作者：Alexander Murauski

如果你曾经参与过录音工作，不管是游戏的角色配音还是视频的旁白，你都会注意到这并不是一项便宜的的工作。为了减少额外成本你必须一开始便做好所有事。同样的道理也适用于音频的本土化：因为在这里任何错误都将因为不同的语言数而翻倍。在本文我们将分享如何结合录音与本土化服务，如何优化并加速这一过程，如何减少音频本土化的风险与成本等技巧。不管你是否使用了来自Alconost等公司的服务，以下的所有内容都将对你有所帮助。

Captain C-3PO by Jeff Nickel(from gamedev)

1.项目格式问题

适当的脚本格式能够确保录音过程的顺畅。一种最佳做法便是在表格中编写脚本，如此每一行内容便将对应一个音频文件。

在Alconost，我们有时候会遇到像顾客只希望我们去录下外文内容，他们自己将去同步音频叙述和视频。在这种情况下我们的顾客将需要面对它们并不了解的语言的录制，而我们则需要尽可能地为他们简化这项任务：

–在转录时我们会让编辑者（游戏邦注：以源语言为母语的人）记录下时间（指示每个阶段需要说什么）并将文本划分成不同部分（如每个句子在表格中都占据一个单元格）。这种转录本身便是“从原始录音中提取文本并将口语转变成书面语言”的过程。

–在翻译时我们会让译者（以目标语言为母语的人）保留内容原有的布局（如此我们便能够确保之前内容的每个部分对应着翻译后的每个部分）。

–在录音时我们会让叙述者将录音分解成不同场景（如此为每个新场景的录音将被储存在不同文件中）。

2.创造一列角色并描述他们的特征

如果你需要一些不同的声音，你便可以创造一列角色并简要描述他们，包括他们的名字，年龄，性别和其它特征等等。这将帮助音频工作室提供足够的角色叙述者选择，以确保基于不同语言将有足够的声优，并避免像让男声去录制女音的情况。不过有时候角色的性别不需要与叙述者的性别相匹配：就像拥有创造性声线的专业声优总是能够在假小子般的女生与优雅的女性间转换着。

为了让声优能够以最佳状态去呈现一个角色你就必须准确地描述角色的特性。例如：

角色A：男性，40岁，性格很好，很淳朴（是个农民）。能够善待自己（不会胡闹），对外人比较冷漠（尽管并不会很粗鲁）。他拥有一个雄厚的声音，但却并不刺耳。讲话语速正常。

或者：

浣熊角色：拥有尖锐，如孩童般的声音，但笑起来或私语时声音会变沙哑。他的语速很快，但通常都是可被理解的。他拥有5种类型的笑声（愉快的，有恶意的，歇斯底里的，有魔性的，很大声的）以及3种类型的恐惧叫声。

最理想的情况是顾客所给予的任务包含了一个描述了角色的截图和视频剪辑。

如果声优能够听到基于其它任何语言的角色录音，任务便会变得简单些。并且任何一种说明都是有帮助的。

如果叙述者忽视了我对于角色的说明该怎么办？

如果声优不清楚某个点，他们可能会向你提出问题（这是最佳选择，但是来回的提问与回答可能会延长截止期限）或独自作出决定（即按照自己所认为的答案为结果）。如果是第二种情况，而你并不喜欢叙述者所决定的版本，并且尽管形式标准让你感到满意，你也会对他们未事先征得你的意见表示抱怨。为此你可能需要额外花钱去进行新的录音，尽管声优总是会想办法满足顾客的要求。

3.限制声音的数量

通常情况下，比起最初的音频，人们往往不希望在音频本土化上花太多钱。因为声优的价格通常都不低，所以限制声音数量能够帮助你有效减少成本。例如你只需要两个专业声优，一个男声一个女声，也许他们便能够提供一个包含12个不同被访问者的视频画外音。

4.避免突出多个角色的音频文件

最理想的情况是，录音过程简单且快捷。所以你应该确保能够提供足够详细的任务内容。声优将记录他们的部分，并且使用本土语言的语言学家将检查这些录音，同时还要删除一些不必要的声音和杂音（游戏邦注：如翻页的声音）。录音内容将被划分到不同文件中，并且这些文件将根据特定内容去命名。

