AV Archiving: changes, choices and challenges Ray Edmondson Uploaded 3 July 1997 | 5,900 words Abstract

William D. Routt has argued (elsewhere in this publication) that, despite a tendency on the part of film analysts to assume the text under discussion is stable and constant, `textual plurality is a condition of the study of film'. Professional audio-visual archivists face this problem daily. It is one aspect of the question of textual integrity - a subject which underlies some of the main philosophical issues of their profession. In recent years, these have been matters of increasing debate - in the professional literature, in conferences, and within individual archives as these issues are faced daily at a practical level. One synthesis of this discussion is the author's "A Philosophy of Audiovisual Archiving"(1), and the observations which follow are an exploration of some of the technical issues which it raises. Archival purposes and goals are closely related to the technical nature of the media being archived and this structures both how and why archivists make decisions.		1. Begun in 1994, this document - still in draft form - was developed by the present writer with the collaboration of members of AVAPIN - the AudioVisual Archiving Philosophy Interest Network, and has been extensively discussed in the conferences and literature of the various international AV archiving federations. Work is currently proceeding on draft three with a view to final publication by UNESCO in 1998. A copy of Draft Two is available on request from the author. The observations in the present article relate particularly to sections D1, D2 and E3 of the draft document and are best read in conjunction with those.

Technical/philosophical issues 1. Format progression AV media are exposed to the effects of format progression: the constant displacement of the old by the new. Once the commercial life of a particular format is ended, and the technology is no longer supported by the industry, it must still be maintained by AV archives - or the material recorded in that format will eventually become unreproducible and therefore irretrievable, no matter how stable the carriers. Repetitive transfer of content from old formats to new is economically impractical, so maintaining the technology - and the associated skills - is the only alternative.....at least for as long as it is practical. But if the economic and practical limitation were not there - what then? It is theoretically possible that, with digital formats, content can be copied repeatedly without quality loss or human intervention - and therefore very economically. Format progression would then no longer be a problem: data could be easily moved from obsolescent to current formats at will. For pre-digital formats, this raises an aesthetic and ethical question. Whenever content is moved from one format to another, what is lost or changed and does it matter? When one copies the content of an Edison cylinder to a CD, is anything important of the content or the subjective listening experience lost? I once had the experience of hearing, in a private cinema, a 1932 variable density optical sound track reproduced through period equipment, including a 1930's conical speaker whose outlet was the same size as the screen behind which it was placed: a vast piece of equipment which gave the sound enormous depth. Subjectively, I've never heard anything like it since: the same track played through modern gear sounds to me quite different and (presumably) differs from the experience of audiences who saw the film in 1932. Does that difference matter? Another example is surface noise. The characteristic hiss and crackle of 78 rpm pressings, played by a steel needle, was a part of the listening experience and perhaps there were times when a virtue was made of a necessity. Surface noise was also characteristic of optical sound tracks in the 1930's and later: film producers introduced low-level, unsynchronised background music to their tracks precisely to distract attention from the hiss, thereby creating some distinctive musical styles in the process. If we now digitally remove the hiss, will we in fact unbalance the sound track and misrepresent it? Marshall McLuhan once claimed that the medium is the message. Replace medium with format. How far does it hold true? And how much may we permissibly change the message in order to give access to it - in a newer format, say, or over the Internet? 2. Carrier/content principle Format progression, as well as the inherently fugitive nature of AV carriers, leads us inevitably to the carrier/content principle: the notion that AV archives are able to separate the concept of the carrier (the physical disc, tape or film) from the content (the recorded sounds and/or images comprised therein). Because the former may be relatively transitory, the latter becomes the object of preservation, and the skills of the restorer, the technician and the collection manager are bent towards ensuring the permanent survival of the content with minimum loss of image or sound information in the process. Since such loss means - to that extent - the loss of the work itself (or its transformation into a different work) the retention of a collection in status quo condition involves an appreciation of these realities. However, AV archives do this of necessity rather than choice: it is the "least worst" preservation