[Translating and the computer 10. Proceedings of a conference … 10-11 November 1988, ed.

Pamela Mayorcas (London: Aslib, 1990)]

Language conversion in the audiovisual

media: a growth area with new technical
applications and professional qualifications
Dr Georg-Michael Luyken
Deputy Director, European Institute for the Media,
Manchester, UK

INTRODUCTION
The European Institute for the Media is an independent research and
development organisation, concerned in the main with media
development on the European scale, including satellite television, cable
television and new electronic information delivery systems. Methods and
mechanisms of language transfer in the mass media are among the topics
researched by the Institute. In this paper I want to acquaint you with
foreseen developments in the media field over the next five to ten years
and their impact on language conversion and language transfer within the
mass media.
The Institute has undertaken two major studies which examine these
developments. The first, The future of the European audiovisual industry1,
is an industry study which tries to establish the likely development of the
European film, cinema, and television industries over the next five to ten
years. The other related study Overcoming language barriers in television:
dubbing and subtitling for the European audience2 is concerned with
dubbing and subtitling, voice-over and narration techniques as well as
their respective merits, costs, audience appreciation, and new technical
and editorial developments. To my knowledge this is the first research
and development project ever undertaken to explore the various methods
of language conversion within the mass media in a comprehensive
manner, and therefore breaks new ground.
I have divided my review into three main parts: first and very briefly, I
shall outline the main developments in television in Europe in order to
give substance to the assumption that within the media field, particular

136
Language conversion in the audiovisual media 137

demand for language conversion will increase dramatically over the next
five to ten years. Secondly, I shall explain the main methods currently
used for language conversion in television, with special reference to cost
and economics, audience reaction and future developments. Thirdly, I
would like to outline the emerging job and professional profile in the
translation field which will combine translation and journalism: the
journalist/interpreter or translator/media editor.

THE AUDIOVISUAL WORLD

Changes and growth in the audiovisual media

Television is currently going through its first profound transformation
since it was established some 30 to 40 years ago: a transformation marked
by two main characteristics. Firstly, the growing internationalisation, or
in the European context, Europeanisation of television, and the growing
commercialisation of television.
If we look at the European television landscape (Table 1) we see that
already some 46 television programmes are relayed via satellite, and are
receivable in not just one national coverage area, but all over Europe. In
1995, we will have 60 to 80 satellite transponders capable of delivering
television programmes, and this might rise in the year 2000 to 140-150
able to carry television programmes across Europe3. The common
denominator of all these satellite-relayed television programmes is, of
course, that they reach some 20 countries in which between seven and ten
major languages are spoken, in addition to numerous dialects.
A second development is the growth of the VCR video recorder.
Although Europe lags behind the United States in VCR penetration,
40 to 50 per cent of all households in the major European countries are
equipped with VCRs and the significance here is that people become their
own programme controllers. They no longer rely on what is transmitted
from national broadcasting stations, be it RAI in Italy, the BBC in the
UK, ARD in the Federal Republic of Germany or any of the other
traditional programme providers. People can either buy programmes or
go to the shop and rent whatever kind of programmes they want. At the
same time, and as part of this development, there is a big increase in the
demand for new and more audiovisual programmes.
These developments in television hardware and infrastructure result in
a considerable multiplicity of channels, which in turn, has severe
consequences for programme production. The crucial question at the end
is not how many satellite transponders we will have, but what kind of
programmes will be shown on them and where will programmes come
from? Again, in the European context that has major linguistic
implications.
138 Translating and the Computer 10

1989 1995 2000

Satellite channels available for TV 46 90-100 140-150
Households equipped with video
recorders 40% 75% 80%
Hours of TV programme transmission
per annum 275,000 350,000 400,000
Table 1. Growth in the audiovisual media in western Europe 3

The more television channels we have, the more fragmented the

audience becomes; the more fragmented the audience, the smaller the
revenue base for any given television programme provider to recoup
programme costs. At the same time we can already observe a significant
increase in programme production costs: home-produced programmes
have increased by some 300 per cent over the past 10 years; the rights to
broadcast the Olympics, football championships or other major sports
events have increased up to 1,000 per cent over the last three years;
programme imports, formerly commonly labelled ‘cheap American
imports’ are still cheaper than home productions but have also increased
significantly in price over the past few years.
This means that television programme providers can no longer rely on
a large income to produce their own programmes, but have to buy in more
programmes, or to engage in even more co-productions, co-financing
deals, or pooling arrangements with other partners in the increasingly
fragmented European television market. These movements are in the first
place dictated by the new economics of television, but they also have a
distinct linguistic component. A television operator will, in the first
instance, search for other partners in their language region: for example,
ARD or ZDF of Germany will look for partners in Austria and
Switzerland where they can easily produce a programme in a language
which reaches the whole German foreign market; the same is true for the
French, English, Spanish and Portuguese.
However, the available market for this kind of monolingual co-
production is becoming too small to generate enough finance to recoup
ever rising production costs. Hence, broadcasters as well as programme
producers, are looking increasingly to the European market. This has to
do with the reasons given before, but also with 1992. Here we are right at
the centre of the new importance of the linguistic element in European
television. A programme has to do with language, with content, and if
producers engage in co-productions with partners from other cultural
and linguistic areas, they have to somehow tackle the language problem.

