"Gnhhh!" ... "Whhhhaaaah!" ... "Bah….ah...gahhhhhh"... "Hmmmf!"... "Ehiehhh" ... "Mhaemm!".. These grunts, sighs, squeals and miscellaneous other vocalizations compose roughly 1/4 of the dialogues in the early hours of Final Fantasy XIII.
On one hand, they're to be expected; Japan is known for its plethora of exclamations and onomatopoeiae. On the other - when translated literally - they make for a poor localization.
These sounds are often louder and longer than their English counterparts, or they simply have no equivalents. As such, they're difficult to remove or replace and are usually left untouched. They've even become something of an accepted "quirk" among the more dedicated fans of Japanese media, but that doesn't mean they couldn't be handled in a friendlier fashion.
I think the localization team for FFXIII wanted to give Vanille a unique voice — much like the Björk-esque Fran in FFXII — but the voice actress' performance is a bit of a mess.
As things stand, vocalizations often come across as alien and awkward. They break the flow of conversation and the suspension of disbelief, and can leave a new audience feeling put off.
Sure, one can always argue for the purity and cultural authenticity of any given product, but that's being a bit of a stick in the mud. Literal translations lack context and social nuances, and those fully familiar with them might as well experience the original versions. In order to make the products easily digestible by a different audience, though, some things need to change. FF XIII in particular is a title Square Enix wanted to be a global blockbuster, not just a Japanese game released to a niche audience outside of its home country, so it stands to reason that they'd want to iron out these kinks.
So how can this be done?
A couple of points: If possible, simply remove the exclamations altogether. The ones that could easily be cut are left in to keep things consistent and speed up the localization process, so getting rid of them shouldn't be a big issue. Use local equivalents of the vocalizations if available. For example, make a character surprised by a hand on his shoulder utter a short "Huh?" instead of the original, "Mnhaaa?" Use actual words or sentences for sounds that have no local counterparts. A character crying out "Gwahhhhhhhhhhhh!" for three seconds after witnessing a car crash could easily be replaced with a quick "Oh my god!" Meld the exclamations into the speech itself. I'm not an expert, but I noticed many of the vocalizations were isolated within the dialogue, whereas in English they'd part of it, e.g., "Mmmm, I don't know about thaaaaaaaat." Finally, keep these points in mind when developing the game, and provide the team(s) with the tools necessary to port it. Automated lip-synching is already widely used, but I'm sure other functionality - or just the permission to alter the in-game cutscenes - would be appreciated.
Of course there are more issues to consider as well - perhaps toning down on the dramatic, clenched-fist poses with characters uttering such phrases as "I'll do my best!" - but those are a whole other topic…
Radek Koncewicz works as a videogame design consultant, and also as the creative lead of Incubator Games, a small development studio he started with a few former coworkers. In his spare time he writes for blog Significant Bits.