option currently available. If a film or tape could be made to last indefinitely, other things being equal, this would be far preferable. Now the dilemma: how far can carrier and content be divorced? In the museum world, for example, a copy of the Mona Lisa - photographic or otherwise - would never be an acceptable replacement for the original artefact. AV carriers can have artefact value in their own right. There are physical and evidential characteristics that do not transfer when content is copied to a different format. How far may we ignore or discard these? Let me cite some audio examples. The Beatles' Sergeant Pepper LP contains a repeating sound loop on the off-centre playout groove in the middle of the disc, creating both a visual and an aural impression as the pickup arm oscillates - the point of this effect is lost in any other format. Graphically, the Sergeant Pepper sleeve is designed for the LP format and loses impact when miniaturised for CD or cassette. Or again: cardboard records produced as advertising gimmicks or ordinary releases have a visual or tactile character which can't be divorced from their audio content. And consider that the size, shape and practical limitations of the cylinder and 78 rpm disc, and the associated recording systems, affected the content: it determined the standard length of a musical number or even the tempo at which an orchestra played a piece of classical music. When we move to film, we will find claims that the aesthetic impact of a tinted nitrate print cannot be exactly duplicated by any modern copy; or that the vibrancy of an old-style Technicolor imbibition print or other superseded colour processes cannot be captured by the chemistry and optics of modern colour printing. Further, original negatives of films from, say, the 1930's exhibit physical evidence of editing and production processes which are lost when the moving image content per se is lifted onto another format. 3. Quality principle The copying of images and sounds is an art as well as a science, and involves a myriad of subjective judgements. There may be questions of fine tuning that can subtly change the content of a work. There can be more basic issues, such as copying from a superior to a qualitatively inferior format, or with poor quality control of the process, which can dramatically change the quality of the content. There can be deliberate manipulation of content which can amount to misrepresentation, or the deliberate falsification of history. There can be practical or economic imperatives which dictate that an inferior copy of something from a deteriorating carrier is better that no copy at all. Many AV collections bear evidence of this kind of choice. When does the quality loss become too great to be acceptable? Copies in smaller, lower quality formats than the original will often be made for presentation or access purposes: the reasons can be practical and/or economic. Audiences or customers can be profoundly misled by such practices - " it sounds like that because it's an old recording" or "it looks pretty good for an old film". What obligations do practitioners have to forestall or correct such impressions? How can it be done? Is there an outer limit beyond which we should not go? Obvious examples are the audio transfer from, say, CD to compact cassette: or the transfer of a 35mm film image to a domestic VHS cassette. Even when this is done well, the difference is palpable to those who know what has happened. All too often, however, the transfer is done badly: how many of us have had the misfortune to buy a VHS copy of a vintage film, only to find ourselves watching a murky or disjointed travesty of the original work? Does the average viewer know that it was never meant to look like that? How much effort we, as AV archivists, should invest in correcting such impressions - in educating our audiences - is a difficult question. Within limits, we can determine the quality of the work we are responsible for. We can also invest resources in "educating" -making the quality equation clear in writing, for example, when we supply a video to a client. Could we go further? How about lobbying for an industry code of practice so that every cassette label bears a statement declaring the nature of the master material from which the cassette was derived? Is this our proper role, or beyond the bounds? 4. Survival principle An AV archive does not put at risk the survival of a work which it is preserving in order to meet short term access needs. It has a long time perspective: it may have to resist short term access demands in order to meet them effectively in the long term. It does not exploit today with no thought for tomorrow. It's simple to say: not necessarily simple to do. "Risk" means different things in different contexts: the choice may not be between black and white, but between the lesser of two evils. At one extreme, an archive which provides no access at all - in order to avoid putting any of its collection at risk - is useless to its constituents and may therefore be putting its own survival, and that of its whole collection, at risk. At the other extreme, providing anything and everything on demand - regardless of danger to the material - may be a total negation of preservation responsibility. If archives cannot afford to make access copies of material in demand, should they always say no? Under what circumstances, if any, should they say yes? Should they ever put a unique carrier at risk to satisfy an access request? 