当你拥有一个包含多个声优的音频文件时，所有内容便会变得更加复杂。因为每个声优都是各自记录文件的，所以你需要在最后阶段从新划分并组合音频文件。而这将会提高任何情况出错的风险。毕竟编辑文件的工程师并不是很熟悉录音所使用的所有外文。

5.准备一份语音指南

当你大声读出文本时，你便会很快意识到有些单词的发音很不同。特别是那些缩略表达。外文总是会让任务变得更加复杂。

有些读音规则是通用的，但有些发音则需要使用公司做决定。你的工作室应该认真研究文本并列出所有可能存在差异的发音的内容。如此翻译者和编辑师便能在录音前明确这些差别。

如果录音有时间限制的话，你必须提前跟声优讲清楚这点。

6.在你的视频中留点空间

如果你对本土化很熟悉，你便会知道英语是最简洁的一种语言。如果你想将同样的文本翻译成俄文，最终出现的文本长度可能会增加10%，如果是翻成法文或西班牙文的话，长度则会增加20%。

如果你不能改变视频的长度，并且你不能加快语速或缩减最初文本的长度，你便很难提升本土化音频的速度去匹配视频的设定。出现这种情况的话你的用户将很难理解你的视频。此外，基于特别帧率的文本或行动有可能不能匹配本土化音频。这对于那些需要花更多钱去延长或编辑的视频来说的确是个大问题。

因此对此的最佳选择便是在你最初的视频中留出一些额外空间：在适当的地方添加几秒钟的暂停。这便能够简化并加快本土化过程。就像我们在Alconost为客户翻译或编辑文本时便总是会想办法根据短语的长度去调整单词的长度。如此将能够避免不必要的暂停或过于匆忙的配音。

7.提供大量的原材料

这是对方经常会要求你做的事。这种情况下你必须牢记：如果你希望配音的风格和基调与最初内容设定相一致，你就必须提供最初音频的样本。

当提到视频时，你必须确保你拥有源视频：基于这种方法你便能在本土化的时候加快或减慢场景从而保证动画能够匹配基于新语言的声优的语速。在这种情况下如果叙述者的声音能够区分于音乐和音效，你便能够在移动动画的同时有效去调整音效的设置。

此外，当提到视频时，除了声音外，屏幕上的字幕也必须进行本土化。如果你想要翻译这些内容的话，你便可以通过提供给对方视频源文件而让他们的工作变得简单些。

8.确保脚本与视频的有效匹配

这对于视频特别重要。配音通常都需要匹配书面脚本，但通常情况下你都是最后才去修改视频片段。所以便经常会出现音频和视频间的不和谐。为此你需要仔细检查最后的视频内容。

9.选择能够说外语且拥有音频本土化经验的专业人士

通常情况下都是由作者的朋友，作者本身或者那些并非以目标语言为母语的人士为文本进行翻译。如此看来你便有必要在配音前邀请一个以你的目标语言为母语的人去校对你的文本内容。的确有些声优在看到基于自己母语的文本带有语法问题时会拒绝为此配音。有些声优甚至会提供文本校对服务，这当然比什么都没有更好。不过最理想的情况还是招募你自己的编辑者（即以你的目标语言为母语的人）去校对非本地人所翻译的内容。如果翻译质量很糟糕，或者翻译内容改变了原意，那么比起校对你最好使用编辑服务（即重新编辑文本以符合最初文本的含义）。

外语通常都会加重整体录音过程的难度。当你在招募声优并确保他们能够与其他同事相互协作，创造脚本，决定配音技巧并从语言上去测试最终产品时，你需要非常重视音频的本土化。举个例子来说吧，如果你委托一个亚洲语言的本土化项目给一个工作室希望他们将其翻译成英语，你可能会发现在最终的作品中出现的是一个韩国声优而非中国声优，而对方可能会因为对目标语言的不熟悉删减了某些短语或使用了一些错误的搭配。

当你在选择一个服务供应商去本土化音频时，你必须确保他们能够使用专业的当地声优或编辑者，并且这些人已经拥有好几年的成功经验且获得了来自客户的正面评价。就像Alconost的团队便能够提供这种优质的本土化服务，即帮助你们更好地完成本土化工作并确保它们拥有高质量的标准，即使该项目所涉及的语言是较小众的语言。

