Audio-video localisation

Tone and tune: giving your media the right voice

Audio and video localisation is needed everywhere, for example for films, series, e-learning and training videos, computer games, advertising and social media. Whether you require subtitles, audio description or video localisation, we have supported many clients and will provide the right solutions for you, including using AI.

What can you expect in audio and video translation?

Most clients require our assistance with audio transcription, audio recording, video localisation (including on-screen text) and video subtitling. We are also happy to localise complete e-learning courses for you from start to finish. AI is particularly helpful today for subtitling and audio projects. Clients send us a wide variety of texts. These often include presentations, training videos, marketing films and informational videos.

Here are two great examples of multilingual subtitling projects that we were able to carry out for clients. We provided the subtitles for an advert for the Canton of Grisons, which has been viewed online over 500,000 times. We also had the privilege of providing subtitles in Standard German, English, French and Italian for the moving documentary “Die fehlende Lunge” (The Missing Lung) for our client Swisstransplant. Click here to watch the Trailer with German subtitles.

When we talk about subtitling, the topic of spotting often comes up. Clients ask us how this works and what they can expect. These days, spotting and human transcription are carried out with the support of AI. The editing is done in tools that display the video and provide space for making corrections.

First, AI generates highly accurate timecodes and a rough translation of the spoken text. Then the humans take over: our language professionals carefully compare the draft text with the video. They ensure that the content has been transcribed correctly and that the translation is accurate. Then they check whether everything is easy to read and appears on screen at the right moment and for long enough. Often, the text has to be shortened, when the talking in the video is too fast, or when there’s simply too much text to keep up with. This requires real professionals, because it’s not easy to condense a statement while fully preserving its meaning. We therefore use the best combination of artificial and human intelligence – for faster, more efficient results with top quality.

And let’s not forget that video localisation is an important means of inclusion for people with hearing and visual impairments. When creating subtitles for the hearing impaired, for example, background noises are also transcribed. If you require any assistance or advice with your audio or video texts for these audiences, please do not hesitate to contact us.

What tools are used for audio-video localisation?

We select our software precisely according to your needs. Target audience, timing, tone of voice, image-text coordination and technical implementation – we take care of all these aspects for you. Here are some of the tools:

For audio, we use the following: Audacity, Ardour, oceanaudio, Adobe Audition, Adobe Sensei, Avid Pro Tools, Logic Pro, Descript, Sonix
For video localisation, we use among others: Adobe After Effects, Adobe Premiere, Final Cut Pro, Filmora
And for AI voices, we recommend: Amazon Polly, Resemble, Lobo, ElevenLabs, Murf AI

Expertise

Any questions about audio and video localisation?

Do you have an audio or video project, perhaps involving several different languages? Not quite sure how to get started or which solution is best for you? Please feel free to contact us. Céline Rodriguez Barbur, Head of Project Management, is always happy to advise you.

beratung@diction.ch
+41 81 750 53 33

What are the advantages and disadvantages of using AI in audio localisation?

Artificial intelligence makes many things easier in audio and video localisation, but not everything. It is crucial to consider which points are important to you.

Advantages of using AI in audio-video localisation:

Delivery times are significantly shorter
Costs are lower
Quick changes are possible if recordings need updating

However: the higher the quality required, the more expensive AI becomes. In addition, certain languages and complex texts are still difficult for AI to process.

Advantages of audio and video localisation by humans:

Natural voices are more pleasant because they better convey the required tone
Regional pronunciations are much better
Variations can be created easily when needed – whereas AI would require specific prompts

However: even though quality is clearly a plus, delivery times and costs are higher than with AI; updates and corrections take more time.

Who checks whether the audio produced by AI is correct and sounds good?

We select all voices carefully and perform meticulous quality checks, using (definitely😊) human professionals.

The voice selection for AI also includes a testing process that checks the pronunciation of difficult words
During the quality check, we examine pronunciation and intonation, including comparisons between the different languages
In addition, everything is reviewed once again by linguists (who were not involved in the process so far), who provide feedback

If you are unsure what is suitable for your audio project, simply contact us. We will give you honest advice – sometimes AI is worthwhile, sometimes it does not save time or money.