5. Documenting principle AV archives need to observe high standards in the documenting of acquisition, access and other transactions so that they can be seen to be accountable and trustworthy in their dealings. Because of the complexity of their collections, precise housekeeping records are essential. In technical work, because it involves subjective judgements, an "audit trail" is essential, and requires clarity in concepts and terminology, as well as accuracy and consistency in documenting. It begins with an accurate analysis of the technical characteristics and condition of the carrier concerned. It continues with the management of vaulting and movement. When a copy is made, the documenting of the choices and judgements made by the technician is essential if the new copy is to be understood and evaluated in context, and the trail back to the original is to always be clear. Unless this is done, subtle changes may accumulate and the work may change gradually over time until it becomes impossible to see or hear it in its original form, and indeed to know that this change has taken place. This is especially crucial in cases where significant restoration or manipulation is involved to produce a result of acceptable image or sound quality. What processes were used? What choices were made? Is the result believed to be an approximation of the original or has it been "enhanced" in any way? How widely is such documenting practised? Does it/ should it apply equally to analogue and digital recordings? Ethical issues 1. Responsibility to the public Collection material can be made available to users in a variety of forms: reconstructions, enhancements, compilations, excerpting, abbreviation, and in a range of formats different to the original one. The 35mm feature film released on VHS cassette is a common example; the compilation of excerpts from vintage radio broadcasts on CD or cassette is another. It is very easy for the consumer to form a totally wrong impression of the nature of the original work by being exposed (for example) to an inferior quality copy, or one from which essential context is missing. It follows that it is incumbent on the archive concerned to correct such impressions when they occur, or better still, take steps to avoid giving the wrong impression in the first place. One way of doing this is to present material only in its original format, though for many usage requirements this is, of course, impractical. Another is to ensure the essential context information is readily available to the user or the audience so that misperceptions can, as far as possible, be avoided. This can be done in a variety of ways - sleeve or slicknotes, screen titles, accompanying documentation. Taking it a step further, there is a case for developing a standardised pro forma statement or code (that might, for example, be adopted by all members of the International Association for Sound Archives - IASA (2)) for major projects, such as a feature film reconstruction or a CD reissue of a collection of vintage recordings. This might, for example, set out the parameters and purpose of the project (which guide all the technical and artistic decisions), a description of the work done and the research undertaken, description of the source material and its condition, explanation of judgements and choices made, a statement on how closely the result matches the stated parameters, the time frame and completion dates of the project, and complete credits setting out the contributors to the project, and their roles.		(2) IASA: International Association of Sound and Audiovisual Archives. Secretary general: Albrecht Haefner - email: haefner@swf.de

Tastes and expectations vary over time. An enhanced or reconstructed version that works for consumers in the 1990's may not work nearly so well twenty years hence. When the next enhancement is done, we need to go back to the source and the audit trail..... 2. Moral rights Copyright has been a fact of life for AV archivists since the beginning. More recent in origin is the issue of moral rights, especially those which are increasingly recognised as accruing to indigenous peoples whose cultures involve perceptions of sound and image recordings which are quite different from those of Western cultures. For example, to listen to the recorded voice of a person now dead is not always something to be taken lightly. Or again, to permit female members of staff to be involved in the duplication or examination of film footage depicting sacred ceremonies revealed only to men is an unacceptable breach of cultural norms. There may, therefore, be particular reasons why individual technicians in an archive may not, with integrity, be permitted to access and duplicate certain collection items. Managing and observing such strictures is difficult, and often inconvenient, but is an aspect of professional integrity which is is essential to observe if the archive is to gain the trust of such communities. New problems/ new solutions? Digitisation looms on the horizon as the apparent answer to an AV archivist's prayer - the answer to quality loss, to expensive storage dilemmas, to unlimited access. How true this proves only time will tell: it's not the first time a panacea has been hailed and I suspect it won't be the