Redefining Early Recordings

This page intends to provide more information on the phonograph-recording sessions conducted as part of three of the symposia from the network. Most if not all researchers in the field will acknowledge how experimental and subject to variability early recordings can be, and how important it is to learn about as much as possible about the context in which a particular recording was made, as well as about the means by which it was preserved, restored and digitized (if applicable). Here, we give an overview of the decisions and processes that went into making the sessions possible.

Who actually made the recordings that can be heard in the individual session pages?

In charge of the phonograph in all sessions was Duncan Miller from Vulcan records. For recording, he used the Fireside Edison (Model B) Phonograph, and New Edison recorder from 1901, and a range of recording funnels. For playback he used the Standard Edison (Model A) Phonograph at Huddersfield and at Guildhall, and at City University, courtesy of Dominic Combe, a Triumph Edison was used.

The digital transfers of the cylinders were made by Inja Stanovic, Duncan Miller and Adam Stanovic.

Did the performers have experience of recording on phonograph?

For most performers, this was the first time they recorded on a phonograph. However, most of the performers featured in the session are experts on early recordings, nineteenth-century performance practice or both. This expertise was in some cases influential in adapting their performance to the phonograph (see performers’ commentaries and interviews in the pages for individual sessions).

How was the repertoire chosen? Some of the pieces seem quite obscure

For the most part, repertoire was chosen by the performers themselves in consultation with the organizers. Generally speaking, we tried to gear the performers towards pieces that were at some point recorded on phonograph: we felt that having an existing phonograph recording of a given piece would provide sufficient proof that the piece would “work” on a phonograph and hence that the session would be productive.

Going to the original sources for inspiration on repertoire also meant moving beyond what we would regard today as the canon. In fact, in the late nineteenth and early twentieth century concert programmes and recordings paired pieces and composers that we would today regard as canonic (Beethoven, Brahms, Chopin, etc.) with others that are frequently less programmed today (Spohr, Lucantoni) or that we would normally consider lighter or less “serious”. The process of searching for suitable repertoire opened up questions about how the present-day canon was shaped, and how recordings contributed to shaping it.

Does this mean that performers were trying to copy or emulate existing recordings?

Emulation is a well-known tool in practice-led research into historical performance practice, and David Milsom and Neal Peres da Costa, among others, have extensively written about it. However, emulation of specific recordings was not the goal of these phonograph sessions. The commentaries from performers in the individual sessions pages show that most if not all performers strived to create a performance that would reflect their own artistic choices, often relying on aspects of nineteenth-century performance practice (e.g. portamento) to different extents; a key factor was making sure that these artistic choices could be technologically captured by the phonograph, which sometimes led to modifications in technique or approach.

Were several takes made? Do the videos show the best takes only?

In the late 19th and early 20th era, making phonograph recordings was an expensive process, and it would likely have been unthinkable to repeat the process multiple times until the perfect take was achieved. Most of our sessions started with a short “test”, i.e. the performer singing or playing a short 20- or 30-second excerpt from the piece that allowed Duncan and the performers to introduce initial modifications to their positioning or to their technique. This initial “test” is not shown in most of the videos. After an acceptable set-up was achieved (as can be seen in the photographs in some of the individual sessions), most sessions consisted of one or two takes of the piece only. In their commentaries, several of the performers spoke about how the set-up led them to adopt an “in the moment” approach, as opposed to the more controlled approach that can appear in present-day recording practices (in which a theoretically limitless number of takes can be made).

Are the digital transfers meant to sound like original cylinders?

Transfers, either of original cylinders or of newly recorded cylinders, add a further layer of mediatization to the sound we hear, which mean they must always be interpreted carefully (and, incidentally, so does video recording – some details might sound different in the videos than they did live!). The phonograph used when making the transfer (which can be the original phonograph used to record, or a different one) will influence the result, as will the skill of the person doing the digitization. You might notice, for example, that the cylinder recording as heard in the video sounds slightly different than as heard in the digitization (for example, a “flutter” might appear). Please bear in mind that this is part of the variability inherent to both early recordings and the methods we have to access them in the present. On some occassions, we provide more than one digitization of the same cylinder, which illustrates these differences.

What conclusions can be drawn from these sessions? What are the main discoveries?

The phonograph sessions were intended to give performers and researchers in the field of early recordings and/or performance practice first-hand experience of the phonograph technology, which is otherwise difficult to obtain due to the small number of engineers as well as the costs. Sources pertaining to early recording technologies suggest – and this was confirmed in the sessions themselves – that early recordings reflect a high degree of variability: the type of wax the cylinders are made of, the type of phonograph they are recorded and played back on, the temperature and set-up of the room and the type of instruments or voices recorded can considerably impact on the final result. Therefore, rather than providing definite answers (which would necessitate a more systematic approach, more prolonged in time), most of the sessions were useful in suggesting ideas and possibilities that can then be pursued through further research; these pertain to both specific aspects of performance practice, technique and technologies (e.g. the ability or otherwise of the phonograph in capturing vibrato and accurate pitch), as well as more general aspects of musicality and artistic vision.

We count, however, the “public phonograph session” as one of the findings resulting from the project. The sessions were attended by audiences between 10 and 30 inviduals and, while these were free to attend and open to everyone, most attendants had some previous expertise in recording technologies and/or nineteenth-century performance practice. Crucially, after a take was made it was immediately played back on the phonograph, and both the performer and audience were invited to comment and ask questions on what they’d heard. We found that seeing the technologies in action, followed by a short burst of very focused listening, catalyzed conversations and facilitated exchanges between experts in ways that were difficult to achieve through more conventional formats (e.g. paper presentations followed by Q&A). Consonant with the range of expertise of attendants, we found that the phonograph sessions did not only stimulate discussion on matters of performance practice, technique or technology, but conversations often encompassed broader questions concerning listening practices, canon and repertoire, recording practices, etc., therefore allowing perspectives on early recordings coming from Historical Performance Practice to interact with others coming from cultural and contextual musicological study.

While professional recording sessions would have likely been made in private and not attracted audiences, amateur recording sessions (such as those conducted by Julius Block and Ruperto Regordosa) often took place in the context of parties or other gatherings, and so it is not out of the realm of possibility that an audience was present, as was the case in our sessions.

Can I use any of the videos, photos or transfers for teaching/research?

We would be delighted if any of the materials provided in this website helps other researchers and enthusiasts of early recordings! The only thing we would ask is that you credit us as appropriate.