Abstract:

Today we are left with physical records (disc records or cylinders) that contain signals in some manner representative of an event that took place in front of a recording horn any time between 1885 and 1925. Anecdotal accounts exist of how recording took place, but very little factual information has been available that demonstrates how the result: a listenable (on contemporary reproducers) and possibly commercially valid product was actually obtained in such alien circumstances. How did the process of recording influence the result? For instance, from the beginning re-scoring of works was performed in order to obtain a desired musical balance.

The paper will look at some of the steps in the process of developing record production as a concept and scratch the surface of material related to commercial enterprises like the Victor Talking Machine Company (USA) and the Gramophone Company (UK). The paper will trace some of the considerations regarding improvements and standardization of work practices as well as the various attempts at quality control. This very early type of material has not been considered to any noticeable degree before. (1)

Introduction.

With the risk of being less precise than music sociologists would accept, I think it reasonable to think of music consumption in a European country since 1870 as being socially stratified, with some overlapping of repertoire. The common man would know liturgical music (psalms), popular songs, dance music (weddings, harvest), and military music (marching and performing bands). The "classical" repertoire would be represented by catchy tunes and arrangements or adaptations for brass bands. The bourgoisie and lesser nobility would know a fair amount of the classical repertoire via the piano literature and going to orchestral and operatic performances. The popular music of the day would be presented in revues and restaurants. The upper nobility and industrialists would travel to foreign opera houses to experience famous performers. And more likely than not they would invite performers to their homes and pay orchestras to perform for social events. Non-professional performance was at one end represented by colliery bands and local choirs and at the other by private chamber music performance.

Commercial recording.

Commercial records did become available from ca. 1895 and eventually would provide a fairly wide range of repertoire. There is no doubt that it was desired to give listeners an experience that in certain - hopefully most - respects was a valid representation of a live performance. Recording was emerging, it had never been done before, and very many skills had to be learnt from afresh. There was no experience that could be learnt from. Later generations had far less problems and could enjoy the full advantages of simpler methods of recording and actually had a simpler task - until they deliberately made the tasks more technically difficult for commercial reasons.

Most recordings were created to represent some band music, a humourous dialogue, a little song, some virtuoso classical music performance. Obviously there were freak recordings or recordings making use of the technology of recording as such as part of the entertainment - a puzzle record was for instance available before 1900. The instruments used were those that were useful in the recording situation. Even for classical music, it was not really very important which instruments were used to carry the melody and provide the harmony - arranging and adapting was already very widespread, in particular for brass band or the piano (or four hands).

In other words, it was the melody, harmony, rhythm and tempo that was important, and certainly it was possible to convey expression and interpretation by means of these elements.

The famous performer on record was the exception, but the famous performer validated the concept of recording as such, the proof that recordings were able to represent artistic expression. Certainly the testimonials issued by famous performers indicate their satisfaction with the end results. However, they bear no witness to the birthpains of the recordings. To some performers recording was just another venue for performing; to others it was a a painful divergence.

Recordings were taken seriously by some composers, for instance, we know that Puccini was eagerly waiting for a selection of the Japanese records recorded by Fred Gaisberg in 1903 before he wrote his music to Madama Butterfly. (2)  

There is not much reliable literature on the manner in which recording took place in the acoustic period, the period that is commonly regarded as displaying the most artificial recording situation. However, it is also the period in which the authenticity of the record as representing the sound in the recording studio (or rather in the recording horn) is simplest to prove.

We know about acoustic recording mainly from modern books like Edison, Musicians, and the Phonograph by Harwith & Harwith (1987), Joe Batten's Book (1956), or the Gramophone Jubilee Book (1973). The first two of these represent experiences that were from 30 to 60 years old at the time of their collection. Some information may be had from reminiscences in the journal the Gramophone, but it has to be interpreted correctly. (3)(4)

Ernest Stevens who was with Edison recounts the systematic work that Edison did to obtain the best recordings. He tells us that Edison marked squares on the studio floor and tried each instrument on each square and them made a whole ensemble play according to the best placements, and correcting if the interaction were detrimental to the sound. The studio was totally deadened, and he also experimented with the influence of the velocity of sound in dependence of the temperature. Many recordings were made inside the horn, so to speak, in that the mouth of the horn was one wall of the studio. Later, and independently, Carl Reich's recordings of canaries 1910 and later also placed them inside the recording horn - actually to scale!

Anna Case and Rosa Ponselle describe how dynamics is controlled, actually compressed, by the soloist's distance from the horn, and Samuel Gardner who played the violin felt insecure because he had to be so close that he might strike the horn. Irving Kaufman describes how a soloist was made to stand out by letting him move in closer into a space created by the other players.

