If analysis means studying something in order to gain knowledge and understanding of it, then there are any number of ways of analysing recordings, and any number of reasons for doing so. Performers, recording engineers, historians of recording technology and historians of performance practice listen to recordings with quite different kinds of knowledge and understanding in mind: analysis means different things to them. The same applies to acoustic scientists, record collectors and archivists, or communication theorists, not to mention people in the A&R divisions of record companies whose job is to spot the next big hit. The list goes on.

This chapter basically assumes that your reason for analysing recordings is to gain a better understanding of them as culturally meaningful objects, and more specifically that you are primarily interested in the effect of music as experienced in performance, whether live or recorded. In that sense its orientation is musicological, although that too is a term that can be defined in different ways. Recordings are a largely untapped resource for the writing of music history, the focus of which has up to now been overwhelmingly on scores, and recent technological developments have opened up new ways of working with recordings – ways that make it much easier than before to manipulate them, in the sense that we are used to manipulating books and other written sources. I begin by introducing software that makes it possible to navigate a number of different recordings, and to create visualisations that help to heighten aural understanding of what is going on in the music.

In transforming musical thought and action into sound recordings, sounds are converted into electric current and sent along a signal path to a storage medium from which they can be reconstituted at the push of a button. This bit of technological magic has introduced a new figure into the music-making process – not a musician, nor a composer, arranger or songwriter, but one who nonetheless exerts a measure of influence over a listener's musical experience: the recording engineer. The tasks performed by engineers, while practical, have an aesthetic dimension as well, which amounts to an expressive voice in the sound-recording project, an example of what Hans-Joachim Braun calls the ‘technologization of musical aesthetics’. The voices of recording engineers, always present though historically ‘silent’, have long influenced the ways in which we perceive musical sound. Indeed, their accumulated work has shaped essential contours of our recorded musical landscape. Sound recordings are renderings of sound events and, like any rendering, they embody the attitudes, skills, habits and aesthetic stances of those who make them. Some renderings aspire to acoustic realism, others to fantasy, but whatever the case, the sound of a recording has much to do with the technical abilities and aesthetic choices of those whose hands control the signal path.

The performer's view of the microphone is a unique one. Invisible to the listener and a tool of the trade for the producer or engineer, the microphone is the representative of potentially countless future audiences. As such, the microphone confronts the performer as an inhuman critic.

The following addresses the recording situation from the performer's perspective. Throughout I will make reference to my own experience and draw upon that of others who have read and commented on earlier drafts, but I will also set those comments within a broader theoretical framework.

Though predominantly trained and employed in the classical field, I have corresponding experience on the lighter side of music (pop and musical theatre backing vocals, film sessions, etc.). It is here that I have met and worked with performers with a different, pop background who have become more general recording artists or ‘session singers’. I have no direct experience, then, of working as a pop musician, but a comparison (rather than a clear-cut opposition) between classical and pop musicians provides a useful methodological contrast and reveals much about different conceptions and attitudes towards recording and technology. It also underlines some of the organisational assumptions, attitudes to technology and working practices of both. Much of what I say will be true for instrumental musicians as well as for singers, but I will limit myself to specific observations about ensemble singers.

Introduction

Recordings show us that music we think we know intimately sounded quite different in the past. When music sounds different it is different, because music's meaning depends to a very important extent on its sound. Even if you sit at home and read an orchestral score (let's assume you have exceptional powers of musical imagination), the sounds you imagine are those made by a modern orchestra playing as orchestras play today. So however you hear it, there's no experiencing music except through the way it's performed: when the performance changes, the music changes. I think we have to assume, therefore, that pieces we believe we know rather well actually felt different a hundred years ago. We can get some sense of this by reading what people thought about pieces then. Scott Messing's studies of Schubert reception give some idea of how views of him and his music have changed. I've argued in another study that we can hear these changes also in recorded performance, indeed, that in some cases it was recordings of powerful performances (those of Dietrich Fischer-Dieskau especially) that shaped the things people thought and wrote about the composer, bringing to him a new seriousness and psychological depth that was not there in earlier commentary or – it would seem – for earlier listeners.

