Residential symposium at Royal Holloway, University of London, Egham - 20-22 April 2006


Day 1

Thursday 20 April 2006
3.30-4.00pm: Tea
7-8.30pm: Dinner
  • 8.30-9.30pm: Special session on comparing transfers

Day 2

Friday 21 April 2006
10.45-11.00am: Coffee
1.00-2.00pm: Lunch
3.30-4.00pm: Tea
7-8.30pm: Dinner
  • 8.30-9.30pm: Tully Potter, 'Dubs and flubs: transfers I have known'

Day 3

Saturday 22 April 2006
11.00-11.30am: Coffee
  • 11.30-1.00pm: Final discussion
1.00pm onwards: Lunch



Peter Adamson and Peter Craven

University of St Andrews and Algol Applications Ltd respectively

Crackling good stuff: changing expectations

We are confronted with listeners' increasingly modernised expectations - in particular, one of 'no background noise'. The public believes what it is told; reviewers are used to modern recordings and struggle to assess old recordings on the sole basis of reissues; musicians may have an understandably jaundiced view of any recording of their art. All these groups are likely to have a low expectation of 'old recordings' and have no opportunity for learning otherwise: their expectations need to be informed.

Electronic removal of crackle and other background noise is now commonplace. It can sometimes be helpful but in most cases interacts with the music and leaves a residue that, consciously or otherwise, confuses the ear. The opposite approach is to reproduce the crackle with the highest possible fidelity, and let the ear do the removal.

Crackle aside, some modern processing techniques (including digital bit-rate reduction such as MP3) have the apparent effect of 'simplifying' the signal. This raises concerns for the 'presentation' aspect as well as the 'archive' aspect of transcription, these aspects both being important for present and future study of recorded performances.

With the fearless tread of irresponsible but informed dilettantes, we shall present an alternative array of simple approaches, of ghastly results - and of high-resolution audio containing 'accurately' reproduced crackles, such as might cause consternation to a public used to transcriptions that do not have "the distraction of intrusive surface noise between the performer and the listener".

So, what are your own expectations? We hope that some of our demonstrations may surprise you, even if you are used to listening to old recordings. For instance, we may ask you to judge directly by listening: just how much noise do you think we can comfortably leave out? The answer may not be what you expect.


Roger Beardsley

Let them sing - the records will if you allow them

The purpose of the presentation will be to show that most recordings are actually very good, and that it is just poor replay techniques that cause problems. Too often there is a serious lack of knowledge about the equipment used for replay, and the way in which the myriad available pieces of sound equipment operate, culminating in highly inappropriate processing. Additionally, there is all too rarely any real understanding of the way in which the original recordings were made.

These factors, together with an ignorance of the sound of live music, are almost wholly responsible for the many poor transfers. Good, simple replay techniques, coupled with the very best of modern equipment, will, when used properly, make the difference between, say, listening through an open or a closed window. We should not have to put up with over-processed, lifeless reproduction, when the originals are so good. Better a little residual noise than hideous artefacts and sterile sound.

By the use of many examples, various different problems will be highlighted by means of comparative extracts from a variety of different recordings; the high quality of early electrical recordings will also be examined. The audience will be asked to 'assist'!


David Breckbill

Doane College, USA

Issues of documentation and experience in re-releasing historical recordings

Scholars who regularly work with historical recordings both in original form and in modern transfers have recognized two significant areas in which reissues require care from both producers and listeners. The first concerns documentation, which ideally should be both comprehensive and transparent (that is, offering an explanation of the rationale/basis for all decisions made in reproduction and documentation). Such documentation includes: notes on the source's material and the technical ways in which that material has been tailored and cosmetically primped for modern release; identification of personnel, date, and place of recording; a description of relevant discographical information (company, catalogue number, take number, and the like); and justification for the decisions made in deciding the pitch/speed at which the recording has been reproduced (including assumptions concerning prevailing 'standard' pitch, score pitch or transposition, and so on). This latter area is especially important, since collectors of original recordings can work to correct or confirm a producer's decisions so long as the speed decided upon is indicated in the modern release. A survey of contemporary practices shows variable levels of comprehensiveness, reliability, and (particularly) transparency in documentation.

The second realm of concern is that of experience, a concept fundamental to understanding many recordings being currently revived. Just as the ideology of 'progress' can distort our perception of any given moment in musical or compositional history, so the assumption that early recordings were merely primitive ancestors of modern recordings deserves a re-evaluation. In particular, the fact that time limitations of early recordings prevented large-scale compositions from being recorded without interruption has discouraged proper appreciation of recordings intentionally devoted to shorter pieces or excerpts, which ideally require a different sort of listening from that ingrained in listeners to modern formats. When playing an early recording from an original pressing, the physical actions required of the listener to commence and conclude that playing have a way of focussing attention; in addition, the knowledge that early discs and cylinders are subject to wear and sonic deterioration through frequent playing further heightens one's awareness and interest in what one is hearing. The impulse to transfer such recordings to a more convenient format that allows for both casual acquaintance and in-depth study is understandable and worthwhile, yet the experience of hearing early recordings in modern reissues is substantially different, especially when the reissue containing (say) a comprehensive survey of the recordings of a particular performer, and when the reissue inevitably presents an individual recording as one of many such recordings linked together in an uninterrupted sequence. Moreover, recordings that in original format are easily distinguishable from one another as physical entities lose significant facets of their individuality in this reissued environment. Collectors who value stories concerning the provenance of their copies of recordings and of the minutiae of those recordings' physical characteristics may seem obsessed by an arcane passion, but such interests help to demonstrate the persistent validity of an earlier conception of what constitutes a recorded product. These considerations give rise to reflection concerning the changing ontology of recordings in relation to performance, the listener, and the commercial products that enshrine them, and additionally prompt numerous suggestions for the presentation of early recordings in modern formats.


