By Seo-ho Yun (Ruudi@inven.co.kr)
This interview was originally published on Inven
*In Korean subtitle of “Avengers: Infinity Wars” Nick Fury’s line “Motherf...” was translated into “Mother”
The mistranslation in the latest Avengers movie has been a hot issue in South Korea recently. Due to this mistranslation, the translation industry has been highlighted anew in the contents industry recently. In fact, there has been a problem with translation not only in movies but also in games serviced in South Korea. The mistranslation makes fans misunderstand the meaning of contents or understand them in a completely different way. Furthermore, in some games, users criticized that voice actors and characters are not matched at all. Due to these kinds of incidents, users started to check the quality of translation and localization of foreign contents more thoroughly.
On the other hand, the demand for localization in South Korea is increasing due to the increasing export of Korean games and contents. Some companies have their own localization specialists, but localization works are usually done by localization agencies. Logrus IT is a localization agency participating in Play X4.
Based in Philadelphia, USA, Logrus IT has been engaging in localization for 25 years now from small text translations to large projects by cooperating with other branches including local dubbing projects. I met with KyoungYeon Won, the Korean branch representative from Logrus IT Korea who was in this Play X4 B2B booth to hear about the translation and localization process in various languages.
Director of Logrus IT Korea KyoungYeon Won
Seo-ho Yun: Can you please tell us about Logrus IT briefly?
KyoungYeon Won: It specializes in localization and it is based in Philadelphia. The main products that we've localized recently are Ubisoft’s “Assassin’s Creed Origin” and “Far Cry 5.” We not only worked on translation but also participated in dubbing and recording projects.
There are branches in 9 different countries, one of them being South Korea. The Korean branch was established in September last year. We look for domestic clients and perform localization of contents and games according to clients’ needs. We do not limit ourselves to games only; we also deal with car ads, airplane ads, etc.
Logrus IT has performed localization with various firms
Seo-ho Yun: You said you want to retain Korean clients. Which countries do Korean firms usually target?
KyoungYeon Won: As many of you know, many companies want to localize their products to Japan and China. In addition, the number of companies setting their sights on SE Asia has recently increased. They think it is easier to access the market there.
Nonetheless, it seems that companies favor Asian markets too much these days. It's not like they don’t completely exclude other options, but they just don’t consider them too much. It's my opinion, but approaching other markets is recommended when approaching localization. The world is wide, and there are various markets in the world.
I think the reason China and Japan are the number one option is that it is easier to localize directly. It is easier to find language specialists for Chinese and Japanese who are knowledgeable in the field. The same goes for English, too. Since it is not easy to find language specialists for other markets, companies just avoid taking even the first step.
In addition, I guess many companies think that direct localization is the only way to localize. I want to say that there's a way to localize using anchor languages. For example, English can be an anchor language for Korean-English-Russian localization.
Seo-ho Yun: Which methods are often used for localization and translation in Logrus IT?
KyoungYeon Won: First, we use direct localization. In other words, directly translating or localizing each source language to the target language. In Logrus IT, each local language must be handled by the corresponding local branch. This principle is a key to our localization projects.
Note, however, that we often provide localization services using anchor languages. As I said earlier, local languages are handled by the corresponding local branches, but there are languages that can’t be translated, or clients may want to localize into multiple languages. In that case, we use English as an anchor language.
Our method uses translation based on the anchor language and communication with the branch employees who use the local language. We can check the quality of our translation through this. This is how we minimize the problems when we use the localization processing applying an anchor language.
In addition, this review process allows us to ensure consistency of terms and double-check the entire process. We review multiple times using this method.
Seo-ho Yun: How many translation specialists are there in the Korean branch?
KyoungYeon Won: The Korean branch hasn’t been out there for a long time, so we only have two. Therefore, we usually translate projects in English, and then change to other languages. Note that we send Chinese-Korean and Korean-Chinese translation requests which are made very often to the Chinese branch.
Seo-ho Yun: In fact, many people think of translation first when they think of localization. This is because the localization of the written language is heavily weighted. Still, you mentioned dubbing. Can you please elaborate?
KyoungYeon Won: Dubbing in the local language is a big part of localization. The headquarters heavily emphasize this part of the operation. Thus, the headquarters have its own studio to perform recording and hire professional voice actors.
