In addition to the use of engaging narratives, the sense of realism in modern games has been enhanced by the use of audio, in particular the human voice for in-game dialogues, with more and more console games incorporating lengthy cut-scenes which may be subtitled and also dubbed in full localization. Accordingly, audio localization has become a major issue in game localization, reflecting a general trend of increased numbers of audio projects in translation requests to localization companies over the last five years. From a translation perspective, the full and regular incorporation of human voices in game systems made game localization more closely associated with AVT. Recordings by celebrities have become common, especially among movie-licensed games. This may add an extra challenge if any retake is needed at a later stage, both from the point of view of actors' schedules as well as the high costs. While the use of pre-rendered cinematic features began in the 1980s, technological limitations in the early days meant that such cinematic sequences were uncommon and mainly without human voiced dialogue. Ninja Ryukenden (1988) is often credited as a successful early attempt at using cinematics between different levels of a game to tell a story albeit without the use of audio. While cut-scenes in modern games have become much more sophisticated, their use of AVT modes is generally not well informed by the body of knowledge now available in this field. This is partly due to a lack of awareness but also because of some game-specific contexts.

Re-voicing in the form of dubbing, which is commonly referred to as "voiceover" or VO in the game localization sector, now forms an integral part of game localization. Subtitling in video games has developed largely independently of AVT. This may be partly due to the specific functionality of cut-scenes and other uses of voice in games not being considered central to the product as a whole or required by the gamer to engage in interactive play. However, this perception is rapidly changing, with audio occupying an increasing part of the game. The immediate difference between subtitles used in games and in other AV materials is the faster speed at which the subtitles in games are displayed. Also subtitles in games tend to be longer as the number of characters permitted in game subtitles is mainly dictated by the length of the original script and is independent of the viewer's (player's) estimated reading speed, which is prioritized in AVT. This may also be due to the fact that the subtitles used in a game can be paused and restarted by the player, although this functionality is also available to the viewer watching a film on a DVD. Game localizers are typically given the number of characters calculated from the length of the dialogue and instructed to translate script lines working off an Excel sheet often without visual information. This mode of working also tends to affect segmentation or what are known as“sense blocks" in AVT. It is not unusual to find a character's dialogue in a game segmented into two or even more subtitles in a manner which does not respect semantic units, further hindering smooth comprehension by the recipient. Despite the increasing resources put into the production of cinematic sequences in games and their multiple functionalities including more affective reasons to engage the gamer, the translation of cut-scenes has generally not been treated with sufficient care by game developers and publishers, and AVT norms are clearly not adhered to. However, this may change as more game developers and publishers are aiming for full localization.

As is well recognized in practice and in the literature, revoicing through dubbing is costly if it is to be professionally executed. For this reason, the dubbing mode has normally been available only in fully localized versions for territories which are considered to be of sufficient commercial significance. Translating voice scripts for a game is an extremely time-consuming and challenging process. We have cited several examples to show the increasing scale of the task of audio localization. A further illustration of the increasing size and the shrinking timeframe to achieve full localization is the aforementioned Final Fantasy XIII-2 (2011) involving over 18,000 spoken lines to be translated and voiced. Furthermore, this game was shipped in English, French, Italian, Spanish, Greek, Chinese, and Korean within seven weeks of the original Japanese release. This contrasts with an earlier title in the same series FFXII (2006), which took two translators nine months to translate from Japanese into English, working on the script alone (Smith quoted in Jayemanne 2009, n.p.). Under a sim-ship scenario the schedule is becoming increasingly tight while the number of words to be translated and voiced is expanding. Further challenges are posed by greater numbers of languages.

责任编辑：admin