George Brock-Nannestad

Patent Tactics, Denmark

Using recordings for documenting performance - an overview of the field

Moving away from traditional approaches to the study of performance is a laborious process, and CHARM has taken a huge step by stressing e.g. the disciplines of discography and automated analyses of performances. The objectivisation of performance analysis by means of existing recordings rather than by live experiment or live performances occurs in a multidisciplinary context, and the present overview attempts to identify the main classes of disciplines that will usefully contribute to insight and to guide interpretation of facts.

Approximately fifteen subject areas have been identified that each have their own fertile research traditions, each have publications in their own rights, however we need to be able to translate their communication space or frames of reference in order to be able to utilize their results.

The presentation will put these subject areas in a suitable context, and audiovisual examples will be given. Time will be provided (albeit brief) for audience participation in the discussion of each subject area.


Per Dahl

Associate Professor, Department of Music and Dance, University of Stavanger

Love is in the air (ear?) Musical expression and soundscape in the recordings of Grieg's opus 5 no.3: Jeg elsker Dig / Ich liebe dich / I love thee

In order to detect changes in the interpretation of this song by Grieg in the gramophone era, I started with making a chronological discography of the recordings and tried to develop a theoretical frame that accepted the listener's choice of interpretation. In this process I had to work out some registrations of musical expression from the soundscape in each recording. This has included a registration of tempo deviation in every phrase of the song, registration of language, accompaniment and form/arrangement, as well as personal evaluation of the soundscape (whether it belongs to the concert tradition or the comers tradition, simple parameters on reverberation, microphone placement, balance between singer and accompaniment, dynamics as well as characteristics about the singer's voice production). I will present an old recording which is incomplete and with the wrong speed (how can I adjust it to fit my inquiry?), different pressing of the same recording (which one should I use for my evaluation of musical expression) and a modern recordings with strange sounds in the studio. (What is happening in the air/ my ear?).


Michael H Gray

Head, Voice of America Library and Audio Services

Behind the studio door

During the first two decades of electrical recording, what occurred behind the closed doors of the recording studio was a well-guarded secret. Proprietary technologies and techniques, and sensitive relationships between company representatives and recording artists, were not to be revealed to the public at large, and certainly not to the eyes and ears of one's competitors. Only the musicians making the music, and their collaborators behind the glass window of the recording booth, were privy to the secrets of the recording room.

This talk will open those doors. Using rare published and unpublished photos, recording session papers, internal documents, and published evidence, this presentation will explain in a non-technical framework how electronic tools, including the disc-cutters, amplifying equipment, and microphones, were used to convert sound in the air into physical modulations in the groove of a 78rpm recording. In particular, the presentation will examine in some detail the environments in which recording took place, and how these environments, along with characteristics and placement of the microphones in them along with the equipment through which their signals were transmitted, 'created' the sound waves captured on the wax 78 master disc.


Ted Kendall

Remastering made easy

The student of performance practice needs to understand the principles of remastering, since this is the gateway through which he will experience most historic recordings. However, like many specialised techniques, remastering has acquired the aura of a black art, with its own arcana and obscure jargon, to say nothing of a rampant cult of personality. Whilst it is true that, for remastering at the highest level, there is no substitute for aptitude, skill and the right equipment, the fundamentals of the subject develop logically and are simply tamed by a systematic approach and the correct application of basic principles. In the spirit of Sylvanus P Thompson's Calculus made easy , the speaker will attempt 'to present to his fellow fools the parts of the subject which are not difficult. Master these, and the rest will follow.'


Ward Marston

The challenges and the joys of remastering acoustic recordings

Acoustic recordings have always fascinated me, both from the standpoint of collecting and remastering. The first commercially available cylinder recordings were made in 1890 and it wasn't until 1925 that electrical recording began replacing the old acoustic process. During this 35 year span, a huge number of important records were made. The real musical value of these early recordings is, to say to least, a controversial subject. Today's music enthusiasts who are not 78 collectors find it tedious to listen to them, and music scholars tend to dismiss them as too primitive to be useful in determining performance practice.

How much can we actually glean from acoustic recordings? In my short talk, I will play acoustic recordings as far back as 1888 through the end of the era demonstrating how the sound of acoustic recording evolved. I will play one particularly flawed recording and discuss a recent restoration solution. I will also play some astonishingly good examples and raise the question of whether extreme interventionist methods are necessary or even advisable. Through these examples, I hope to convince my audience that there is more on these scratchy old records than immediately meets the ear.


