video and film to digital formats. For audio and video this has been for
preservation. The old carriers are obsolete, they are deteriorating over a timescale
measured in decades, and so most audio and video collections have projects for
transferring content onto digital carriers: CD, DVD, digital audio tape (now
itself obsolete), digital video tape, datatape and hard drives (e.g. the wax cylinder
project of the University of California, Santa Barbara13).
Almost as though technology were wilfully intensifying the pressure, the price of
hard drives has dropped faster than that of other forms of mass storage. Discrete
media such as DVD require robotics to create a mass storage system. Datatape
also requires robotics, but can produce relatively larger storage systems because
of its greater storage capacity (a modern datatape can hold 300 times as much as
a CD14). Hard drives now compete in both capacity and price with datatape,
though computing total cost of ownership (TCO) over 20 years is not easy - for
any media.
The point is not whether disc or tape is cheapest. The point is that disc storage is
now (2004) roughly 100 times cheaper that it was in 1997, and so it is possible
to put large volumes of storage online at relatively low cost. The benchmark
figure is US$1 per gigabyte for purchase (construction) of a petabyte of disc
storage with all system components for supporting online access15, though
average mass storage systems currently have a TCO of more like US $10 per
gigabyte per year.16
EC Projects
In addition to the general technology developments just reviewed, there has been
a decade of concerted European effort, with considerable EC research support,
toward developing the technology needed to preserve audiovisual media and
make it available online. The Euromedia project grew out of discussions at the
FIAT-IFTA conference in 1994, and received funding from 1996 to 1999. The
problem was that broadcast archives are documented at the programme level,
and online media really needs to be accessed in smaller units - at the shot level
for video. So Euromedia promoted concepts and a workflow now taken for
granted: ingest of conventional media to a computer hard drive; automatic shot
detection and keyframe extraction; presentation of these results as a story-board
for manual documentation; encoding of proxy versions of the video at reduced
bitrate; delivery of keyframes (and catalogue text) in response to a search; click
on a keyframe to see the video.
The follow-on from Euromedia was Amicitia (2000-2002), which has a website17
(Euromedia was too soon!and concentrated on the use of online media in TV
production, with a European perspective. It looked at how to use a common
catalogue, with multi-lingual searching, by work teams across Europe: UK,
Germany, Austria, Netherlands. Projects since then have addressed a
constellation of issues: the manual effort needed in indexing, and standards for
metadata (Echo18), the general development of European digital libraries
(Delos19), image restoration (for video and film; Aurora20, Brava21 and
Diamant22), audiovisual preservation (for broadcasting: Presto23; for film:
First24; and for all audiovisual collections: PrestoSpace25). Related projects of
significance are: Sepia26, which showed that non-specialists could be trained in
photographic collections management and preservation; Erpanet27, which is still
active and provides information and training in all areas relating to digital
preservation (preserving things that are already in a digital form); Birth28, which
is the first project to collect TV material from across Europe for a public-access
website. It has faced, and solved, many issues seen as preventing public access to
broadcast material.
Changes in Legislation, Policy and Attitudes
Today's small online collections, and projects like Birth, show what can be done.
Yet all current online collections, in all media and for all purposes, still show
only one particular kind of media: the kind with minimum copyright problems.
Real change requires shifting the copyright boundaries, and their interpretation.
Significant changes are occurring, in the areas of copyright legislation, Creative
Commons and common repositories.
In general, the period of copyright has been extended, in the US and in inter
national legislation. This has protected certain significant works, specifically the
Disney Studio's copyright in Mickey Mouse. It has also greatly increased the
number of 'orphan'works: media which has no identifiable copyright holder, so
cannot be legally used - or even copied for preservation - but is deteriorating on
shelves for lack of investment in preservation. The problem is that interested
third parties would make such investment - but would want some form of access
in return. Under current law there is no way to provide access to these orphans,
and so they wither in obscurity. This issue is taken up in greater detail on various
campaigning websites.29
A novel and rapidly growing approach to rights is to create positions between 'all
rights' and 'no rights'. This approach encourages rights owners to make material
available to the public under specified conditions - to donate their works to the
'creative commons', knowing that they can still have legal control over allowed
and disallowed usage. The Creative Commons30 has been the single most
productive force for enlarging public access, and has been used by many persons
and institutions, including the BBC.31
This paper cannot go into the complexities of digital preservation, which has its
TOEGANG
13 http://www.library.ucsb.edu/speccoll/pa/cylinderstech.html
14 0.7 GB for a CD, 200 GB for a high-capacity datatape
15 httpwww. archiveorgwebpetabox.php
16 Dietrich Schueller, Phonogrammarchiv, Austrian Academy of Science, Vienna, personal communication,
June 2004
17 For Amicitia see: http://amicitia.v2.nl/amicitia/, http://www.amicitia-project.de/
18 For Echo see: http://pc-erato2.iei.pi.cnr.it/echo/
19 For Delos see: http://www.delos.info/
20 For Aurora see: http://www.ina.fr/recherche/projets/finis/aurora/resume.en.html
21 For Brava see: http://brava.ina.fr/
146
RICHARD WRIGHT ACCESS TO AUDIOVISUAL ARCHIVES
- NEW METHODS
22 For Diamont see: http://diamant.joanneum.ac.at/
23 For Presto see: http://presto.joanneum.ac.at/index.asp
24 For First see: http://www.film-first.org/first/
25 For PrestoSpace see: http://www.prestospace.org/
26 For Sepia see: http://www.knaw.nl/ecpa/sepia/home.html
27 For Erpanet see: http://www.erpanet.org/
28 For Birth see: http://www.birth-of-tv.org/
29 http://cyberlaw.stanford.edu/about/cases/kahle_v_ashcroft.shtml; http://www.boycott-riaa.com/
http://www.tompaine.com/feature.cfm/ID/6101
30 http://creativecommons.org/
31 www.bbc.co.uk/pressoffice/pressreleases/stories/2004/0 5_may/26/creative_archive.shtml
147