Audio-Visual Translation and Translation Technology

Published on 01.09.2021

By Angela Starkmann – Opinions expressed in this article are solely those of the author.

This discussion about using technology for audiovisual translation (AVT) is old, and new at the same time. At present, the effective use of translation memory (TM) technology is really an issue in the exploding and technology-savvy audiovisual market where the time and price pressure is huge, and there is an evident lack of linguistic talent for some languages. The question is: Should (and could) translation technology (and TM memory technology in particular) be used for translating audio-visual content and “creative” content in general? Can we benefit from such technology to optimize our processes, improve performance, increase volumes and achieve faster turnarounds? If the answer is yes, then how do we go about it? And should we, as linguists, worry that such technology might make us obsolete even before our professional career comes to its natural end? This might still be treated as a controversial topic, so I’d like to make a few points before I discuss TM technology. This article is about audiovisual translation and the use of TM technology. Amazingly, the AVT segment of the translation market is still treated as very much separate from other translation segments in the market. When talking about audio-visual translation, the focus is on three areas: entertainment, training and marketing. This is usually not part of the regular workload for most translators. It is assumed that most translators typically work on technical, legal documentation, life-sciences and marketing materials, software, games and website localization etc. The general assumption is also that audio[1]visual translation is mostly handled by a different community of translators (and translation companies) and it is subject to different rules and the process is different. These assumptions are not necessarily true, and many translators and translation companies have already crossed over to this segment to supplement their current offering. Many others are contemplating this move. In most other translation and localization segments the use of translation memory (TM) technology (and I am referring to professional CAT tools in particular) is a given, but in entertainment we are still working hard to convince both clients and the translation community that technology (beyond machine translation) is also beneficial to everybody and should be used by all professional translation service providers.

In order to make the case for TM technology and its adoption, translation service providers need to understand the concept of ‘TM leveraging’ in order to understand how a CAT tool (e.g. like memoQ) can be used to support the audiovisual translation process. Leverage is how previously translated segments can be saved and reused for any number of reasons: think re-use of previously translated text (words, phrases or sentences), quick lookup for terms in a termbase or within the translation memory itself through concordance, standardized translation of units of measurement (as in distances, times, currencies), and many more. All of these ‘re-uses’ might be different for audio-visual projects, as they are structured differently than other projects, but the principle is the same. They can still significantly reduce the translation volumes and save you a lot of time in the translation process while at the same time improving translation quality. TM technology is particularly useful if you want to maintain consistency across many similar projects, and in case of audio-visual translation this is absolutely crucial if you’re working on a film series (for Netflix or Prime, for example). The actual usefulness of leveraging and QA (Quality Assurance) may differ across different TM tools and projects, but memoQ, for instance, combines a really user-friendly text editor with a robust QA functionality, and it can easily be used together with one of many available machine translation engines as well as voice-recognition tools.

I have decided to focus on memoQ here because I have been using this particular TM tool for years now and, apart from the features enumerated above, I believe that, as compared to other TM tools, memoQ is designed to make the translation process easier rather than more complex, the interfaces are so intuitive that it does not require any great effort to start using it from scratch, and it is great value-for-money. When I first moved into the audio-video translation segment I thought I might have to just use one of the tools advertised as specifically designed to deal with AVT, so I was really surprised when it turned out that memoQ was actually much better at dealing with subtitling than any of the specialized tools out there and I can continue to use it. This is important because the standard subtitling tools are very simple and do not offer much in terms of functionality. On the other hand, memoQ deals with subtitling and other aspects of AVT just as well as it deals with localization or any other type of translation out there. It is also operable so it can be used with any other type of tools seamlessly and file conversion works really well. Last but not least, my TM tool of choice is not only much better in terms of functionality than other CAT or ATV tools out there but also much more user[1]friendly. Additionally, you can easily customize memoQ to suit your personal work environment and projects. This is extremely important if you wish to reuse your legacy resources and manage different projects or similar projects quickly.