John McBride

Electro-Mechanical Research Group, School of Engineering Sciences, University of Southampton

Non-contact surface scanning systems for the retrieval and protection of archived sound recordings

The aim of the research is to develop methods for the rapid non-contact scanning of archived sound recordings such that the sound contained can be stored in a digital format. The project is focused on the development of sensing methods and the implementation of the sensing methods into two measurement systems, one for cylindrical scanning and one for planar scanning.

The aim of the sound archive project is to recover important and damaged recordings of interest to the general public, historians, linguists and musicologists. The proposal is timely and critical, since at present many sound carriers are archivally unstable and at risk of deterioration.

Initial studies are presented on a cylindrical scanning system, and the associated data processing method.


Mark Obert-Thorn

Transfer fundamentals

This presentation will trace the transfer process from the procurement of good source material to the finished master, focussing on the steps which are (or should be) common to successful restorations. Among the topics to be examined are: selecting the proper stylus; issues related to pitching and playback speeds; recording curves and equalisation; side joins; and computerised noise reduction. In addition, some of the 'finishing touches' restoration engineers must decide upon will be considered. Recorded examples will illustrate the step-by-step procedures.


Renee Timmers

CHARM, King's College London

Listening to historical and modern recordings: the effects of age and recorded version on the perception of performance

Listening to historical recordings seems a skill that one has to develop. This development does not necessarily concern getting used to the hisses and cracks of an historical 78, but rather getting used to unfamiliar performing styles and limitations on transferred overtones.

A series of experimental studies was conducted to investigate the sensitivity of listeners to performances on historical and modern recordings. The experiments examined the effect of 'age' of a recording and 'version' of a recording on different perceptual dimensions, including the perceived quality of a performance, the perceived emotional affect, and the perceived clarity and variety of performed sounds. All experiments used the same material: four fragments from Die junge Nonne, which is a late song by Franz Schubert, sung by six performers from the beginning, mid, and second half of the 20th century in a clean and a noisy version.

The results differentiate between perceived dimensions of performances that are and are not affected by the age and version of a recording. Generally, affective dimensions are less affected than perceptual dimensions and judgements of quality. For example, the communication of emotional activity was successful for historical recorded performances as well as for modern recorded performances. In contrast, the evaluation of the quality of a performance depended strongly on the age and version of the recording.


Martha Tupinambá de Ulhôa

Universidade Federal do Estado do Rio de Janeiro

What can we hear from Casa Edison's old records?

Working with historical phonograms of Brazilian popular songs, most of them done for Casa Edison, the first record company in the country - active from 1902 to 1932 - I became interested in learning more about the recording process of old 78rpm mechanical recordings. Casa Edison, commanded by Fred Figner (1866-1946) produced between 1902 and 1927, circa 7,000 double-faced phonograms, 728 items of which were done in 1902, what makes this repertory one of the largest in the world. A great part of this repertory is available online (, having being already transferred to a digital media. Having chosen to study closely two song genres through recordings (the modinha and the lundu) now it's time of identifying which musical aspects can be analyzed, and what has been the transfer process impact on the resulting sound. The session will consist of a presentation of a sample of Casa Edison's repertoire and the kinds of problems our research group have been encountering while dealing with historical recordings.


Simon Trezise

School of Music, Trinity College, Dublin

Emotional and musical responses to mutating sound quality in Vaughan Williams' recording of his Fourth Symphony

Musicologists tackle historical recordings in a number of ways. Initially the emphasis was on timing information, which is both easily extractable and readily categorised. Even this avenue of exploration cannot be taken for granted. The conductor Herbert von Karajan frequently demonstrated to students that the same tempo, measured metronomically, could be made to sound livelier by changes of accentuation, emphasis, etc. without departing from the pulse. Similarly a passage might be made to sound duller by the same expedient. The information that the researcher might extract from recordings to illuminate a sense of liveliness in the tempo chosen by the performers (for example) is far more difficult to categorise and present to the reader of, for example, a learned article than the original timing information given in beats per minute. Even if the means may be found to do it - and the information stands on the edge of a vast array of similar strands of elusive information enshrined in recordings - the recording itself is no a single, immutable source of sound for dissection and analysis. Multiple factors make a recording variable as a sonic object. Playback conditions are significant; so too is the manner in which the transfer has been carried out.

In order to provide a theoretical basis on which to make relatively simple observations of the effect of sound and sound quality, I will use a set of simple, Kate Hevner-inspired adjectives to evoke responses to differences of texture, articulation, and so on. Taking recordings of Vaughan Williams conducting his Fourth Symphony (and other illustrations as appropriate) I will illustrate differences in sound of transfers and suggest that, subtle though these differences sometimes are, they may affect the manner in which we respond to the music. There are at least four commercially available transfers available for discussion, and I will also make up examples of my own from the original 78s. The presentation will be devised in such a way as to pose questions to the transfer engineers present and to musicologists, from whom I will be seeking the vocabulary to characterise differences of muscal affect and emotional response. At the same time, aspects of sound quality that might encourage one to regard one transfer as perhaps livelier than another will be examined.