Logrus IT's sound recording studio
Seo-ho Yun: Do Korean companies request test translation only, or do they also consider dubbing when they make a request?
KyoungYeon Won: Case to case. Some firms only want text translation. Some firms want the entire process. We sometimes even do localization marketing.
It doesn’t apply for every content, but it's crucial to understand the flow of game translation, because the translation won’t be able to deliver clear messages to users. It's important to understand the overall flow and build-up based on the foundation of the flow. It’s the same for users in every country.
One more thing. There may be issues with fonts and visuals when translating into foreign languages. For example, Arabic has a different way of reading as well as how fonts appear compared to other languages. Most languages are read from left to right, but not Arabic. Thus, we need to do things differently for Arabic. Sometimes things crash during encoding.
In case of text codes, they often crash, don’t they? Moreover, when you change to a foreign language, fonts are applied differently for each letter, which causes graphic distortion. It’s not for Arabic only. This is one of the difficulties in the localization process. Logrus IT has also resolved these kinds of issues as a solution for localization, and it also provides technical support to ensure that translations are applied to games and that they run smoothly when they are compiled.
Seo-ho Yun: I'd like to ask you about dubbing. When you localize, don’t you tend to be more biased toward “translation”? I wonder what Korean companies’ approach to dubbing is in the process of overseas localization.
KyoungYeon Won: I think there’s some prejudice about dubbing. For example, when we promote dubbing in Korea, we emphasize celebrity, a famous person. I think that’s the reason why people think it’s expensive. Oh, of course, in the case of foreign countries, it’s obvious that when a famous actor or a voice act participates in dubbing, they publicize. That cost is beyond imagination (laughs).
One thing to point out, however, is that celebrity dubbing may attract users at first, but it's hard to hold their attention continuously. After all, in gameplay, whether the dubbing matches the character is the first thing to consider.
When recording, the game music will be the same; but if the company works with several agencies, the price will be steep. For example, if we work with studio and music company separately, we need to pay each of them. The bigger problem is that when you make corrections, you need to call the voice actors separately and borrow studios separately, which hikes up the cost further. I would like to say that people need to think about the cost as well. If the company is equipped with all that, then the overall cost will be reduced (laughs).
As a cost-saving tip, when developers request voice recording, people tend to do it every time the draft on each character is completed. Instead of that, in a state with a certain range of settings, I think it’s better to organize voice recording more systematically through consulting or something like that. Even if you do not consult, I want people to know this one point. Voice actors can make various sounds more than we think. Especially for those who are professionals, the variation is good enough to wonder if he’s human. Of course, those professionals are expensive, but they can make various sounds; in my experience, it’s sometimes more cost-effective to assign as many as possible to them than to give those to many voice actors.
The reason for the savings is that, as I mentioned earlier, the effort, time, and cost of recording and editing by renting studio are reduced. That’s how the effective work is done. As you may know, when work is done exactly according to time and plan, time and cost are reduced. With that in mind, I would like to say that people can proceed with the dubbing even at lower cost.
And one of the reasons for the high cost is post-recording editing. For example, there would be additional cost if you request recording again because you don’t like the voice of some characters. To minimize such occurrence, it’s important to give as much information as possible at the planning stage. That way, the dubbing is done befitting the intentions. Note, however, that this part is overlooked a lot in foreign language dubbing. I would like to take this opportunity to point out that this part is also important in foreign language dubbing.
Seo-ho Yun: The cost issue must be mentioned. In fact, if one can offer a variety of services and have the corresponding workforce and equipment, I think companies may think that it might be costly.
KyoungYeon Won: This is where we approach very carefully. But I can tell you that Logrus IT was able to survive for more than 20 years because it was able to meet the needs of customers. The solutions are broken down so that customers can pick solutions that suit their needs. In other words, the cost depends on the solution chosen.
And to talk more about the cost issue, hiring a freelancer for a single language will certainly save you money, although there’s a disadvantage of the quality depending on the freelancer’s ability. Still, translating each work with different freelancers also has a disadvantage of taking a lot of time because the work may not be in good flow and the decision process will be done separately.