I have observed that in the audiovisual translation segment, as opposed to other translation segments where the users seem to be more technology savvy, most AV translators approach technology issues rather intuitively and I’m hearing a lot of: “I think that I am saving time using this subtitling tool, and it might be helping me work faster (by some percentage), smarter or deliver better quality than I would otherwise.” Audio-visual translators are hesitant to demand that technology actually help them work faster and better, and they usually think that their clients might use tools against them to lower their already very low rates or impose a certain process. They also often either blindly accept the tools imposed on them by clients, without providing those clients with alternative solutions, and they usually resist being part of the testing and developing process that might change their work environment and allow for the use of a tool of their choice. Hence, in many cases, AV translators are usually not part of the community of linguists who use tools for their own benefit. After attending a number of AVT conferences and taking part in numerous projects, I have to conclude that many AV translators are not very competent in technology (especially TM technology) and they do not actively seek solutions to technical problems. They usually believe that the subtitling tools provided by their clients and/or used by other AV translators are the best (or only) tools for this sort of work, and they sometimes believe that technology can negatively impact their specialization (which they deem to be creative, and hence different from every other type of translation) and damage their reputation as translators in this ‘artistic’ field.

For the above reasons, audio-visual translators need to check out the TM technology that has been used by specialist translators for years and determine for themselves if it is better than the simple editors offered for the AVT segment of the market. In other words, AV translators should use the same robust tools as all the other translators, and think in terms of their own convenience and benefits rather than hearsay. What we need are tools and workflows that save the linguist’s time and money, allowing them to enjoy the advantages of both TM and other technologies (e.g. MT). Professional translators have to embrace technology and work with state[1]of-the-art tools competently and confidently, and demand recognition for their professional competence and performance rather than expect praise for doing a lot of repetitive, time[1]consuming work for low rates. We also need to educate our clients so that the best tools are used by all the stakeholders and to dispel the perception that translation technology can soon replace professional translators.

The fact is that any kind of translator or translation project can greatly benefit from translation technology, whether it is in the field of art, marketing, or technical documentation. There are always plenty terms and phrases that a human translator shall select for the purpose of producing a specific target text time after time, and translation memory (TM) which enables reuse of previously translated segments, and terminology, which can be accessed with one click, can only make this process easier, faster and more enjoyable. The rejection of advanced technology just because we are scared our clients might leverage us and take these benefits away from us, or impose a certain translation process, and offer us lower rates is irrational and totally passé nowadays. If we can increase our productivity on some projects by up to 50% and improve quality at the same time, then even if leveraging might result in lower remuneration this is really not an issue for a professional translator. The benefits are even more obvious if we take into account all the useful resources which we collect when using TM technology and how we can utilize them in the future.

As part of audio-visual translation a huge amount of work is processed on behalf of many highly visible international clients and in many languages. This is a specific segment because of very fast turnarounds, need for great consistency across L-T projects and considerable price pressure, but not because it requires any unique technology. If CAT tools can be used for any type of audio-visual translation (and I have personally determined that memoQ can deal with any projects in this area) then this benefits both clients and linguists alike and there is absolutely no case for creating new, less advanced, technology for handling ‘creatives’ and dealing with the relatively minor technical issues associated with this process. I love to think of myself as “an augmented translator” (i.e. one that is in a unique ‘central position’ and uses any technology and environment at their disposal). I also believe that the uptake on technology should be much faster for AV translators—because this needs to happen if we want to adjust to the changing market environment and continue to be effective professionals.

I have used a few CAT tools but consider myself to be an expert on memoQ. So, in spite of the fact that I have been told by many colleagues that audio-visual translation requires special tools, I know that this is not true. Just as in case of website or games localization, memoQ is perfect for audio-visual translation, including subtitling. It deals with all typical file formats and offers reliable TM technology that has been used for specialist translation and localisation for decades. It combines robust functionalities such as a fully functional termbase and advanced for terminology management functionality, automated QA which can be customized to suit your needs using regular expressions, Live Docs which make the creation of TMs really fast and easy, concordance, doc preview, PM functionality and MT integration options for your preferred engine. This tool can be used for different processes and manage translations in many different file formats and according to any specifications. So, if you need to deliver your audio-visual translation together with any form of collateral material such as metadata files (e.g. content summaries and other info), this can be dealt with by reusing any content that had been previously created in the translation process.

AV translators should embrace established TM tools that are affordable and actually work for translation services providers of all sorts and this is the best way to move forward as AV translators community. This is the alternative to treating highly skilled and experienced workforce as working bees in a process that is becoming less and less controllable by the people who should really be in charge of it. If you use reliable and operable technology, all the players in the service chain can benefit from this and it’s clearly a win-win situation.