I was recently recording a CD for a veteran blues singer, and one of the last sessions involved a duet with a guest vocalist. The instrumental tracks had already been recorded and, as is common in pop music production, the final vocals were being added at the end of the process. The two singers were in the studio – each acoustically isolated but positioned so that they could see each other through a window. As they listened to the instrumental tracks on headphones they traded choruses, each also ad-libbing along with the other's lead vocal. At the end of the song, during the vamp, they traded ad libs by drawing from the lyrics of the previous choruses. All well and good in principle – but the reality was a bit of a mess. The first take had great spirit and enthusiasm, but also a lot of ‘problems’ – the vocalists were stepping on each other's lines (the ad-libbing ‘comments’ obscuring parts of the other singer's chorus melody) and there were a variety of nonsensical passages when a less than apt choice of phrase was used to respond to the sung lyric. The trading in the vamp was equally spirited – and equally flawed.

We made a few more complete takes of the vocals but the flaws remained, and where the singers had been loose and excited on the first take, in the subsequent takes they lacked that spontaneous enthusiasm. In the earlier, tape-based era of record production we would have had to resort to recording each section piece by piece, repeating the performance on each part until a satisfactory take was recorded. Eventually we would have got a performance that had sufficient compositional integrity to bear repeated listening, but it would surely have lacked the genuine excitement of that initial run through. Instead, I simply sent the singers to the lounge (they weren’t interested in the process, only the result) and went about reconstructing that first take by rearranging the parts so that they made musical and lyrical sense. This involved moving and adjusting vocal lines so that they fitted neatly into a call and response kind of duet performance without conflicting with each other.

The past thirty or so years that I've worked as a recording producer have certainly formed my view on the relationship between music, recording and interpretation. Perhaps most of all, I have come to see how important it is for producers – themselves a form of musical catalyst – actively to engage with repertoire rather than contenting themselves only with performance. The longer I work in recording, the more convinced I become that composers rather than performers provide our music. The crisis the ‘business’ is now experiencing is due to the lack of platforms available to composers who have something to say to an intelligent, receptive, paying public. In a world where audiences have the freedom to choose what they wish to hear, it cannot be enough to rely on tax-payers to provide whatever meagre exposure composers have as their only means of dissemination. All my observations on performance and interpretation must therefore be taken in the context of an environment offering a surfeit of different musical views on a relatively static number of works. Indeed, discussing whether the studio adds or detracts to someone's performance of a Beethoven sonata recorded hundreds of times already is akin to calculating the number of angels dancing on the head of a pin during the Grand Inquisition. Pop music has stolen the initiative, told us what people are prepared to buy and left intelligent music lovers with less choice than before. If we were dealing with the written word, we would be living in a world of comic books with literature inhabiting the margins and virtually nothing in between except ungratifying verses by some worthy, publicly financed poets.

Today's CD buyer/internet downloader demands, and mostly gets, a ‘perfect’ soundworld: sonic sumptuousness is as important as compositional content, and the performer's prowess goes without saying. If, in so-and-so's new Chopin recording, the piano is too distant, or too tinny, or in an acoustic so reverberant as to blur detail and condense the dynamic range, then it will simply collect dust on the shelf. Recordings with performer errors are similarly undesirable. I am, however, fond of Mary Garden's 1904 recording of Debussy's ‘L'ombre des arbres’ from Ariettes oubliées with Debussy at the piano – an important document of course, but also memorable because after the opening piano introduction Garden enters on the wrong note. This is swiftly corrected by Debussy, who also (just in case) gives her the starting note for the second-verse entry. It always raises a smile. Wrong notes, untidy ensemble or imperfect intonation in live performance are, to some extent, the fragile nature of the business, and Garden's recording is a performance preserved for posterity. Yes, audiences are impressed by impeccably virtuosic playing (a spectator sport akin to gawping at a freak show), but we performers are also curiously and necessarily fallible: attempting, say, the extraordinarily slow speeds Messiaen asks for in the solo movements of Quatuor pour la fin du temps, or a real tutta forza fffff in works by Birtwistle or Maxwell Davies, is certainly technically challenging, and these kinds of dangerous moments in performance, even when they don't quite work, are exhilarating for performer and audience alike.