On the other hand, in our case, we have branch offices, and the decision-making system with the branch office is compressed. Thus, when we translate into various languages instead of one or two, we have the advantage of it being done with more prompt feedback. I think time and money are saved in this part.
One more advantage is that domestic vendors often request secondary translation to other local vendors for Russian localization. We are able to offer lower rates as we can contact office in Russia directly. I personally believe that it would help to take note of this as we assess the potential of the East-European bloc and Russian localization highly. At Logrus IT, we assess the marketability of Russia and the East-European bloc highly as well.
Seo-ho Yun: That's something I want to hear more about. In fact, what should I say about Russian and Eastern European countries… Well, there’s an interest, but it seems there’s some distance.
KyoungYeon Won: When I meet Korean clients, I realize that. We have headquarters in the United States, but the Russian market is so big that we have done a lot of work related to it. When I had such experience many times, I saw a lot of possibilities regarding the eastern region. In addition to Russia, Ukraine and countries under the influence of the former Soviet Union are also really interested in games.
If we think about it, isn’t it Russia that made Tetris? In addition, there are more technicians in the software field. Surprisingly, many games are made in Eastern Europe.
I don’t know if it’s well-known in Korea, but in fact, they are holding many events like game shows in Eastern Europe. What I realized when I got there was that there’s definitely a potential. Some have asked whether the infrastructure is lacking. And some wondered if the Internet works well. Some are scarce, but these days smart phone games are a big deal, aren’t they? They had quite a lot of smartphone users there. And when countries where the infrastructure of the previous generation is not available to learn new things, they skip that step quickly and accept the next generation of infrastructure. Eastern Europe is a bit like that. Personally, I do believe that they are ready to accept it as mobile rather than online or PC stage.
Of course, this is only my view, so I'm going to tell you again that it may not be correct. Note, however, that Korean companies tend to look only at China, Japan, and Southeast Asia when they think of the game market. From that point of view, why not try to look at a new market? In that sense, I would like to say that the Eastern Europe market is worth a try. Of course, we have a certain degree of network or translation service in Russia or Eastern Europe (laughs).
In fact, we have corporations in Ukraine and Kazakhstan besides Russia, and we work directly in each country. We work in cooperation with each branch office instead of alone, so if you are interested, please contact us. We offer a variety of solutions besides translation (laughs).
Logrus IT also participated in the Russian localization process of “Assassin's Creed.”
Seo-ho Yun: This time, I would like to ask about how you are looking at localization in Korean from abroad. But recently, more and more games are imported in already Koreanized state. You have been working in an overseas localization company. I wonder what overseas markets think of the Korean localization market.
KyoungYeon Won: Of course, Koreanization has been going on in the past, but I feel that overseas markets are paying attention to the potential of the Korean market. Previously, there were not that many start-ups like now, and there were few attempts to debut. The domestic situation is not known for sure, but at least I think overseas markets feel like that. And companies interested in overseas markets increased. On the other hand, the foreign market started to pay more attention to each Korean content.
First, Chinese companies are interested in Koreanization. Some time ago, certain Chinese companies thought that they saw a lot of possibilities by turning the prototypes into Korean and seeing user feedback. In addition, overseas companies are increasingly aware that Korean language is important in the Korean market. Beyond simple gameplay, I think customers are increasingly focusing on the story, translation and localization. As customers focus on the story, they see that the customers’ level of interest in translation and localization became higher.
In fact, looking at the old Korean game market from the outside, they thought that the Korean market focused on gameplay rather than the feeling of great discomfort about language. But now, the way users approach it is different. They emphasize immersion. And the way to provide such immersion is, of course, familiar localized content. It means that the translation must be very complete. It seems that users are now demanding high-quality translations. Stories, plot lines -- to make all thing fused smoothly, high-level translation is essential.
Overseas clients are now approaching Koreanization with such awareness. They don’t simply translate or roughly transcend but take it more seriously. Users are also giving a lot of feedback about the translation. They post to the community that the translation looks a bit erroneous. When I see such things, I realize that there’s a lot of interest in localization in Korea.
Seo-ho Yun: In terms of localization of overseas contents to the Korean market and localization of Korean contents to the overseas market, is there any approach that you want to recommend to Korean developers when they localize?