Methods of language conversion in the audiovisual field

There are basically three methods of overcoming language barriers in the
audiovisual field, i.e., the cinema, television, video and of course, the
growing area of corporate videos: first, dubbing, or more precisely, lip-
Language conversion in the audiovisual media 139

synch dubbing; secondly, subtitling; thirdly, voice-over and narration

techniques. (See Figure 1)4
At present all three techniques are used in the various European
countries. There are typical subtitling countries, such as Holland and all
of Scandinavia which have been subtitling very successfully for years.
And there are the classic dubbing countries, such as Germany, France,
Italy and partly Spain where lip-synch dubbing is the common way of
transferring a foreign language audiovisual product. The reasons for this
division are partly historic, partly economic. Dubbing, which is very
expensive, is only cost effective if a minimum of 20 million television
households or more can be reached.
In the following, I should like briefly to discuss each of the main
language conversion methods: or, the mechanical process of a given
method; the inherent economics, that is, how expensive it is and how
much manpower is needed to produce it; and lastly, how viewers react?
Dubbing in television always means lip-synchronised revoicing which is
the most expensive of the language conversion methods available at the
moment. The process of dubbing a foreign audiovisual production
requires three main steps. First, a raw translation is prepared, which if I
may say so, is usually very sloppy. Often, these raw translations are
contracted out at low cost to students, or to people who happen to know
another language; professional translators are hardly ever used. The raw
translation script is then adapted to lip-synch requirements: if the
original voice says ‘ah’, you have to have an ‘ah’ or similar open movement
in the target language which matches the time and lip movement in the
original. This adaptation work is done partly by the dubbing director and
partly by the dubbing actors in a dubbing studio. The third step is the
final mix of the newly generated voice track which has been created in the
dubbing studio, with the original music and effects (M and E) track.
The soundproof, multi-track dubbing theatres needed for this process
cost in the region of £100,000. This capital expenditure comprises some
23 per cent of the entire costs involved in the dubbing process. Dubbing
actors themselves represent the highest cost factor, some 64 per cent: star
dubbing actors are paid very high sums. The technical side of the
dubbing process accounts for 13 per cent of the entire cost involved. (See
Tables 2 and 3.) One hour of lip-synch dubbing will cost anything from
£10,000 to £20,000. In view of the rising programme costs and smaller
revenue resources referred to above this kind of expenditure is, especially
for some of the smaller broadcasters in Europe, an additional important
factor.
How do television audiences react to dubbing? In the main, they love it,
because it is in their own language, and therefore easy to understand.
However, a dubbed programme only produces an illusion of the original:
it never gives the real soundtrack of the original film but only produces
the illusion of seeing the real film. Therefore, from a cultural point of
140 Translating and the Computer 10

Figure 1 The Two Main Language Conversion Methods and their

Subdivisions
Language conversion in the audiovisual media 141

view, quite a bit of concern has been expressed about having too many
dubbed versions.
Subtitling is a much cheaper method of language conversion than lip-
synch dubbing by a factor of about 1:15. The process goes through three
main steps. First, the registration and verification of speech in the original
voice track, or where necessary, transcription of the dialogue; second,
translation and composition of subtitles. The important point here is that
subtitles are never an accurate translation of a given dialogue, but a
reduction into key sentences or even only key words. Opinions on
subtitling vary, and depend on the context and nature of the original.
However, a subtitler should have not only excellent translation skills but
also the journalistic skills to reduce the often complex dialogue to short
sentences which can be printed in not more than 60 or so characters.
These must also be easy to read and understand in around four seconds of
exposure. The third step in the process of creating subtitles is the time
and picture insertion of the subtitle into the appropriate place in the film.
The facilities needed to produce subtitled audiovisual versions are a
central studio costing around £80,000, plus as many workstations as are
necessary, each equipped with VCR, television set and a personal
computer. Each station costs around £8,000 (See Table 2). However, it is
becoming increasingly easy for subtitling to be done at home with a
personal computer. Software is available and the hardware equipment
can be purchased off the shelf in most computer shops. Considerable cost
reductions can be made by using new equipment, in particular personal
computers and increasingly, remote workstations in the translator’s
home.
Dubbing Subtitling
Equipment Sound-proof Sound-proof Full-scale Each additional
multi-track outfit for subtitling workstation
dubbing voice-over unit with or set of home
theatre and narration 1 workstation subtitling
equipment
Approx cost
in ECU* 135,000 45,000 105,000 10,000
* ECU = European Currency Unit (approx US$)
Table 2. Investment costs for language conversion equipment in
ECU*
In the case of subtitling, equipment accounts for about 25 per cent of the
total language conversion cost. Interpreting labour accounts for about 60
per cent, lower than for dubbing, where the translation labour accounts
for 64 per cent. Technical labour accounts for about 15 per cent of the
total cost in both methods.
One hour of subtitles produced in the way described above will cost
anywhere between £500 and £1,300. Hence, it is considerably cheaper
than lip-synch dubbing.
142 Translating and the Computer 10