An important source would seem to be literature contemporary with the technology. However, here we come up against a problem: whereas e.g. the electrotechnical or the chemical fields were well-documented by contemporary literature and patents, commercial sound recording very much relied on trade secrets. In practice this means that we may find literature that addresses the problems of home or amateur recording on the only significant medium, the cylinder, but virtually nothing regarding commercial recording that intended to manufacture disc records with a suitable durability and quality on the gramophones available to customers.

But even the "standard" contemporary literature is very hard to find, because the whole industry was at the fringe of everyday life. Talking Machines & Records by S.R. Bottone (1904) was not written by a specialist but by a technical journalist who also had dealt with many other technical subjects. Gramophones and Phonographs by B. Clements-Henry (1913) came in several editions, but again it was not written by a specialist - these two books demonstrate some fundamental technical misunderstandings. Die Modernen Sprechmaschinen by Alfred Parzer-Mühlbacher (1902) was another book by a generalist, but it has sensible instructions for recording. The Art and Science of the Gramophone by Harry A. Gaydon (2nd Ed. 1928) was written by a consultant in the field, but as opposed to the first two, there is no mention of recording situations. The standard work in the field The Reproduction of Sound by Henry Seymour (1918) was written by yet another consultant and this book is very satisfactory in many respects. (4)

Although these books are all exceedingly rare, they are in the public domain, and so merit less focus than what may be gleaned from contemporary business papers in the gramophone business and verified by modern experiment, and these are the sources that we shall now address.

Two players in the field

The Gramophone Co. (GC) in the U.K. and the Victor Talking Machine Co. (VTMC) in the U.S. were massive players in the field, tracing their roots back to ca. 1895, they were organized very differently, and the owner of VTMC, Eldridge R. Johnson also had a substatial share of the GC at various times. They developed very differently from ca. 1930. GC merged with Columbia Graphophone Co. in 1931 to form what was later to be named EMI. The GC (and EMI) were always very conscious of the back catalogue, and vast quantities of masters as well as the business documentation have been preserved to our days. Columbia material has not been preserved, partly due to a fire before the merger. (5)

The two "old" cooperating companies manufactured both gramophones and records. We know the same type of "complete line" from Edison, but he was the sole supplier of both his discs (1912-27) and the proper reproducer and was able to optimize a system, i.e. the complete combination of record and reproducing mechanism. His approach was that of the industrial scientist, and in some respects concerning actual recording and repertoire, quite idiosynchratic. For this reason the vast Edison material from this period that has survived only permits us to draw fairly narrow conclusions. (6)

We must not forget, however, that other, but somewhat smaller, companies were active in the early years, such as the International Talking Machine Company (notably Odeon) and Pathé.

A matrix exchange agreement was in place beetween the GC and the VTMC, which meant that a recording made under one company's auspices would most likely be available in another part of the World with the other company's labels. However, in order to be able to sell a high quality product irrespective of the source, there had to be certain standards of quality, and we know from prior research (7) that the philosophies of the two companies were very different. VTMC originally had a couple of recording locations on the east coast of the US, whereas GC had recording locations in a number of large cities of Europe. In the US only a half dozen experts were doing the actual recording, in Europe there were three times as many. The VTMC philosophy was mainly durability, i.e. customer satisfaction for as long as possible after the purchase. This required a reduced dynamic range for the recording. The GC philosophy was to re-create the excitement of the rendering, and they generally had a higher dynamic range.

Contemporary considerations

I shall quote at some length from two letters constituting internal reports of the GC.

The first is a report (8) on a trip abroad:
"..... During my visit to Paris to see X [author's anonymization] I also took along with me a variety of Sound Boxes so that we could test which were the most suitable fro [sic] French Band Records. We succeeded in picking out two that gave splendid results and X is now using these for his Band Recording; the tests to my mind both as to firmness of Sound Waves and loudness of tone are the same results which we are abtaining [sic] in London, and X seemed very pleased.
     Heard a number of X's records when in Paris, and if loudness is the main object in the French Recording, I think X's Vocal records hit the mark.
      From an artistic standpoint there is tremendous room for improvement, I am surprised that X gets the results that he does without any artistic assistance, of course I hav'nt [sic] the slightetst idea what arrangemeents the French Company have with their Artistes, but certainly there is needed in Paris a better division between the ordinary work and the Operatic. From X's explanation it seems his main object is to get as many records in one Session as possible, this is alright up to a certain point, but in recording Operatic numbers speed is impossible. I understand that X is in the habit of mixing up Comic Singers and Operatic Singers and recording them in the same Session, using the same Conductor for the work; this of course might be possible to do, but not under X's supervision, and my experience has been that if there is no one present during the Recording Session who knows whether a song or Selection is being rendered properly both as to the singing and the proper Orchestral effects the Company are [sic] liable to issue records that are musically incorrect. In London here I find it to my advantage so surround myself with people who are familiar with that style of Record which I am recording at the time."
...........
"It also struck me in the Band Session that I heard, the Instruments were anything but in tune, why this should be I cannot understand - a Band Conductor allowing a Band to make records when the Band is entirely out of tune should be changed."   ............