（本文为游戏邦/gamerboom.com编译，拒绝任何不保留版权的转发，如需转载请联系：游戏邦）

9 Tips on Localizing Audio UNDER REVIEW

By Vladimir Kupratsevich

If you have ever dealt with audio recording, whether character voice acting for a game or a voice-over for a video, then you probably noticed that it is not cheap. It is important to do everything right the first time in order to reduce additional costs. The same thing applies to the localization of audio: every error is multiplied by the number of languages. In this article, we will share some tips on how to interact with recording studios and localization services, how to optimize and accelerate the process, and how to reduce the risks as well as the costs of audio localization. It does not matter whether you are ordering these services from Alconost or from another company — knowledge of all the following pitfalls will stand you in good stead.

Captain C-3PO by Jeff Nickel

1. Project Formatting Matters

The proper formatting of the script ensures a smooth and problem-free recording process. One best practice is to format the script in the form of a table, where each row corresponds to a single audio file.

We at Alconost are sometimes faced with situations where a customer only wants us to record narration in a foreign language, and they leave it to themselves to synchronize the audio narration with their video. In this case, the customer must work with audio recorded in a language they do not know, and we for our part must do everything possible to simplify this task for the customer:

When transcribing we ask the editor (a native speaker of the source language) to mark the timing (indicating what phrase should be said at what second) and to break the text into segments (for example, each scene occupies a separate cell in the table). The transcription itself is a process of “lifting” the text from the original recording, that is, of converting the spoken text into a written one;

When translating we ask the translator (a native speaker of the target language) to preserve the original layout (so that it is clear how each part of the original corresponds to a respective part of the translation);

When recording we ask the narrator to break the recording down into scenes (e.g., the recordings for each new scene are stored in a separate file).

2. Create a List of Characters and Describe their Personalities

If you need several voices, create a list of characters and briefly describe them, including their name, age, gender, and other characteristics. This will help the audio recording studio to provide an adequate selection of narrators for the roles, to ensure that there are enough voice actors given the different languages, and to avoid mistakes such as recording a female voice instead of a male one or vice versa. Incidentally, the gender of the character does not have to match the gender of the narrator: professional voice actors with a wide creative range can successfully voice tomboy girls and respected ladies, for example.

Describe the character’s personality in order to allow the voice actor to present the character in the best possible way. For example:

Character A: male, 40 years old, good-natured, simple-minded (farmer), “one of the guys.” He is friendly towards his own (though he doesn’t horse around), and he is curt to outsiders (though he stops short of being outright rude). He has a powerful voice, though it is not harsh. Normal rate of speech.

Or:

Raccoon character: high-pitched, child-like voice, but laughs and makes snarky asides in a husky voice. His rate of speech is very fast, but it must always be intelligible. He must have five kinds of laughter (good-humored, malevolent, hysterical, demonic, and loud laughter), and three kinds of exclamations of fright.

It is ideal when the customer’s assignment includes a screenshot or video clip depicting the character.

Melancolia by Kristina Alexanderson

If the voice actors could have the opportunity to listen to the recordings of their characters in any other languages, this would be plus. Any clarifications are welcome: “not as fast as in the reference clip,” “just like in the example.”

What if the Narrators Ignore My Descriptions of the Characters?

If the voice actors are unclear about a certain point, they will either ask you questions (this is the best option, but the need to exchange questions and answers can make it harder to keep to deadlines) or make a decision at their own discretion (what they think is right). In the second case, if it turns out that you do not like the version the narrator decided on, though the formal criteria have been satisfied, your complaint will be about something that was not stipulated in advance. In this case you may have to pay for another recording, although voice actors will often try to accommodate the customer’s request.

3. Limit the Number of Voices

Often people want to spend less money on the localization of an audio recording than on the original audio. Since voice actors usually have high minimum rates, so limiting the number of voices is a good way to reduce costs. For example, just two professional voice actors, one man and one woman, may be all that are needed to provide the voice-overs for a video featuring 12 different interviewees.

4. Avoid Audio Files that Feature Multiple Characters

Ideally, the audio recording process should be simple and fast. The assignment should be provided with sufficient detail. The voice actors record their parts, and the recordings are checked by linguists who are native speakers and cleaned of any extraneous sounds and noises (such as the turning of pages). The recordings are split into files, and the files are named according to the assigned specifications.