Lip-synch dubbing Subtitling

Equipment and facilities 23% 25%
Technical labour 13% 15%
Interpreting 64% 60%
Table 3. Cost structure in percentages of total language
conversion costs

What about the audience reaction to and appreciation of subtitled

versions? For the audience subtitles are, of course, much more difficult to
cope with. A viewer is primarily interested in watching the picture, but at
the same time will have to read the subtitles. This requires an additional
activity and is rejected by many people, particularly in the traditional
dubbing countries. On the other hand, extensive audience research in the
subtitling countries – Holland, Sweden and Switzerland – has shown that
subtitling is accepted by 80 per cent or more of the audience.
Longitudinal surveys which were taken in 1974 and 1984 in the
Netherlands show that the appreciation of subtitles has grown
particularly among the lower socio-economic groups, from around 50 per
cent in 1974 to well over 70 per cent in 1984.5
This leads to the conclusion that the common argument ‘the people
don’t like subtitling anyway’ is not necessarily true in this crude form.
There is a lot of room for experiment and a more flexible approach to the
cheaper and more cost-effective methods of language conversion for
considerable sectors of the audience.
The reason why we are likely to see more and more subtitling, however,
is growing economic pressure. There is reason to suppose that in the
dubbing countries especially Great Britain, France, Italy, Germany and
Spain, more and more subtitled programmes will be produced in the
future. This is, of course, also desirable from a cultural point of view.

Average Maximum Minimum

Subtitling 7402 1,957 2753
Lip-synch dubbing 11,0004 24,000 3,400
Voice-over 1,500 2,500 600
Narration 1,100 - -

Table 4. Units costs for one hour of conveyed TV programme in

ECU1

1
Including equipment costs, depreciated over three years and relative to the time needed to
produce one hour of conveyed programme.
2
On the basis of 36 manhours and 750 subtitles per hour of programme.
3
For non-electronic equipment.
4
On the basis of 15 working days by translator(s), sound technician(s), a dubbing director
and 10 dubbing actors, part-time as appropriate.
Language conversion in the audiovisual media 143

Future developments in audiovisual language conversion

So far it has been shown that a considerable increase in the demand for
language conversion can be expected in the audiovisual field as a result of
programme exchange among various European broadcasting providers.
Dubbing is very likely to remain the main method used, especially for
mass appeal programmes such as series and those to be broadcast during
prime time. At the same time, we will see more and more demand for
subtitling at the fringes of the programme end. This might eventually
extend to prime time schedules where an audience of sufficient numbers
could be captured with this method.

LANGUAGE CONVERSION MARKETS IN EUROPE

At this point it is useful to look at the main language groups in Europe in
order to assess the potential market size in each of these languages.

Language European Community Western Europe

English 59 59
French 58 60
German 60 72
Italian 60 62
Spanish 38 38
Source: The European Institute for the Media

Table 5. The main language groups in Europe (native speakers in

millions)

Table 5 shows that there is roughly an equal number of 60 million native

German, English, French and Italian speakers within the European
Community. However, if the Eastern European countries are included,
German is the most widely spoken first language, with English the most
widely spoken second language. Table 6 also demonstrates that there are
considerable number of adults who speak a European foreign language
and hence constitute market niche opportunities, aside from the large
linguistic blocks, which can be catered for in Europe with multilingual
audiovisual products. These are as yet little developed and will, no doubt,
have to be further exploited in the years to come. It is hoped that this will
also lead to a much more active exchange of audiovisual production in the
various European languages.
144 Translating and the Computer 10