Audiovisual source texts (both subtitles and dubbing scripts) are not complicated by definition. They can be clear, simple, and well structured. They can also contain some untypical or specialist terms, long phrases, and lots of repetitions. There is absolutely no reason why CAT tools cannot support the AV translation process, at least to the extent where our work can be automated and managed, which is most of the time.

It’s best to illustrate any point by offering examples, so I will just enumerate some of the functionalities of memoQ to show how it achieves the results I mentioned earlier and I can additionally offer the following advice based on my personal experience when handling audiovisual translation:

  • prepare your source texts (e.g. subtitle templates) properly and make sure that you import the files into memoQ properly, so that the text is as short, error-free and consistent as possible, with no strange or mixed up tags, bad segmentation etc., before you even start working on the actual translation;
  • when working on any project, make sure you clean and consistently name and classify (according to projects/categories etc.) your TMs so that they can be easily reused in the future. For audio-visual translation this might involve not only using the client and project name but also, for instance individual series, genres or directors etc.;
  • use memoQ for consistency tracking (KNP tool) to help you bundle your resources properly and make information readily available;
  • remember that memoQ can really help you with translation of any recaps if the same terms/segments are reused in future—using TMs will eliminate the need to search for previous versions, manually copy and paste content, and will therefore make your work easier and more productive, while at the same time reducing the number of potential errors;
  • be aware of and use any structural rules that memoQ might be able to help you with (e.g. automated translation of units of measurements, names, dates, places etc.);
  • remember that you can use memoQ functionality to access resources and files that are being translated by other translators working on the same project and you can also communicate easily with other members of your translation teams;
  • project teams can work on the same files in memoQ in real-time so this greatly improves turnarounds, and because all files, translation memories, and termbases can be shared consistency between deliverables, they can be maintained without the typical QA challenges that usually happen when multiple translators and revisers work on the same project according to tight deadlines.

The above are just some notes and tips on how to handle audiovisual translation with memoQ to optimize the translation process. With this in mind, and some smart workflows and processes, subtitling can become yet another technology assisted process that will give us, creative translators, more time to focus on the real linguistic challenges and the truly creative parts of our work, and save us a lot of the typical hassle associated with using basic technology or no technology. Therefore, audiovisual translators should use really good and affordable technology that is produced by reliable and experienced technology providers.

To conclude, no matter what technology has to offer in terms of actual benefits and what tools we choose to use, we shouldn’t forget that we are in charge of our own process. We do have a lot of choices when it comes to technology (particularly which specific CAT tool we choose to use); therefore, as professionals, we should make rational and informed choices when it comes to the technology we use because this really impacts our performance. There is no point saying that translation technology is not useful or worry that it can replace humans or destroy creativity if we don’t actually know how to use it to our benefit. In this day and age, we simply cannot ignore that we need technology to be on top of all the changes happening around us and adjust as soon as possible to the requirements of the digital economy. The human factor in audio-visual translations will always be crucial, even if it involves editing some segments that are provided from TMs. It is really up to us to decide how we and our colleagues who work on the same project or similar projects will handle any translation, post-editing, or other services.

As translation service providers, we want to be respected for the quality of our work and our qualifications, and we want to be paid fairly for the professional work that we perform as expert linguists. This is what we need to focus on, and it means that we cannot reject useful technology and make our own lives more difficult. No matter what, translators will definitely continue to play a key role when it comes to creative or specialized translation, even if some of it might involve not only translating, but also editing content and making decisions as to the final product. Technology should simply be viewed as the environment in which our professional linguistic skills can really be used in an optimal way so that we can achieve our full potential.

Angela Starkmann studied Comparative Literature (Augsburg, Aachen, Toulouse, Waltham, Mass.) and became a freelance translator after graduation. She then worked as a translator for a major multinational medical devices company in the Netherlands for some years before moving back to Germany. She has been a freelance translator for several decades now and, apart from some specialized translation projects, today she mainly works as an audio-visual translator and translation technology evangelist. She is an enthusiastic memoQ user, and wouldn’t translate or edit a single segment without using this CAT tool. She is interested in new industry developments and how the translation process can be improved with translation technology. She is always happy to engage in discussions, present at industry events, and she writes for industry publications from time to time. Angela lives in Bavaria with her partner, also a translator, and Ginny (the office poodle), and can be reached on Linkedin.