The difference between GC and VTMC was discussed in a letter (9) that demonstrates a great will to get to bottom of the problems percieved:

    "I am going to try to embody my views on this subject in this letter and to compare the English recording with the Victor recording, from a commercial standpoint.
    First of all to make a physical comparison - the recording surface of a Victor record is more pleasing to the eaye. It looks finer and when there is a loud note the vibrations show less on a Victor record than on ours.
    Scratch. Although our records are much quieter than they were, and especially the Clara Butt's, still I do not think that our quietest are as quiet as the Victor Company's. I suggest here that the fault on our side does not seem to me to be entirely with the recording material, because it occurs that the original first samples are often a good deal quieter than the finished record. Ana again, some recods from the same master, pressed from various matrices, vary in scratchiness.
    Wear. I consider under this heading the Victor are unquestionably ahead of us. Some of our Celebrity records are very bad from a wearing point of view."
.................
    "To turn now to the more artistic side I will deal first with:-
    Tone. This I will subdivide under headings of Vocal Music and Orchestral Music.
    Vocal.  I consider, speaking generally, that the Victor in their recording tighten up the tone of the voice. To give a simile, i will compare it to a violin string being tightened up without altering the pitch, if tht were possible - only altering the tone. It would be slightly harder. Similarly, the Victor produce that effect in recording their Artistes. This has a beneficial effect in some cases, but the reverse in others. Take a tenor voice - in some cases this tightening-up effect is to the advantage of the tenor and roduces a record with even a better tone than the original singer's voice is possessed of."
...................
    "Comparing the recording of a tenor voice on this side with theVictor is very difficult because I have not [sic] no recent record of any great tenor in mind which was made over here, but to judge by the ordinary tenor voice we record, we get proportionately a better tone. To come, however, to baritone and bass voices, I think here we score heavily, because the Victor Company with the same tightening process convert a bass entirely into a baritone and a baritone into a would-be tenor."
..................
    "Now, what I have written in connection with men's voices applies to a certain extent to women's voices, but more especially to Mezzo Sopranos and Contraltos.
    Articulation. Here, I consider, that our recording scores, and although I do not think that the question of clear articulation has been sufficiently studied - (in fact, I might say I consider it has been ignored) - yet, from the records I have compared, notably Journet's latest, made in New York and in Paris, the clearness of the words is in the favour of European recording.
    Accompaniment.  Here again, I consider, we are ahead. You very rarely get a really properly balanced Victor record. It is all voice and an orchestra somewhere inn the distance. The balance between the orchestra and the voice from an artistic viewpoint is of great importance, and although we are by no means perfect, I consider our balancing much better and the tone of our orchestra when used in accompaniment is superior. It is brighter - crisper. This remark applies to the orchestra when used as an accompaniment; I cannot compare our solo with the Victor's for I have practically none of theirs.
    So much for comparative criticism, but there is one thing I would specially like to cticise both Companies over, ant that is, it does not seem to me that enough attention is paid to the placing of the voice. By that I mean that you may get a record in which insome parts you have the illusion of the singer being in the room, but in other parts he, or she, might be a long way off, which means that some notes come from the lip of the horn, other notes being buried lower down. As concerns the illusion of the singer being in the room, the horn is to the diaphragm as the human palate and lips are to the larynx, and therefore the importance of how a singer produces being emphasised bbefore he, or she, begins to sing, is very great, owing to the fact that there are may different ways a singer can produce."
.....................
[examples have been left out]

Conclusion.

These very intense letters display the process of reflection and maturing that at least this part of the industry was going through while creating from afresh this field of endeavour. We cannot say that we look back at acoustic recording as a primitive precursor to what we are doing today, we must regard it as the pioneering work that created the mindsets to which we are still attuned today. Admittedly, we have better equipment, we may create better and different illusions, but the foundations on which we stand were created in the early years. For this reason, no discussion of the Art of Record Production can be complete without going back to these roots. The present conference has had two out of ca. 50 papers that take this perspective, and I consider the industry to be too ignorant of its roots. My report here is on work in progress.

