As recording formats become obsolete, sound archivists are rethinking the paradigms around methods of preserving our audio heritage. By Will Prentice of the British Library.

It’s an archivist’s job to take the long view, and by habit and training we inherit the memories of our institutional forebears. In pointing out similarities with the past there’s a danger of implying that ultimately it’s business as usual, so I’ll start by saying that the current state of affairs brings two fundamental crises to sound archives. First, the archival paradigm of putting stuff on shelves and hoping it will last forever is no longer enough. Second, when so much is freely (if illegally) instantly accessible online, what purpose do archives serve?

The first sound archives began at the very end of the 19th century, and within a few years they were grappling with format obsolescence. The writing was on the wall for vertical cut discs and wax cylinders by the outbreak of the First World War, and we’ve been struggling to maintain the ability to play old formats ever since. Technology comes and goes; archival material just accumulates.

Formats do eventually just die, though. There have been many stories over the past few years about the resurgence of vinyl, of the cassette, of the hand-crafted artefact. All of this is healthy, enjoyable and satisfying, but increasingly unsupported by available technology. You and I may have analogue decks today for audio playback, but there’s no professional grade analogue tape equipment being mass produced any more, and very few professional turntables. By professional grade, I don’t mean kit that will flatter by glossing over any weaknesses in the sound being reproduced, I mean kit that will replay the information in the carrier transparently, with as little deviation or subjective ‘enhancement’ of what’s on the disc or tape as possible. An archive aims to preserve history, not rewrite it, so any subjective alterations to the sound captured by the original recordist, deliberate or otherwise, must be minimised and documented.

Manufacturers would continue to make and support this equipment if they could see real evidence of a profitable future in it. Without the numbers, though, no one, however skilful and determined, can produce vital and complex spare parts affordably. Within a few decades, our old equipment will wear out and won’t be adequately fixable or replaceable. If we want our audio heritage preserved sustainably, we need to digitise, and we need to rethink the archival paradigm.

The old paradigm – of careful storage and hoping for an everlasting medium – is no longer sufficient for audio. While many older sound carriers are ironically fairly stable and somewhat accessible, newer media are usually less so, either through degradation (the dye in your CD-Rs fades in daylight) or system obsolescence (when did you last see a new DAT player for sale?).

The new paradigm no longer hopes for a long life: it plans for a short one, and recognises that unlike analogue, digital technology could allow audio information to flow unchanged from one carrier to another into infinity. We have conceptually separated the ephemeral carrier from its everlasting content, and assume that intermediate carriers will be junked every few years.

The argument that older media offer the nearest thing to proven longevity is true up to a point, but compared to other cultural heritage, all sound recordings are babies. Our descendents should be able to listen to sounds that will be as old to them as Stonehenge and the Dead Sea Scrolls are to us today. We have the ability and therefore the responsibility to do so, and the only way at present involves converting the audio to file-formatted digital data. This should be done in multiple copies, in multiple locations, and on multiple storage types.

As to the purpose of sound archives in an era where everything is becoming available online, I think the answer will increasingly be authentication and intelligent navigation – necessary components of research. Instant accessibility still seems miraculous to me, but much online material typically carries very little context or reliable provenance, making it easy to misinform, or to disinform. Archival documents carry their provenance with them in the form of metadata, whether on a sleeve or virtually, with the intention of making clear to the listener exactly what they’re listening to. Historical research demands reliable provenance, and our provision of explicit and impartial metadata will increasingly be where our value lies.

This isn’t necessarily a function of new technology: legitimate physical publications can also revise history. Can reissued their 1971 B side “Shikako Maru Ten” on the compilation Cannibalism 2, and without explanation it was edited, coming in at about a minute shorter than the original. I’m not questioning their right to do this, of course; they’re the artists and it’s their work. But this subtle revision demonstrates the difference between the outside world going about its natural business and the archival imperative.

And of course, increasingly archivists make audio available online for researchers to use anytime, anywhere. The British Library currently has 25,000 recordings freely accessible to all via its website, including everything from the world’s largest collection of Decca West African 78s, to 500 recordings which make up an oral history of jazz in Britain. By arrangement with rights holders, we also make a further 25,000 recordings available to the UK higher and further education sectors.

I mentioned routinely junking intermediate data carriers earlier. This doesn’t apply to the original physical media, which as artefacts in their own right carry more than sound. In the same way that ethnomusicology encompasses not only the study of music but the study of people making music, audio carriers and their packaging tell us something about people making and using music.

Looking at sleeves throughout the ages, we can see reactions to the threat posed by blank media, from the “Home Taping is Killing Music” logo in the 1980s, back to dire warnings from Thomas Edison against making duplicates of his cylinders, or using his blanks as masters in the duplication of records. If you wanted to quantify how the threat of illegal copying has grown over the years, you might look at the changing affordability of blank media. In 1904 a blank cylinder could be bought for seven and a half old pence, equivalent to about £60 today, or the cost of a two terrabyte hard drive. The cylinder typically holds a two minute song; a two terrabyte hard drive typically holds over half a million songs, or more than four years of audio.

Technological evolution driven by the marketplace has been a dominant theme throughout the history of sound recording, and will likely continue to be. To drift into an even longer view and paraphrase John Blacking, there is a relationship between humanly organised sound and soundly organised humanity, and this will transcend crises in the recording industry and in archiving. But it’s in all of our interests to continue documenting and preserving it, somehow or other.

Will Prentice is Head of Technical Services at the British Library’s Sound & Vision Department: access its sound archive at sounds.bl.uk

Leave a comment

Pseudonyms welcome.

Used to link to you.