacture under licence from Edison, the enterprising brothers simply copied the American machine and manufactured it at a local engineering firm in Belleville, in the outskirts of Paris. The enterprise was so successful that, by the end of the century, their factory in Chatou, west of Paris, employed over two hundred workers to meet customer demand for cylinders and machines. From this one factory, the Pathé brothers are said to have produced between 30 and 45 million cylinders between 1900 and 1910. This was the start of the media giant Pathé. Here is a link with more information about Pathé discs and how to deal with them in Stereo Lab.
An excerpt of an example of a transfer of an Edison cylinder made in 1904 using Chamoux's Archéophone is given here. The song is Je suis républicain (I am a Republican) written by Henri Christiné and sung by Victor Lejal. This recording is a tiny part of the vast Phonobase on-line database project which gathers sound excerpts and photos taken from early European commercial cylinders and records made from 1888 to approximately 1920. Even this short excerpt demonstrates both the Archéophone and its ability to coax the best sound quality available from Edison's system. The result is certainly impressive.
However, weighing almost 25kg and costing over US $10,000, Chamoux's Archéophone is clearly a specialist's tool and his machines are built to order. An alternative product - but equally one aimed at institutional collections - is the Cylinder and Dictabelt Machine made by Endpoint Audio Labs of Burbank, California.
BURP!
But what if you are not the Library of Congress?
Happily for the more modest collectors who might nevertheless wish to archive a small collection, or simply experiment with cylinder recordings, much more modest alternatives have been constructed from non-specialist materials and parts.
One such is the BURP-ONE (Basic Utility Record Player) which was described by its inventor James Nichols in an AES Conference paper4. Almost militantly "low-tech", James describes a practical wax cylinder transcription system using readily available parts which affords the modest archivist a pragmatic means for preserving the heritage of antique cylinder recordings at low cost.
Anyone wishing to experiment with a cylinder player for electrical recording should also read a truly excellent article5 by Hans Meulengracht-Madsen of the Physics and Engineering Laboratory in New Zealand who developed a machine for the purpose of transcribing old cylinder phonograph records from 1919-1923 of a series of Maori musical events, starting with the Hui Aroha welcoming the Maori battalion on its return from World War I.
Although this article is over 40 years old, it contains unique information, gained from specially recorded cylinder test records, of the transfer-function of cylinder-records, their surface noise-spectrum and their dynamic range and linearity.
Meulengracht-Madsen's machine is illustrated left.
Into the Lab'
Stereo Lab has several features, originally developed for shellac discs which work well with cylinder recordings too. Crackle Removal is very useful when processing cylinder records as is Graδient noise-reduction. Hill & Dale should be selected in the Phonograph-regime parameters (in the Phono EQ tab of the Preferences), this ensures that all lateral modulation is ignored and only the vertical-modulation is passed forward for further processing. (Balance and azimuth should be checked too, as this improves the accuracy of the of the Hill & Dale matrixing).
In addition, specialist techniques may be used:
By employing half-speed needle-drops, discs may be replayed at half-speed (the Archéophone allows this to be precisely selected) and speed-doubled in software. This results in much more accurate tracking with much less tendency to jump - something to which cylinder discs are somewhat prone; especially when damaged.
As an example of how Stereo Lab may be used to process really early and (otherwise largely unlistenable) needle-drops of historical cylinder records, we have processed an early recording in the Library of Congress collection. It is the Passamaquoddy Tribe's song of the Snake Dance sung by Noel Joseph, recorded by Jesse Walter Fewkes in 1890. The original needle-drop (from the Archéophone) is first, then the processed file7. Crackle reduction and Graδient noise-reduction were employed. Also Pavillon (see below).
Pavillon - coming soon!
Pavillon is a pre-process in Stereo Lab intended to correct for the limitations of the acoustic recordings; especially in the irregular frequency response due to the recording horn (pavillon is the French word for an acoustic horn). Pavillon also reduces mechanically induced noise in the mechanism of the recorder because these artefacts are often at frequencies below the LF cut-off of the acoustic horn.
Pavillon should be selected in the phono preferences dialogues. It has been tested and is compatible with Edison, Pathé and Lioret cylinders.