 

Notes and references:

(1) for copyright clearence reasons I bring no illustrations here. The task of obtaining clearance is simply too large for the individual researcher.

(2) letter from Alfred Michaëlis, Manager of the Italian Branch of GC dated 22 January 1903 to William Barry Owen, Managing Director of GC

(3) Edison, Musicians, and the Phonograph by Harwith & Harwith (1987):
Page references regarding acoustic recording and some poignant quotes:

p. 26ff (Ernest Stevens):
"..... on the floor we had marked seventy-five squares. We'd make tests on each one of the seventy-five squares with each instrument - he'd [Edison, GBN comment] listen to them and say "Assemble them all", and we'd record the whole orchestra. He'd pick apart this, that, and the other thing, and we'd have to make more tests." ........"Maybe it would take four hours in the morning, because he couldn't give me any more time than that. Then he'd assign things." ........... "When we experimented with sound effects, it was wonderful to watch his attention to the very smallest of details ....."

p. 43ff (Anna Case):
" ........ At the time I made the recods, they didn't make them like they make them today. You had gauge your own distance from the horn for every note. At first I was guided by a man holding my arm; if I was to get closer, he'd pull me closer, if further away, he'd pull me this way. Finally, I got so that I could do it myself, and I think I did all the acoustical records that way. ......."

p. 50 (Samuel Gardner):
"With a horn. Oh, it was horrible. I'll never forget that. You had to stand right on top of the horn, and they couldn't splice if I dropped a note. We had to stand sideways so that the sound went into the horn, and you had to avoid striking the horn. ......"

p. 55 (Irving Kaufman)
" ...... All the recordings were made in the horn .......... The other musicians were all around the room. When they had to play a solo, they would move in close, and I would move away from the horn so they'd get the sound of that individual instrument more pronounced than the others."

p. 80f (Rosa Ponselle):  
" ....... With my voice we had to do everything by taking distance from the horn into account. We had to make three tests, right through, without any stops, and we were limited to about three minutes in which to record. So you just held your breath for those three minutes until the last note—either from the singer or from the instruments in the orchestra." ......................... "We'd do three tests. We'd hear the first test and decide, "This has to be further back." Being a dramatic soprano, I had to get way back, measure everything by lines. We'd hear that that was too close, so we'd make a mental note or a chalk line to get further back on this tone, and for the high Cs I'd run way back." ....... "The second test often sounded better than the first, so I kept moving further back on high notes and then would run forward for the middle- and lower-register notes. I got plenty of exercise!" ................. "The players were where I could see them they and the conductor all formed a ring. And the conductor would follow me. So we'd try another take, we'd hold our breath, and if it was a little better we'd erase the chalk, and then remark it. If I sang too loud, it would blast, and the wax master disc would be ruined."  ...................... "I'd say for the chest tones I'd be about three feet from the horn, and for my medium and medium-high tones, about five or six feet from it. If it was a duet, they would have two horns feeding into the same recorder. After a while you'd get accustomed to the distances as you had more experience making records. "

(4) further references to early recording:
the Gramophone Jubilee Book (1973) pp. 64-70
Joe Batten's Book (1956) pp. 32-33, ill. facing p. 66
Talking Machines & Records by S.R. Bottone (1904) pp. 66-68
Gramophones and Phonographs by B. Clements-Henry (1913) pp. 120-132
Die Modernen Sprechmaschinen by Alfred Parzer-Mühlbacher (1902) pp. 30-41
The Reproduction of Sound by Henry Seymour (1918) pp. 88-95, 127-137

(5) I was very kindly given access to the EMI Music Archives during extended study tours to this material in the period 1981-96, and it is slowly being digested and disseminated by permission granted by Leonard Petts, Archivist and Ruth Edge, Chief Archivist & Manager, EMI Group Archive Trust.

(6) at the Edison National Site.

(7) Brock-Nannestad, George: “The Objective Basis for the Production of High Quality Transfers from Pre-1925 Sound Recordings”, AES Preprint No. 4610, 103nd Convention 1997 September 26-29, New York. This text contains much archival material and puts into a technical and commercial context.

(8) Letter to Theodore Birnbaum, Managing Director of the GC, dated 20 January 1909 from Will Gaisberg [presumed], Head of the GC Recording Laboratory, London

(9) Letter to Alfred Clark, Managing Director of the GC dated 27 April 1910, from the manager of the English Branch.