Table 6. Adults who speak a European foreign language as a

percentage of the total population

THE LANGUAGE CONVERSION INDUSTRY

How is the language conversion industry organised at present? In some
countries there are special subtitling or dubbing units within each
broadcasting station, which are usually accounted for as a service
department within the post-production process. In Germany the
dubbing industry is highly oligopolistic, whereas in other countries, the
industry is monopolistic with very strong vested interests. These
industries tend to have close working relationships with the established
public service broadcasting stations and little space for development.
The future, however, might see many more independent companies in
this field. This is, of course, only speculation, but is based on the general
trend towards deregulation in broadcasting and a move towards smaller,
specialised companies or enterprises, which can provide high quality
support services in subtitling, dubbing, voice-over or narration
production. It might be just a matter of time – maybe five to ten years –
before we see these developments mature and take off as an audiovisual
support industry in its own right.
For example, the European Television Task Force under the
presidency of Valery Giscard d’Estaing, included in its report, Europe
2000: what kind of television?6 two recommendations:
6.1 Multilingual European television services working in the public
interest should be established. These services should be primarily
concerned with high quality programmes, drawing on the
production resources of several countries as well as with a European
news service.
Language conversion in the audiovisual media 145

6.2 the improvement of the technical, cultural and economic conditions

applying to language transfer should be given high priority.

This shows that there is pressure both from within the television industry
as well as in the political arena for supporting and strengthening the
audiovisual language conversion industry in Europe.
Conversely, there is a danger that if nothing is done to promote the
European audiovisual industries, many of the new television channels
will fill their programmes mainly with American imports. These will
keep translators busy, but only from English into Danish, French, Greek
and so on: clearly, a one-way flow of programme exchanges and linguistic
spread. What is needed is a more balanced exchange of culture between
the various European regions and a more flexible approach to the various
methods of language conversion in the audiovisual field.

TECHNOLOGICAL DEVELOPMENTS
Of course, computers and personal computers are already widely used in
audiovisual language transfer. Improved techniques for simultaneous
subtitling, similar to the Dutch Velotype keyboard and character-
generation system, will have to be developed. However, this technique is
unlikely to be fully and satisfactorily operational for some time.
One way to produce multilingual audiovisual products is to shoot the
production in various language versions. While this might only add some
three to five per cent to the total cost of major productions, it is still very
expensive, and not an option for existing programmes nor for most new
productions.
A recent Screen Digest Forum on these matters agreed that the ideal
technical development
would be a system based on technology whereby the writing of the
translated dialogue to match the lip movements would be rendered simpler
for the translator by being made more automatic. It was envisaged that with
voice recognition technology to detect and analyse the dialogue, a computer
holding a multilingual thesaurus would recognise voice patterns and try to
find appropriate foreign language equivalents which the translator would
just edit and tidy up .... Another potentially interesting technology is to
stretch the film or to compress the image (stretch-and-compression
printing) in order to achieve smoother dubbing.
While these and other techniques are currently being researched in
Europe and Japan, such development will take some 10 to 15 years to
mature, and will only be of limited application to the media, where
vocabulary varies widely, a vast array of fields are being covered, and
language is often colloquial, sometimes even slang.
146 Translating and the Computer 10

A NEW JOB PROFILE: THE MEDIA TRANSLATOR/

INTERPRETER
More immediately, the above developments are likely to create a demand
for a new professional job profile, the interpreter/journalist or translator/
editor who performs both journalistic, editorial and linguistic tasks.
Hitherto media translators have been:
- freelance translators
- programme-makers with language skills
- cinema translators.
Previous experience, for example the Eurikon experiments of the EBU
(European Broadcasting Union), has shown that these professional
backgrounds are essential but not sufficient for high quality performance
in future multilingual television. Therefore training for new media
translators and further vocational training for existing media translators
deserves high priority.7
So far language conversion, and those who carry out the tasks involved,
have contributed greatly to maintaining linguistic and hence cultural
diversity in Europe. It is hoped that this will continue to be the case in the
emerging European television landscape.

REFERENCES
1
Lange, Andre and Renaud, Jean-Luc. The future of the European
audiovisual industry. Manchester: European Institute for the Media,
1988 and 1989.
2
European Institute for the Media. Overcoming language barriers in
television: dubbing and subtitling for the European audience. Manchester:
European Institute for the Media, 1989.
3
Luyken, Georg-Michael. Direktempfangbare Satelliten in Europa –
Gegenwärtigen Stand und Faktoren der zukunftigen Entwicklung,
Medien Perspektiven, October 1987, 615-630.
Luyken, Georg-Michael. The VCR explosion and its impact on
television broadcasting in Europe, Colombia Journal of World Business,
Autumn 1987.
Lange and Renaud op. cit.
4
Luyken, Georg-Michael, In other words, Cable and Satellite Europe,
May 1987, 32.
5
Further details in Dubbing and subtitling for the European audience,
in Overcoming language barriers in television, Manchester: European
Institute for the Media, 1917.
Language conversion in the audiovisual media 147
6
The European television task force. Europe 2000: what kind of
television? Manchester: European Institute for the Media, 1988.
7
Translation, our future; in Proceedings of the XI FIT World Congress,
Maastricht: EUROTERM, 1988.

AUTHOR
Georg-Michael Luyken, Director, European Institute for the Media,
The University, Manchester M13 9PL.