Here's a brief demo of the Pavillon process on a recording on a standard Pathé cylinder (Pathé 3440) of the Berceuse de Jocelyn by Benjamin Godard and sung by Jeanne Marignan. The original needle-drop (taken from the Phonobase database) is first, followed by the version treated with Crackle Removal, Graδient and Pavillon.
Pspatial Audio worked with Phædrus Audio to develop a new Groove Sleuth preamplifier specifically designed for making needle-drops of cylinder records. (Most designers seem to make both signal channels available from a modern pickup when playing vertically modulated cylinders - certainly the Archéophone does. Thus a "stereo" preamplifier is still required.)
The lower modulation velocities encountered with cylinder records demand a higher gain preamplifier for these duties, compared with a preamplifier for disc records. The cylinder version of the Groove Sleuth preamplifier is modified to provide the extra gain required whilst not sacrificing noise level or headroom. The Groove Sleuth preamplifier is available from Phædrus Audio. Address sales enquiries to sales@phaedrus-audio.com.
The standard Edison cylinder bore is a truncated cone with a taper of 1:32. The diameter at the larger end of the bore is around 47mm and the smaller around 43mm.
Cylinder speeds
With the introduction of mass-produced Gold Moulded cylinders in 1902, record and playback speed was standardized at 160 revolutions per minute (RPM). Before this a wide range of speeds were used. Home and office cylinder recording machines spun at nearer to 100 RPM and this is true of many of the field-recordings made on cylinders. It is widely held opinion that the ear is a discerning instrument and general consensus that cylinder phonograph speeds should be set so that the performance - and especially the voice - "sounds right".
Groove dimensions, modulation depth and maximum recorded velocities
For standard "two-minute" cylinders the groove is cut with a tool made from a cylindrical 0.9mm diameter sapphire rod. This defines the maximum depth of cut of 0.000721" (18µm); any deeper and the width of the grooves would be wide enough to over-cut on a subsequent spiral.
Given that the minimum grooove depth is zero and the maximum is determined by the over-cut limit depth given above, the maximum peak-to-peak modulation (in terms of stylus displacement) must be equal the maximum depth of cut, 18µm.
This therefore is the peak velocity limit of a cylinder record. Expressed in the normal cm. units, this is 3.72cm/second which is equivalent to 2.62cm/s RMS. Note that this is verymuch lower than the velocity limit of a disc record. Interestingly, in a (100 LPI) cylinder record, the modulation limit is set by cutter slew-rate considerations, rather than by modulation depth considerations for all frequencies above about 300 - 400Hz.
When Edison resumed serious work on the phonograph in 1887, he investigated various animal, vegetable and mineral waxes, experimenting with cylinders consisting principally of white wax, containing a small percentage of commercial carnauba wax. But this material proved too soft to be practical. Edison and his chemists needed a material soft enough to take a faithful recording, yet stable and durable enough to permit repeated playback.
They found that in a class of compounds known as metallic soaps which afford optimal properties for such a recordable and shavable cylinder blank. These salts of a fatty acid combined with one or more metals yield amorphous materials with a waxy feel, insoluble in cold water. Though chemically quite different from the true waxes occurring in nature, the metallic soaps used in the cylinder record industry continued to be referred to as "wax" because of their wax-like feel and mouldability. (For a comprehensive analysis of the make up of Edison's cylinders see reference 6). See also Appendix 2 - Cylinder Care
The influence of the recording horn upon an acoustic recording was first investigated by G.W. Stewart14. It may be understood in terms of transfer-function of the sound collected at the throat of the horn when presented with frequency sweeps at the mouth. The response shown was obtained in this way from a replica of an early recording horn8. The peaks in the frequency response are due to the axial resonances of the conical horn. Very sturdy, metal-plate conical horns were preferred for recording well after exponential horns had taken over from conical horns in reproducers because the recording engineers preferred the directional characteristics of the conical horn. In the case of Edison's studios, the standard recording horn was 16 inches long without a flare. It was conical with a (relatively narrow) 6-inch mouth tapering to a 5/8" inch end to which a rubber coupling for the recorder was attached.