The Cambridge Companion to Recorded Music is a very broad title. On one interpretation it might take in just about all popular music, the development of which was largely conditioned by recording; on another you might expect an annotated guide to the recorded repertory. We are offering neither of these. Our aim is rather to promote an understanding of the ways in which recording has both reflected and shaped music throughout the twentieth century and into the twenty-first: that is, how it has reflected and shaped not just the music itself, but the ways in which it is produced and the ways in which it is heard. That involves a lot of background information about recording and recordings, which we also try to cover. And ‘music’ in this context is a very inclusive term, encompassing the countless different genres of classical and popular music, all of which have been shaped by the development of recording technologies, originally in North America and Europe, and their subsequent spread across the globe. The way in which music has been shaped by recording is not uniform, however, and comparison of the impact of recording on different classical and popular traditions brings home the variety of conceptions that exists of what recorded music is and might be.

The appearance of this Companion is a symptom of – and, we hope, will further contribute to – the increasing interest of musicologists in music as performance. To someone outside musicology it might be odd to think of it as anything else, but the traditional focus on scores as the repositories of compositional creativity has led musicologists to think of performance as something that happens after the event, so to speak, rather than being a creative practice in its own right. It also signals the discipline’s increasing concern with reception, with the way in which music is given meaning in the act of listening to it, and in the other acts that are informed by listening, ranging from dancing to it to writing about it: recording has fundamentally changed the reception of music, in terms of its nature, its conditions, the places where it happens.

Introduction: Criticism and commerce

For about twenty years, between 1972 and 1992, I practised as a rock critic. While this did mean reviewing concerts and sometimes talking to performers, to be a rock critic was to be a record critic. My first published work was a record review in Rolling Stone, and rock, as a new kind of musical institution, was centred on the record. The founding fathers of rock criticism, Greil Marcus and Jon Landau, both edited Rolling Stone's record review pages, while Robert Christgau, the self-titled Dean of Rock Criticism, started his Consumer Guide, capsule reviews of every rock record released, in 1969.

Towards the end of my time as a critic I began to notice articles about the decline of rock criticism. ‘Where have all the rock critics gone?’ asked Ed Ward, editor of the Rolling Stone history of rock ‘n’ roll, in August 1988, following up his question later that year with the more assertive ‘Rock Critics RIP!’ This was to become a recurring feature-story line. In 1998 Gina Arnold, a leading voice in the next generation of American rock critics, reflected in her turn ‘On the death of rock criticism’ (‘Once it was about passion. Now it's all puff’) in the online Flagpole Magazine, while veteran Italian rock critic, Gino Castaldo, deplored the ‘strong decrease in the demand for critics’.

Introduction

When we consider the use of sound recordings for the transport and presentation of performance, the history of sound-recording technology must be taken into account, but in doing so we need to concentrate on those types of sound recording that have provided an important influence or contribution. Short-lived systems, esoteric systems, and discussions of who invented what are not particularly interesting in this context. This presentation of the historical development will concentrate rather on the performance of the systems then, while they influenced contemporary appreciation and further development, and now, when we have a possibility of providing better signal extraction than ever before. The first is related to the reception history of sound recording, the second to realising the correctable parameters of historical recording.

Readers will have noted the somewhat clumsy expression ‘transport and presentation of performance’. This broadness of purpose is necessary because, seventy years into the development of sound recording, most records being produced began to be edited on a microscopic level into a coherent whole, presenting the performer as he or she would have played in almost ideal circumstances. From that point, performance as a specific sonic event is not created until an edited recording, perhaps in conjunction with a recording of a synthesised sound, is played. Traditional thinking about sound recording and reproduction was that it was a ‘naive’ recording, an image of sorts of a live performance, the later reproduction of an event in time that had really taken place earlier. a large degree the general public is still under this illusion.