When you have a single audio file that contains several voice actors, everything becomes much more complicated. Since voice actors always record their files in isolation from each other, the audio files are divided and recombined during the final stages of the process. This increases the risk that something will go wrong. After all, the engineers who edit the files are hardly familiar with all the foreign languages in which the recordings are made.

5. Prepare a Pronunciation Guide

If you read the text aloud, you will soon realize that some words are pronounced differently. This is especially true of abbreviations. Foreign languages make the task even more difficult.

Some rules of pronunciation are generally accepted, but others must be determined by the company itself. Your studio should study the script and make a list of all the words whose pronunciation may present discrepancies. Translators and editors should clarify any difficult points before recording.

If a recording should have a time limit, then Captain Obvious suggests that the voice actor should be warned about this in advance.

6. Leave Space in Your Video

If you are familiar with localization, then you probably know that English is one of the most compact languages. If we were to translate the same text into Russian, it would increase in length by 10%, and if we were to translate it into French or Spanish, then it would increase by 20%.

If you cannot change the length of your video, then it will be hard to speed up the localized audio to match the video without accelerating the rate of speech or reducing the length of the original text. Both of these options can make it more difficult for your audience to comprehend your video. Moreover, the text or action in a particular frame may not match the sense of the localized audio. This is particularly problematic for videos that would cost lots of money to lengthen or edit.

Therefore, the best option would be to leave a little extra space in your original video: add pauses of a few seconds where this is possible. This will simplify and speed up the localization process. We at Alconost always try to adapt the word length to the necessary length of the phrases when translating or editing a text for our customers. This is how we are able to avoid unnecessary pauses or undue haste in our voice-overs.

7. Provide Plenty of Source Materials

You will most likely be asked for them. In any case remember: you must provide samples of the original audio if you want the voice-over to be done in the style and tone of the original.

When it comes to video, make sure that you have the source video: this way you can slightly slow down or speed up the scenes when localizing so that the animation matches the rate of speech of the voice artist in the new language. It is desirable in this case that the narrator’s voice be separated from the music and sound effects in the sound track so that you can simply adjust the placement of the sound effects as you move the animations.

And one more thing. When it comes to video, often the on-screen captions must be localized in addition to the text read by the voice artist. If you want to translate them as well, you will make this task easier for the service provider if you submit the video source files as well.

8. Ensure that the Script and the Video Match Each Other Exactly

This is particularly important when it comes to videos. The voice-over usually corresponds to the written script, but often changes are made to the video clip at the last minute. This is how inconsistencies crop up between the audio and video tracks. Carefully check the final video.

9. Select Professionals Who Speak Foreign Languages and Have Experience in Audio Localization

Often texts that are to be recorded are translated by the author’s friends, the author him- or herself or by translators who are not native speakers of the target language. All of this provides reason to consider that you should allow a native speaker to proofread your text before recording the voice-over. Indeed, some voice actors may even refuse to provide a voice-over if they see that a text in their own language is not grammatical. Voice artists may offer their own proofreading services, and they may be better than nothing. However, ideally it is worth it to recruit your own editor (a native speaker of the target language) to proofread a text that has been translated by a non-native speaker. If the quality of the translation is poor, or if the sense of the original has been changed or lost in the translation, then it may be better to order editing instead of proofreading services (when the text is edited, it is compared to the original).

Working with foreign languages adds complexity to the entire audio recording process. Localization of audio requires the utmost care when recruiting voice artists and ensuring that they are able to work well with each other, drafting scripts, determining recording techniques, and linguistic testing of the final product. For example, if you entrust an Asian language localization project to a studio that mainly translates into English, you risk discovering in the final product that a Korean voice artist was used instead of a Chinese one, phrases were cut off in mid-sentence, or other inconsistencies occurred due to lack of knowledge of the target languages.

When you select a service provider to localize audio, make sure that they employ voice artists and editors who are professional native speakers of their working languages, have years of successful experience working in this market, and have received feedback from satisfied customers. The team here at Alconost is ready to help you with your localization projects to ensure that they are localized to high quality standards, even if the project should involve less common languages.(source:gamedev)