The resonances were not due to resonating parts of the structure of the horn; the heavy construction saw to that. They were due to the purely acoustic resonances as seen in open-ended pipes. In fact, the first few resonances follow closely what one would expect of sound in a cylindrical pipe. Due to the taper of the horn, these resonances spread out and become generalised above the first few harmonic frequencies. It is Stewart14 who first understood that this generalised resonance is the source of the "gain" of the horn.
It must be said however, that this very "peaky" frequency-response was not passed-on to the record unchanged. The interaction of the horn with the coupling hose and the the recording diaphragm tended to "smooth" the end-to-end response. There is evidence too1, 9 that Edison's recording staff deliberately worked by selecting and combining of horns and coupling tubes and the recording diaphragms so as to place the horn resonances away from the the resonances in the the recording diaphragm; thereby flattening the combined transfer-function.
Horns were also selected so as to ensure the formants present in the singer's voices or instruments were not artificially amplified by the horn, rather as a modern engineer will choose a particular microphone to suit a particular singer. This "creative" part of the recording process sets a lower limit of what is possible in correcting for horn response when post-processing these recordings. If only the recitation which starts so many cylinder recordings were something more like,
Mr Richard Brice sings Music Hall Melodies preceded by a sine x upon x pulse. Edison Records..
we could do so much more!
Appendix 2 - Cylinder Record Care
The UNESCO (UNISIST) document concerning the practical management and curation of Audiovisual archives13 has this to say about cylinder records as part of a collection,
All cylinders are now very fragile. The polymers, both coating and monolithic, and
the core materials are decaying and irreversibly breaking down. Urgently seek advice
if you have any cylinders in your collection.
Unhelpfully, as to whom one would turn for this advice isn't specified! The following is an attempt to collate various sources of advice which we have found useful in our experiments with cylinder records.
Handling cylinder records
Despite being conceived as a medium for the general public to use, early brown wax and Gold Moulded cylinders are fragile and can easily be broken or damaged through improper handling. Obviously, they will smash if dropped, but more surprisingly, if a record is stood on its end and topples over onto its side, it will often crack. So, it pays to be careful and methodical when playing cylinder records.
Do not handle the cylinder by the outside surface. The wax surface of the record is even softer than the plastic of an LP disc, so any scrape with a fingernail will leave a nasty gouge and grease from the skin leaves food for fungal growth to develop. Pick up a cylinder by gently expanding two fingers of one hand inside the cylinder from the title (or thicker) end as illustrated. Always stand cylinders on their ends, never on their sides as this will tend to make them eccentric and damage the recorded surface. At the least, replace cylinder records in their boxes as soon as possible (but see more below on storage of cylinders).
Don't push the record onto the mandrel with too much force: they are brittle and can split. And allow cylinders and mandrel to reach room temperature before using them. If a warm cylinder is pushed snug onto a cold mandrel, it will contract and split. Never leave a cylinder on the mandrel when you have finished with it, such an oversight may result in a split cylinder due to shrinkage as the room cools at night.
Wax cylinder deterioration and preservation
For a truly thorough study of disc record care, we have a report by a team working for the US Library of Congress (LoC) under a research grant from the Rockefeller Foundation11. Although this report was written the end of the nineteen-fifties, it is late enough that PVC-PVA copolymer disc records (LPs and singles) were included in the study as well as acetate and shellac disc records. It remains the standard work from which all more modern advice12 is derived. But, sadly, the report contained no specific work on cylinder records. This is an odd omission from such a thorough piece of work. Perhaps the LoC had not appreciated its cylinder record treasures12 at that time.
Nevertheless, many of the findings relating to disc records are applicable to wax cylinders. In addition to breakage, the greatest factor in wax records deterioration is from fungal growth. Unfortunately, the composition of wax cylinder records make an ideal material for mould spores to grow and flourish. Mildew on wax cylinders makes its appearance as white-grey spots surrounded by filamentary strands. This mould feeds on cylinder material and destroys the undulations or recorded matter on the records. The result is that fungal growth causes a noisy hiss on the records when reproduced. This is covered on the page devoted to the LoC report, since disc records are similarly affected. Advice on the environmental conditions for storage of disc records is applicable to cylinder records.