Introduction

In 1998 the distinguished economic and social historian Cyril Ehrlich suggested that the history of the record industry could be divided into five phases. Each of these has been driven by new sound-recording technologies. They are: the recording horn and the cylinder (1877–c. 1907); the acoustic disc (c. 1907–c. 1925); the microphone and electrical recording (c. 1925–c. 1948); tape recording and the vinyl long-playing record (c. 1948–c. 1983); and digital sound and the compact disc (c. 1983–c. 1998). To the last phase may now be added the computer file, such as the MP3 format (c. 1998–). This chapter will examine each of these periods. It will outline the dominant technologies of each phase, their commercial exploitation and related artistic developments, principally in the fields of musical repertoire.

1877–c. 1907

The inventor of the phonograph, Thomas Edison, saw its future as an office machine for dictation. Its mechanical basis was simple: a metal stylus inscribed sounds transmitted through a speaking tube onto the surface of a revolving cylinder, in this instance covered with silver foil, which could then be played back to the listener using the same stylus. Edison patented his invention in 1878, the year after its first presentation, and set up the Edison Speaking Phonograph Company to exploit it commercially. Both public and inventor soon lost interest. Three years later, another major inventor, Alexander Graham Bell, together with colleagues, set up the Volta Laboratories to study sound recording and reproduction.

The literary form of the Afterword is an attractive one: it gives one licence to reflect on this Companion to Recorded Music as a whole, compendious and diverse as it is, while bringing out and expanding upon some of its core themes. In this sense the Afterword itself recapitulates one of the key properties of media: the capacity to rework existing ideas or cultural material. In relation to literary media, the seminal work of Jack Goody and Ian Watt identified how the transition from oral to written transmission of knowledge and information enabled the development of a series of externalised aids to thought: the production, in written or graphic form, of lists and summaries, groupings and categories, classification and comparison, and hence the potential for a reflexive and critical engagement with past ideas, and for objectified histories. Visual media such as painting, photography and film proffer their own particular versions of these techniques, among them, variously, framing and focusing, close up and long shot, montage and jump cut. Electronic sound and auditory media proffer yet other analogous properties: splicing and editing, sequencing and looping, sampling and remixing. Writers on ‘new’ or digital media, in turn, have identified in them a series of capacities, many of them prefigured by the properties of ‘old’ (or earlier) media, but all of them reinflected by the physical and symbolic architectures of computer software and hardware: automation, modular and fractal organisation, replicability and variability, and transcoding – the repeated translation of a concept, process, object, image or sound from one format into another.

In 1985 Paul Simon called the South African producer Koloi Lebona into the studio. He needed to borrow Lebona's ears, for Simon wanted to reproduce the quality of the accordion that he had heard on an inspiring production by Lebona before coming to South Africa. Lebona sat in the control room in a Johannesburg studio. He sat there for days, listening and advising, while South African musicians played and the tape rolled and rolled, and Simon laid down backing tracks for the Graceland album. He gave Simon the sound of his own mix. Lebona didn't care about basking in Simon's glory, though he admires him, but he sat there for days, listening intently. For him the hook was Simon's engineer, the famous Roy Halee. For producer Lebona, Halee held the secret to production success: the more he could learn from this expert sound engineer, the more control he could have in the studio as a producer, especially as a black producer in apartheid South Africa's recording studios.

In the ideal studio, the producer, as the studio's client, directs studio sessions and has final authority over the sound, so that the engineer serves his or her client. But in practice, in South African studios during the apartheid era, this was not always the case. The engineer, usually white, middle-class and male, wielded significantly more power working the controls of the console than would probably be the case elsewhere, especially in major studios in the metropoles of the global North. In the 1970s and 1980s, during years of increasing political repression and the concomitant mobilisation of resistance out in the streets, Lebona and his then assistant, Monty Bogatsu, recount being marked as radical inside the studio. They argued for what they believed were their rights to quality sound and insisted on their entitlement to use whatever technological facilities they needed.