last. Amid all the anticipated advantages, is there a down side? 1. Digital modification As we are already aware from journalistic practice and television commercials, the digitisation of images permits the undetectable manipulation of images of reality. It is easily possible to change history, to remove objects or individuals from photographs or video images, to change time or place, to synthesise. The process can be subtle: the film, video or sound editor has long been used to reconstructing reality in order to meet the practical requirements of a newscast or a documentary. Now the possibilities can go much further. What is the legitimate area of discretion for the archival technician? What is outside that area? 2. Viruses We are familiar with the physical and chemical "diseases" which attack discs, tapes and films and our collection storage repositories are designed to deal with threats like excessive temperature or humidity, fungus, chemical reactions, physical mishandling and so on. We have less experience with the man-made diseases - "viruses" - which attack computer data. These have already wreaked havoc in the commercial world. Will they ever move in on the image and sound data banks of the future? Is it possible to imagine a virus, created to tamper with image and sound data, moving into a digital archival collection and effectively destroying it over a period - perhaps a very short period - without anyone knowing until it was too late?. It is technically possible. Would anyone ever be so warped as to invent something like this? The answer, unfortunately, would have to be yes: there are enough examples of deliberate destruction of cultural heritage around the world, for purposes of political or economic control, or even for sheer malice, to conclude that digital AV collections would not necessarily be exempt, however far-fetched it may sound at present. Indeed, if all such collections could be reached by the Internet, could a single hacker wipe out the world's digitised AV memory.......?? The "duty of care" which AV archivists exercise over their fragile collections may need to move up several notches as digitisation moves in. Degradation which we now detect through visual, aural or tactile examination may in future need to be detected by the skills of data diagnosis. Moreover, digital damage can be widespread and instantaneous - unlike chemical or physical decay, which takes place progressively. On what carriers may we safely keep our digital collections? If there is data loss, will it be recoverable - and will we only be aware of it after the event? Will insurance mechanisms, such as the creation of multiple copies - perhaps in different formats and stored in different places - be a necessary and/or sufficient safeguard? In the digital future I expect AV archives will have an ever greater reliance on the ethics and integrity of their technical staff, in the storage, copying and manipulation of sounds and images. They will exercise great power over the continuity of history. This places great responsibility on them and emphasises the need for appropriate codes of ethics and formal training. Film archiving in Asia - the challenge of tomorrow To relate philosophical and theoretical considerations such as the above with the real world in which audiovisual archivists work, let us now turn to a case study - the present and future situation of film archiving in Asia. Film archiving began in Europe and North America in the 1930's and is traditionally most fully developed and best resourced in those parts of the world. That, at least, is the import of one survey conducted several years ago by FIAF (3). More recently, in 1996, the National Film and Sound Archive of Australia, under contract from UNESCO, conducted a survey which gives a much more detailed picture of archiving in Asia, and makes an interesting comparison (4). By their nature, of course, the results of surveys are only as good as the information which goes into them, and they are at best an incomplete snapshot of reality at a particular time. They are useful not so much for detailed statistical comparisons, as for more generalised observations which may have implications for future action.		(3) FIAF: International Federation of Film Archives. Secretariat: email fiaf@mail.interpac.be (4) "Survey of Asian film archives", conducted December 1995-January 1996 for UNESCO by the National Film and Sound Archive of Australia. UNESCO contract number 372.057.4

It has been said that film archiving was born in poverty and reared in adversity: certainly, I know of no archive anywhere which claims to have all the resources it needs. Of course, some archives are better off than others, but it remains a question of degree. However, I would make three general observations. First, that - although there are some notable exceptions - film archives in Asia/Pacific are generally younger than their Euro/American counterparts. Second, that the last few years has seen rapid growth in their size and number, and great energy and commitment poured into developing the field. Third, that these circumstances coincide with the technological changes discussed above. Against this backdrop, I would like to turn to several challenges that I think face the region in the immediate future.

		Go to'Part 2'

Home | First release | Classics & re-runs | Reviews | Trailers | Cast & credits | Search | Comments