Packaging and storage
There does not seem to be a standard reference for how to store cylinder records, however very good, and sensible, practical advice is available from the website of the City of London Phonograph and Gramophone Society (CLPGS). The CLPGS is the oldest such society in the world, so they have a wealth of experience to call on. Introducing their advice they say hearteningly,
..... cylinder [records] appreciate the same sort of conditions we humans like to live in, namely; stable, without extremes of heat or cold, and neither too damp, nor too dry. Extremes of temperature can bring about chemical degradation and physical stress resulting in rough cloudy surfaces and fissures, which will eventually cause the record to split. Extremes of humidity promote the growth of mildews, which eat into the record surface, leaving it noisy or even unplayable, and condensation, which can degrade the record chemically. Wax records suffer these reactions more readily than celluloid, but eventually, celluloid records will degrade too, not to mention shrink and split. Encouraged? Read on, it need not happen in your lifetime!
Their general advice is to ensure that, if cylinders are kept in their original boxes, the boxes should be cleaned of dust and grit, ensuring especially that the cotton liner is absolutely clean and dry or it will encourage mildew. They also caution against storing cylinders on their side (even once in their boxes) as this too encourages damage to the playing surface where the weight of the record is upon the cotton lining. Summing up, they say,
..... all things equal, cylinders stored in peg boxes or cases where nothing touches the surface and dust and air cannot circulate, seem to have fared the best.
1. A Complete Manual of the Edison Phonograph. Tewksbury, G.E., Ringer, New York, 1897
2. Taken from Edison maintenance information.
3. If Edison's cylinder format had bridged the years into the era of electric recording and reproduction, the lower volume of the cylinder would have been irrelevant.
4. A High-performance, Low-cost Wax Cylinder Transcription System. Nichols, J. 20th International Conference of the AES Archiving, Restoration, and New Methods of Recording (October 2001). Permalink: http://www.aes.org/e-lib/browse.cfm?elib=10052
5. On the Transcription of Old Phonograph Wax Records, Meulengracht-Madsen, Hans. JAES Volume 24 Issue 1 pp. 27-32; February 1976
6. Cylinder Record Materials. Wile., R.R. ARSC Journal XXV1 I ii 1996.
7. Recorded in Calais, Maine on March 15, 1890 by Jesse Walter Fewkes. Passamaquoddy Tribe's Snake Dance song. Digital preservation copies, MAVIS nos. 2031764-2-1 and 2031764-4-1, from original cylinder on Archeophone #27. Library of Congress, 2016 March 28 and 2016 May 18. 89.05 and 88.89 Mbytes BWF. American Folklife Center, Library of Congress, Washington, D.C.
8. Computational Phonogram Archiving Editor: Bader, R. Springer 2019. Chapter 16: How to Interprete Early Recordings? Artefacts and Resonances in Recording and Reproduction of Singing Voices Kob, Malte (et al.)
9. Electrical Reproduction of Acoustically Recorded Cylinders and Disks, Fesler J.C., J. Audio Eng. Soc., Vol. 31, No. 9, 1983 September also Presented at the 70th Convention 1981 October 30-November 2 New York
10. Electrical Reproduction of Acoustically Recorded Cylinders and Disks, Owen, T., Fesler J.C., Presented at the 70th Convention 1981 October 30-November 2 New York
11. Preservation and Storage of Sound Recordings. Pickett, A.G.; Lemcoe, M.M. (US) Library of Congress Report 1959
12. ARSC Guide to Audio Preservation. Brylawski, S. et al. National Recording Preservation Board of the Library of Congress
13. St. Laurent, Gilles. 1997. "The Care of Grooved Recordings." Pp. 250–258 in Audiovisual Archives: A Practical Reader, edited by Helen P. Harrison. Paris: UNESCO. Available at http://unesdoc.unesco.org/images/0010/001096/109612eo.pdf.
14. The Performance of Conical Horns. Stewart, G. W. Phys. Rev. 16, 313 – Published 1 October 1920. Published by The American Physical Society. url = {https://link.aps.org/doi/10.1103/PhysRev.16.313}
The invention of the cylinder-based phonograph (and of recorded-sound itself) is ascribed to Edison in 1877 and generally this is a fair attribution. Although the French physicist Charles Cros presented a similar idea to the Academie des Sciences in Paris the same year, Cros never made of his idea a practical reality. In turning ideas into practical inventions and then promoting and selling them, Edison excelled. It is to Edison we owe the word phonograph.