Every collector, whether of rare porcelain, jade or Old Masters, dreams of that one find that will astonish their peers and bring fame everlasting: their name will be spoken with bated breath for as long as there is interest in their field. Collectors of gramophone records are no different. For the instrumental enthusiast it might be a previously unknown recording of Johanna Martzy; the vocal collector may dream of finding the two fabled Fonotipia sides of the nineteenth-century tenor, Jean de Reszke. Whatever it may be, each and every collector has the belief that once in their lifetime he, or sometimes she (and it does seem that collecting is a male feature, or problem, depending on your point of view!), will make ‘the great discovery’ that will place them at the forefront of the craft.

At this juncture I must confess to being a ‘record collector’ – there, I've said it! My excuse is that I am a second-generation collector and my father inspired me with stories of his visits to Covent Garden to hear Supervia and Chaliapine (that's how it was spelt then): my earliest memories are of thorn needles being sharpened prior to wonderful sounds coming from an elaborate radio-gramophone. I still remember my first records when I can have been no more than five years old: a job-lot box from Morphet's sale room in Harrogate that included, oh joy of joys, Charles Penrose's ‘The Laughing Policeman’. That's probably why my professional life has revolved around records and recording for more years than I care to think. Love of the artists, and their performances, has been why I am in this strange business of ours. But although I had been on the receiving end of luck in finding interesting and sometimes rare records, I still awaited The Great Discovery.

In 1978 if you played in a Manchester ‘new wave’ band, like I did, you recorded in two ways. You spent eight hours in an antiseptic radio studio taping four songs for what were called ‘Peel sessions’ (whether or not they were for use by John Peel, the kingpin BBC Radio 1 DJ, on his nightly orgy of the aural ‘other’). Otherwise you made records, mainly seven-inch vinyl 45 rpm singles or ‘extended play’ (EP) discs, often packaged in a picture-cover sleeve designed with either little thought (New Hormones label) or too much (Factory Records). Compilation discs, where groups shared track space with others, were an accepted way of gaining attention and they showed the inclination at that time for collective action. Live ‘gigs’ built up a local audience, the radio sessions a national one, but the records brought you local, national and international attention in one bound.

However, bands like ours – which sought to be ‘creative’ in the recording studio – soon encountered an unexpected problem. One night we played in a Munich club where a fan asked for a certain song from one of our records. We obliged. Afterwards she complained that we hadn't met her request. We insisted we had. ‘No,’ she said, ‘I didn't hear it. You know, the one with the bells.’ Ah, the bells, on the record: ‘Sadly we didn't bring our tubular bells to Bavaria.’ To her mind the song was what she heard on the disc, not what we, who made it, played before her ears and eyes. So much for authenticity.

I remember my excitement when first asked to make a commercial recording. My previous studio work had been for the BBC, where editing was frowned on and retakes permitted only in dire emergency. Now I would be able to repeat as often as I liked, subsequently editing my very best playing into a flawless whole. The theory sounded simple, and I approached the sessions with confidence, despite the difficulties of the repertoire, the piano works of Havergal Brian, a substantial footnote to the composer's thirty-two symphonies which includes the monstrously awkward Double Fugue. Optimism began to fade when I arrived at the venue, the concert hall of a conservatoire, to find the recording engineer gloomily assessing various sources of extraneous noise. Besides passing traffic, there was a soft but intrusive hum from the lighting and a murmured conversation of random creaks from the wooden floor, while the otherwise excellent piano had a squeaky pedal and a buzz on one of the upper strings. I learned quickly that in order to make a good recording one's best playing has to coincide with all-too-seldom moments of silence.

Another snag became apparent as soon as we began work. Neither I, nor the fledgling record company, had seen the need to engage a producer, and without an expert second pair of ears I found I would have to listen to each take, a time-consuming process that would make it impossible to establish any momentum. Fortunately, a handful of Havergal Brian enthusiasts were present out of interest, and they quickly formed themselves into a knowledgeable, if argumentative, advisory committee. After my first take, a run-through of the C Minor Fugue from 1924, a lengthy discussion produced a list of recommendations, including a new tempo, a rebalancing of the contrapuntal entries, and the addition of a grandly rhetorical final allargando, all of which I implemented, working for a further hour until everyone was satisfied.

