From Script to Recording: Localizing Voice Overs for Video Games

Despite a strong presence of English content in global media, the art of dubbing is far from dead. From movies to videogames, the target audience for entertainment products would be limited without localized dialogue.
Dubbing VOs will help any title reach a wider public by enhancing the gaming experience for international players.

In videogames localization, some languages are referred to as ‘‘full loc’’ or ‘‘fully localized’’: the term highlights that the VOs of a game will be dubbed for a specific idiom as opposed to having translated subtitles only.

Localizing VOs takes its own space within the localization process: while it shares some steps with the rest of the translated content, it has its own best practices and quality assurance checks.

The foundation of any high-quality localization is accuracy. This applies to dubbing as well: translated dialogue sent to the dubbing studios should be reviewed and proofread to ensure not only accuracy of translations per se, but linguistic consistency and dialogue continuity as well.

A good practice is to have all voice lines at least proofread – if not directly translated - in chronological order; this step would identify any unnatural change in the flow of the dialogues or inconsistent vocabulary.

The details provided to both translators and the dubbing directors are also vital. Key factors such as a character’s gender, age, emotional state can change the translation and/or influence the dubber’s acting choices.
Similarly, the translation of a dialogue might vary depending on who the character is talking to.

A well-structured text database can significantly support the localization process and help maintain a consistent tone for each NPC and protagonist.
For instance, naming dialogue strings based on which character pronounces them ensures that they can be easily filtered and reviewed.

Audio validation (AV) is a step of the QA process aiming to ensure adequate quality of audio strings. It is not limited to translated dialogue only, since original VOs and ‘‘barks’’ lines - audio lines for characters’ sound effects such as panting, screaming etc. - can benefit from this QA process as well.

For the game's original VOs, lines should be tested for general audio quality (e.g; volume levels and clarity of recordings). Additionally, discrepancies between subtitle strings and the audio recorded should be avoided as they can disturb the gaming experience and player’s immersion.

In localized languages, the same criteria apply, along with identifying and correcting mistranslations that might have slipped through the first reviews, as well as harmonizing any discrepancy between VOs and subtitles.
The latter is a rather common occurrence: voice actors and dubbing directors might need to steer away from the provided translation due to dubbing limitations. When that occurs, the easiest route is to update the subtitle string in the database to ensure it matches the recorded line.

During the AV process, all issues should be systematically cataloged based on their severity and with clear description of the error at hand.
This approach aims to identify which lines require re-recording, and which can be accepted as they are.

Understanding the gravity of each issue helps LQA managers address the most critical problems promptly. Re-recording doesn’t just take money, it takes time: dubbing studios and actors must be booked, and availability is not always guaranteed on short deadlines.

After audio implementation, VO testing is done along with regular LQA work.
Even if all the steps described above are perfectly followed, this testing phase is still pivotal to ensure that dialogues are smoothly integrated into the game for the best player experience.

Conducting thorough in-game testing allows the LQA team to identify any continuity issues, sound discrepancies, or translation errors that may have slipped through – often due to the lack of in-game context in previous stages.

Natural playthrough is time consuming but worth the effort: it allows LQA professionals to witness the flow of dialogue like any normal player would, getting a first-hand feeling of any unnatural elements disturbing the experience.

Nonetheless, nothing makes an LQA tester more grateful than good localization debugs.

In regards to VO tools, developers should consider including ‘‘cinematic triggers’’ and ‘‘jump to dialogue’’ options in their debug rooster. Cinematic debugs should allow LQA experts to review the most dialogue and VO dense parts of the game efficiently by launching any cut scene without having to reach a specific area.

Dialogue debugs are particularly helpful when a game presents multiple dialogue options: a good debug would trigger all the potential questions and answers one after the other, allowing LQA testers to efficiently review all possible lines without of having to go through each different option manually.

A special mention should be made about how gender influences VOs dubbing and game localization as a whole – and it goes beyond having to adapt a translation based on the gender of the two people talking.
It is increasingly common to play games allowing players to choose the gender of their main character and provide 2 or more options to their audience.

This factor must be taken into account as early as possible in the localization process, especially if the title is being localized in gender sensitive idioms – like Latin languages, which are more gender specific than English.

The localized strings and VOs should use both gender-neutral language and gender-specific variations in dialogue strings to provide a well-rounded experience to all players, no matter their language or character creation choices. Otherwise, the risk is to serve players with inconsistent translations and easily noticeable errors that will impact negatively their gaming experience.

After this rundown of all the steps and requirements to ensure well localized VOs in videogames, one might be left wondering whether it is worth the effort. Aren’t subtitles good enough? The answer is no – not yet at least.

There might be a future in which all global audiences are so used to subtitles that the cost of dubbing videogames will no longer constitute a good investment.

That future is not here and might never come true: especially in games that aim to give an immersive experience, localized VOs play a crucial role in catching and keeping the players’ attention. The cost of good localization – dubbing included – is still a worthy investment to market a title to a bigger slice of the gaming audience.