All these other machines fell by the wayside. The public's imagination was captured - and commercial success gained - with the introduction of the first Spring Motor Phonograph (1896) which evolved into a machine called the Triumph. Later Edison phonograph machines were: the Home (late 1896); the Standard (1898); and the Gem (1899), the latter an economy machine which became very popular.
Blue Amberol ("Everlasting" or "Indestructible") cylinders are made with a plaster (or sometimes cardboard) core with celluloid (in Blue Amberols, nitrocellulose) plastic playback surface layer. These records originate from 1912 to 1929. They were mass-produced via moulding process. Not subject to attack by mildew, like the "wax" records, Blue Amberols may suffer from absorbed moisture into the core material which causes the interior to swell and distort.
Just like a disc phonograph record, cylinder records are a physical memory storage device. Edison’s phonograph recorded the modulations as variable depressions in the cylindrical medium, a method called vertical-modulation or hill-and-dale modulation. Berliner chose lateral modulation for his discs in which the cutter - and subsequently the stylus - “wiggle” from side to side in sympathy with the sound. This choice of vertical-modulation was later to become a hindrance to Edison and ultimately led to the truimph of the disc medium over that of the original cylinder as we'll see.
The principal working parts of the phonograph (apart from the motor) are illustrated right¹. Of these be sure to identify: the lead-screw (6, red), tapered mandrel (1, yellow) which runs on the main-shaft (6, red), reproducer carriage (14, green), the reproducer (15, blue), back rod (4 brown) and its sleeve (34, light-brown), half nut(s) (28) and spring (3 light-blue), lift lever (18, purple); and swing arm (11, orange). This last part temporarily swings away from supporting the main-shaft so as to provide access to the mandrel when loading and un-loading cylinders.
The French connection with the cylinder phonograph goes back to the very earliest days of the format and its adoption by two brothers, Emile and Charles Pathé (Les Frères Pathé) who owned a successful bistro in Rue Fontaine in Paris where they entertained their customers with an early, imported Edison cylinder phonograph; an innovation which proved so mesmerising that many of their clients asked if they could buy one.
Stereo Lab has several features, originally developed for shellac discs which work well with cylinder recordings too. 
Pspatial Audio worked with Phædrus Audio to develop a new 
For standard "two-minute" cylinders the groove is cut with a tool made from a cylindrical 0.9mm diameter sapphire rod. This defines the maximum depth of cut of 0.000721" (18µm); any deeper and the width of the grooves would be wide enough to over-cut on a subsequent spiral.
The influence of the recording horn upon an acoustic recording was first investigated by G.W. Stewart14. It may be understood in terms of transfer-function of the sound collected at the throat of the horn when presented with frequency sweeps at the mouth. The response shown was obtained in this way from a replica of an early recording horn8. The peaks in the frequency response are due to the axial resonances of the conical horn. Very sturdy, metal-plate conical horns were preferred for recording well after exponential horns had taken over from conical horns in reproducers because the recording engineers preferred the directional characteristics of the conical horn. In the case of Edison's studios, the standard recording horn was 16 inches long without a flare. It was conical with a (relatively narrow) 6-inch mouth tapering to a 5/8" inch end to which a rubber coupling for the recorder was attached.
Despite being conceived as a medium for the general public to use, early brown wax and Gold Moulded cylinders are fragile and can easily be broken or damaged through improper handling. Obviously, they will smash if dropped, but more surprisingly, if a record is stood on its end and topples over onto its side, it will often crack. So, it pays to be careful and methodical when playing cylinder records.
Apple Certified Developer. Stereo Lab, Aria 51, Aria 20, Head Space, Groove Sleuth, iLOOP and FRANCINSTIEN T-Sym are trademarks of Pspatial Audio. FRANCINSTIEN and Bride of FRANCINSTIEN (BoF) are trademarks of Phaedrus Audio. Macintosh and the Mac logo are trademarks of Apple Computer, Inc.