BBC SPORTS LIBRARY


         The Library was set up to catalogue tapes selected for Archive preservation in the mid 1970’s.
At the time there were only production log sheets from which the Department could retrieve
information of their holdings – which obviously meant going through piles of paper to find the required
material. Other Libraries in the BBC didn’t put Sport on their priority lists for cataloguing !

        The Sports Library, therefore, was set up by Sports Dept itself as an overhead to tackle the 10
years or so backlog of cataloguing and indexing of its video tape holdings. A card catalogue was
formed taking descriptive information from the production log sheets and indexing them by
Personalities, Teams, and Sports Subjects (by dividing the Sports up by their respective tournaments
– such as Olympic Games, World Championships, and World Cups etc)

         Eventually, a word processor was used to duplicate cards and the data generated was stored
on floppy discs (when they really were floppy!). By coincidence, when our physical catalogue was full,
the BBC had developed a retrieval system on its ICL main frame computer where it was possible to
recreate alphabetical card type indexing systems to retrieve text which otherwise wasn’t searchable.
The data from the floppies was added to the main frame and retrieved electronically again by
Personality, Team, and Sports Subject. As now, the cataloguing was done on a tape by tape basis, so
that a search on say George Best would reveal a list of tapes attached to descriptive content displayed
in chronological order.

Major Events:

         The problem was that this system was operationally slow in terms of cataloguing and indexing
and consequently unable to keep up with an ever increasing volume of video tape due to more
sophisticated production techniques. More importantly it couldn’t be used outside the BBC network. So
to provide an onsite stand alone system without the need for indexing, a system called Cardbox was
employed which simply required descriptive data input – each tape catalogued became a record within
the system. Searching was now free text, but it was essential that the database didn’t get too big and
that prescribed keywords were used within the text to limit the amount of hits obtained.

         This worked fine for about 7 or 8 years – but the problem was by doing so many major events
on site on a stand alone basis such the Olympics, Wimbledon, Football World Cups etc., we ended up
with about 26 different Cardbox databases which were formatted differently from each other according
to event, and incompatible with the main frame archive itself.

The Solution:

         We therefore needed a single database that could be added to on site and include the
catalogued archive that could be consulted on site. This needed connection to the BBC Network on
line. Cataloguing was still done on a tape by tape basis as before, but without the need for controlled
indexing. We now have, therefore, a web based system with free text, Boolean logic, and advanced
searching options housed on the BBC Intranet – but accessible on line from any site that has a
connection with the BBC Network. In this way, the on site cataloguing would go straight into the
Archive database, without it being held separately in yet another database. The archive from the old
main frame database and the Cardbox databases have now been uploaded to the New System (which
still hasn’t got a name !) and was used for the first time on an Olympic Games in Athens.


        Four Librarians were seconded to Athens to catalogue into this New System the tape
recordings made by the Production team. The whole production was based there including a studio
and a VT editing area based in the IBC. Programmes were a mixture of live and pre-recorded material
which were broadcast in the UK from roughly 0700 – 0000 UK time. Several hundred archive tapes
containing Olympic and other material relevant to the Olympics were despatched from London, the
content of which was accessible through the New Library System. Feeds from 40 different sources
were logged (or not !) by production staff on paper and handed to the Library staff for cataloguing as
soon as possible after recording. This would also include edited tapes used for transmissions. There
were also recordings made of “lesser” events from the different feeds in a “VT bank” which went
unlogged in detail but tagged by keywords in case needed by Production.

           As well as inputting as much descriptive detail as possible, it was vital for searching purposes
that correct metadata was entered – in fact descriptive text cannot be entered unless this has been
filled in. The metadata includes obvious things like the programme title and recording date – much of
which was preset in the system before the Games began. One of the most important pieces of data is
the type of tape recorded eg COMPILATION, EDIT, MAIN, etc. as this information determines to the
producer the “clean” and “dirty” factors they need to know especially when cutting Opening and
Closing Title sequences, montages, previews and reviews etc.

         A sign of the future was the increasing use of hard disc recording (LSM – EVS) which at this
stage was used for quick access to Opening Titles and other “great moments” clips. At this stage
however, retrieval to this material is not by our Library catalogue but by simple keyword tagging
available within hard disc systems. However, these discs were also used for continuous recordings,
but were intended for quick turnaround packages and not for a long term archive. At the moment,
these recordings, where necessary, are dumped on to tape for long term storage and only then
catalogued in our video tape catalogue. In many cases discs were recording the same feeds as tapes
– though for different purposes – adding to cataloguing and retrieval problems. Obviously old (tape)
and new (hard disc) technologies are converging, and the time has come when the latter will replace
the former, implying metadata and logging systems to be associated with the new hardware. I think I’m
talking here about “asset management” systems which are now being used in the BBC in certain areas
such as Natural History (pre and post production) and BBC News, but not yet for the archiving and
cataloguing of live and edited sport.


        In London a separate Library operation was employed to log/catalogue recordings of the
BBC’s Olympic output as it happened from 0700 – 0000 UK time. The BBC showed the Olympics
continuously during this period swapping from one terrestrial channel to the other as the day
progressed. The New Library System again was used and with a built in time code grabber it was
possible to log/catalogue the tapes “live”. Speed of cataloguing was therefore essential and possible in
the System as the there was no indexing necessary as in the older systems. With metadata and
keywords preset in synchronisation with the Athens team, the “live” logging was a great success. The
main problem was getting competitors names spelt correctly (there were 10,000 !) because
commentators’ words and on screen graphics come and go so quickly. But the Athens website was
very useful as we were able to cut and paste start lists and results straight into the System.

         The London logging was done through the keyboard of course – and in the “live” situation,
there is a limit to how much you can type in a given duration of time. So we experimented with Voice
Recognition using Via Voice in an attempt to log in far more detail than we could just using the
keyboard. The experiment was impressive though it highlighted the problem of time required for voice
training and the adding of names and sports terminology to the inherent vocabulary. Although due to
technical difficulties we weren’t able to do as much as we would have liked, the concept has huge
potential, particularly in the area of shot description (eg good close ups and reactions etc) and analysis
of team games such as Football and Rugby etc.

         In addition to the terrestrial transmissions on BBC1 & 2, 4 streams of interactive coverage
were also provided to digital viewers continuously through the 0700 – 0000 transmission periods. They
could choose which Sport they wanted to watch irrespective of what was being shown on the main
channels. These streams were also recorded for posterity on DV Cam tapes, but not catalogued in
detail (we had run out of staff!). Only keywords were put into a tape storage database to identify the
Sports recorded. All other tapes used were digibeta.


         With only four librarians in Athens and four in London coping with approximately 3500 tapes in
all, our New Library System was an obvious benefit over previous efforts. The fact that it was being
added to and read by the Librarians in both countries simultaneously obviously became helpful to the
Production overall. However, and as ever, technology is changing towards hard disc recording, editing
and eventually preservation. It’s vital, therefore, that a similar logging and retrieval system is added to
this new hardware and a centralised archive catalogue maintained. Hence the current interest in
“asset management systems” that can appease the needs of Production and the Archivist alike. But
how is the video archive to be stored? It seemed comfortable to catalogue something that had a
physical presence, but one doesn’t have the same affection somehow for a mass of digits held in


To top