My first experience with the gramophone was a total failure. It must have been around 1960, when I was seven or eight years of age. My father had given me a second-hand radio set with an integrated record player, and I happily went to a second-hand dealer and bought one of those seven-inch discs for just a Mark – a single with Harry Belafonte singing his ‘Banana Boat Song’ of 1957. The battle of the speeds, which had been waged in a country unknown to me at the time, now reached me, and I became one of its casualties. I owned a record player and I owned a disc, but the two did not go together as my record player had only one speed – 78 rpm – for which no records were still made. Instead, the characteristic Belafonte sonority was transformed into a flickering castrato each time I put the disc onto the turntable. To an artist like Paul Hindemith, this might have been a stimulus for experimenting with sounds formerly unknown and unheard of. To me, it was a frustrating experience which led me to neglect the gramophone altogether for some years. Instead, I looked at the back of the radio case through the little ventilation holes to search for Heinzelmännchen (munchkins) making music inside the box.

Later, when I was thirteen or fourteen, I wanted to buy my first classical record. We still had no proper record player: in fact, the old one had been disposed of by then. But I was eager to get a recording of Schubert’s beautiful and haunting Piano Trio Op. 100, even before I had a record player to play it on. I had grown to love the Trio’s slow movement because it was practised by older pupils at my school during a week in a holiday camp devoted to music-making. But the local shop did not have a record in stock; the sales assistant looked it up in a catalogue – it must have been the Bielefelder – and informed me that there were two recordings for 25 DM and 21 DM respectively.

It may be a shame that many significant performers did not arrange to live their lives in times when recording technology was fully developed, but that is no reason to dismiss their performances. The aim of remastering is to make these performances accessible to a modern audience, in terms of both physical availability and the most informative sonic presentation possible. This I call the music to muck ratio, as its determination is necessarily a subjective matter.

I shall outline the processes involved in remastering from 78 rpm originals, with particular reference to a specific example, George Formby Junior's recording of ‘Rhythm in the Alphabet’, issued in 1938 by Regal Zonophone (catalogue number MR 2890).

Having decided on the material required, a suitable source disc has to be found. Curiously enough, this is often easier with art music than more popular fare which, although it sold in greater numbers, was played far more often and looked after less well, being essentially Gebrauchsmusik. In the case of Formby, the situation is not helped by the ubiquitous banjolele solo, recorded close to the microphone and starting half way through the side, when the average steel needle would already be past its best. Aside from the question of wear, there is also the matter of pressing material to consider. The largest market for George Formby's recordings outside the UK was Australia, renowned for producing beautiful laminated pressings of superb appearance, although this was sometimes achieved by over-polishing the metal parts, causing some loss of high-frequency detail.

My serious involvement with recordings, which has already lasted fifty-five years, has given me much of my musical education as well as endless inspiration. Growing up in the South African countryside, I was largely dependent for entertainment on what we had in the house. The Broadwood baby grand piano was vital: my father played by ear (hence my predilection for James P. Johnson and the Harlem stride style) but my mother played only from printed music – we regularly sang songs round the piano to her accompaniment. I do not recall that we listened to the radio much but we had two cupboards full of 78 rpm discs. Among the twelve-inch records, housed in strong cardboard boxes, were classic sets: Toscanini's 1936 NYPSO Beethoven Seventh; Stokowski's Philadelphia Petrushka; the Grieg Concerto with Moiseiwitsch; Faust with Marcel Journet; Fritz Busch's Glyndebourne Don Giovanni; Lucia di Lammermoor with Pagliughi; the plum label Rigoletto with Piazza, Pagliughi and Folgar; and, perhaps most important for my early development, La traviata with Guerrini, Infantino and Silveri. Our ten-inch discs were often not even dignified with paper covers but stood in convenient piles from which eight would be selected at random to be placed on the ‘record changer’ in the radiogram. We always had background music at dinner and it might be anything from Roberto Inglez or Amalia Rodriguez to Sir George Henschel or Richard Tauber or, at the lighter extreme, Layton and Johnstone, Bing Crosby, Danny Kaye or the Andrews Sisters.