Audio Production
Back to LAB main
Behind the Scenes: Voicing “Like a Dragon: Infinite Wealth”
Like a Dragon, formerly known as Yakuza, is a well-known Japanese RPG series which has seen much critical and commercial success since its debut in 2005. In Japan, the game is known for its star-studded voice cast. So when SEGA announced the addition of Mandarin Chinese audio for the eighth and latest entry in the series, Infinite Wealth, the game naturally become highly anticipated among Chinese players.
The momentous task of providing localized voice production for Like a Dragon: Infinite Wealth thus fell upon our SIDE Shanghai team. They were involved in the entire Mandarin Chinese voiceover process, from casting, recording, and directing to project management and post-production. What’s more, our teams at PTW collaborated in providing Localization and Localization QA for the game’s script for Chinese script in both Simplified and Traditional Chinese, as well as Korean.
To get a better idea of how the Mandarin VO was produced for the Like a Dragon series for the first time, we sat down with key members of our SIDE Shanghai team to learn about their process and dive into all the time and effort invested into preparing for this special project.
The Like a Dragon series has been around for close to nine years. Many of the Japanese voice actors have become revered personalities in the games industry and among players alike. Therefore, finding the best Mandarin Chinese voice actors for the job was undoubtedly challenging.
In Infinite Wealth, over 400 characters have voiced lines—among them, 53 are special characters that require specific vocal qualities, including characters from previous titles and brand-new characters. After multiple rounds of auditions, we successfully cast 84 professional voice actors.
To ensure that we would cast the best possible talent, we conducted extensive research by immersing ourselves in the series’ previous titles and carefully dissecting the plot and background of every character. Based on characters' personality traits and the vocal qualities of the original Japanese actors, we cast three candidates for each special character. Samples were then recorded and sent to SEGA to choose the most suitable actor.
For example, casting the actor for Yu Namba proved to be a challenge, given the unique vocal characteristics of the original Japanese voice actor. The character is a shabby, middle-aged man in the game, so we initially recommended several actors of an older age for the role; however, the SEGA team ultimately went with a younger actor whose vocal qualities more closely matched the original actor. In contrast, the casting for the main characters Ichiban Kasuga and Kazuma Kiryu went much smoother, with the Mandarin Chinese actors' voices, personalities, and characteristics aligning seamlessly with their Japanese counterparts.
Considering the complex character groups in the game, we also had to take second language proficiency into account when casting the Mandarin Chinese actors. Some characters required that the actor be fluent in English, or proficient in both Japanese and Chinese. There is even some language-mixing across some of the game’s language versions—for instance, the game features a group of Chinese characters called the Yokohama Liumang, whose voices we recorded, and can be heard in the English and Japanese versions of the game.
Definitely—the game's rich story features many character interactions. What people might not know is that the audio for each character was recorded separately. The lack of live interaction added an extra layer of complexity for the voice actors. For instance, the five most important protagonists in the game were recorded from scratch without any reference recordings, and they were the ones that set the overall tone for the other voice actors.
To ensure coherence, actors not only referenced the original Japanese audio and Mandarin Chinese recordings, but also studied the emotional context of other characters' lines. An interesting example is the recording of a scene between the characters Utamaru and Seonhee. The original recording schedule had the voice actor for Seonhee come into the studio first. Unbeknownst to them, there were many improvisations made during the recording of Utamaru that the director wanted to retain for dramatic effect. This meant that the actor for Seonhee had to return to the studio for re-recording.
Preparation and research were just the tip of the iceberg—the most significant challenge was providing authentic localization and replicating the same feeling of the Japanese VO for a Chinese audience.
In particular, syncing the audio between languages proved difficult due to the differences in expressions between Japanese and Chinese. For example, Japanese often has long strings of continuous interjections, while Chinese tends to be more concise. The rhythm and pauses in Chinese and Japanese also differ, especially with Japanese honorifics making sentence endings longer and certain syllables having a quicker pronunciation. Sentence structures also vary considerably, with some sentences being inverted between the two languages. Therefore, to ensure the dialogue fit well with the original audio, the script and lines needed careful refinement and polishing. Actors also had to add some extra performance elements to perfectly match the original script.
As an example, below are comparisons between some of the Japanese and Chinese voice lines from the game:
1. Original Japanese:
まずは余計なことを考えずに打席に立ちな ホームランは後からついてくるさ
Literal translation:
First, don't think of anything else. Just stand on the batter's box. A home run will naturally come later.
Adapted for colloquial expression and to match the line length:
Clear your mind and stand there. A home run will come.
2. Original Japanese:
アルツハイマーだそうだ。 57のはずなんだけどな。
Literal translation:
...She has Alzheimer's. She's only 57.
More natural with inverted word order:
She's only 57, but she has Alzheimer's.
3. Original Japanese:
お初にお目にかかります。 手前生国と発しまするはジャパンの東京・神室町。
Literal translation:
Nice to meet you for the first time. I'm from Kabukicho in Tokyo, Japan.
Translated based on context and to match original line length:
Greetings, everyone. Nice to meet you all. We've never met, but I hail from Kabukicho in Tokyo, Japan.
Thankfully, the collaborative work between the recording director and the PTW localization team allowed us to quickly overcome any challenges that came up regarding localization.
In particular, SIDE’s proprietary script management plugin “Tracker” allowed the team to quickly locate recorded lines and replay voice files, allowing for accurate and efficient modifications at any time.
Of course, and the karaoke scenes were some of our favorite parts of the entire production process—even if mapping the Chinese syllables to the original Japanese syllables was a challenge at times. To tackle the localization of the karaoke, we studied the melody, lyrics, and breaks of all the songs, providing voice actors with demo recordings as reference. For example, aligning the Chinese rap with the Japanese beat was a tedious task at times, where syllables had to be marked one by one and reassigned with Chinese characters. As a result, players may notice differences between the Japanese lyrics and the Mandarin Chinese voiceovers in the final version of the game.
While we were fully engaged in dubbing Like a Dragon: Infinite Wealth into Mandarin Chinese, the text translation and Localization QA for the game were also entrusted to our teams at PTW Shanghai.
As per usual practice, the voice acting team only receives the scripts related to the audio. This created a few problems early on since certain scenes were only recreated based on the lines provided. Fortunately, however, our colleagues from the translation team were able to provide us with quick assistance for any missing context. Many of them were avid players of the Like a Dragon series, so they were able to not only clarify a lot of background information about the game but also assist in polishing the Chinese script in advance to meet the strict timings of the original Japanese voiceover.
A heartfelt thanks also goes to our audio post-production team in Malaysia. The recording process was divided into fifteen batches, including cutscenes and files with strict requirements for synchronization. Each batch required meticulous inspection, recording, and archiving according to specific requirements. Issues are inevitable in batch processing, including missing files and mismatched lines. Fortunately, our team in Malaysia provided excellent post-production support. With a well-integrated system of spreadsheets, recording software and plug-in macros, issues were swiftly identified, solutions found, and accuracy ensured.
To hear some of the localized VO in action and to get to know our SIDE Shanghai team members that contributed to this project, check out our deep dive behind-the-scenes video below!