KyoungYeon Won: When Korean companies do translation and localization, they tend to look for a direct result. Instead of considering the localization from the beginning, they show their work to us and say: "Please, do it by a certain date." At that stage, the first thing to consider is the cost savings. Their approach is how cheap it is as well as the costs that must be paid right away. And they should also consider how quickly you can get it out.
Of course, I understand that many companies requesting this kind of work cannot afford it. Indeed, it's important for indie and small or medium-sized developers to put out products first. In terms of the content industry, however, this approach alone doesn’t seem to help much in the long term. They should look at the future. I have a strong desire for a deeper approach to localization. I just wish companies have an in-depth approach to localization.
From the outside, what can I say about domestic users? I get the impression that they are focused on playing quickly. Should I say that they want to resolve things asap? They sometimes put it that way. Note, however, that there are many users in the overseas market who are looking for the details of the game or the elements hidden in various places. There is also a high percentage of users who enjoy stories and plots. In fact, in Korea, once they see the story, they consider it “good” and that’s that. But overseas, they have a high percentage of users reviewing stories one by one. Considering that, I think it would be great to invest a little more in overseas localization so that overseas users can immerse themselves more.
Seo-ho Yun: Of course, it would be nice to do that, but in reality, many small and medium-sized companies are complaining about cost issues and various problems. What solution would you offer in such situation?
KyoungYeon Won: Actually, we are a bit sensitive here, too. When talking about a solution, if I highlight a total solution or a bundled solution, or something like that, those people do not even consider it. They don’t even try it because they think it will be expensive since there are a lot of options attached.
But there are different kinds of solutions, and the costs are different. Even in our case, the most basic steps are not so expensive. We just introduce to users this kind of content. In terms of the number of characters, about 80,000 or 100,000 characters are not so expensive.
Plus, Korean companies tend to calculate the cost from a short-term perspective. There might be a lot of things going on here, but one thing I want to say is that translation/localization is not just a one-off process, especially online and mobile games. When a title is released, it's not the end. You need to keep it updated, right? You will need to manage this on an ongoing basis when updating.
Localization is done through translation-editing-QA process. If you’re developing games, you might be familiar with this. Note, however, that localization is perceived as a little out of place. If you’re a game developer, you realize that reducing the lead time in this process is the key to cost savings. Time and cost savings are always the developers' homework, and so is localization. If you consider the process to be out of alignment, the lead time will be longer when we go through each of these processes. But if you don’t go through this process, you won’t get satisfactory quality.
Seo-ho Yun: Some say it’s hard to translate Korean into another language and to express it with abundant nuances that complete localization is difficult. I consider these acts as a barrier which makes people think that they just need to convey the meaning.
KyoungYeon Won: There is such tendency. As a matter of fact, Korean is one of the difficult languages to learn for foreign countries. I have heard that foreigners personally think Korean is as hard, or harder, to learn as or than Thai.
But more important than that, it's a way of approaching localization. If you’re a translator, you probably have experienced it. They need a lot of context to express a term in some way or to interpret it with such nuances. In other words, it's the intent of the developer, or the whole context. When the localization is done with such data, expressions that fit the intention and which can appeal to the local people will be possible.
To provide advice to developers, when you give a script for translation or localization, it helps to localize when there is some commentary about the important parts of the script instead of simply giving the script. In the field, there are many cases wherein a 1:1 interpretation of words or concepts is not possible. But to find the corresponding concept or word, you need to dig it out.
It's like the basics of translation.
If not, it's a way to have feedback back and forth with us and correct it together. As our company has been doing this for more than 20 years, we have various solutions. We call it “spell checker” or “terminology” linguistically, right? We organize the terminology. And we classify how this can be used. We also had such system to apply in detail. What do I have to say, because I always emphasize these things? I think it gives people the impression that it’s expensive, so they avoid such approach (laughs). Anyway, I can tell you that I can provide you with the quality that you want, if you have this system.
Seo-ho Yun: Finally, what are some important things in localization that you want to point out?
KyoungYeon Won: I am saying this repeatedly, but translation/localization is not done in a single moment. As you go through each game and context, the game becomes more enjoyable, and high quality of localization will be realized.
After all, the important thing is to understand the context and refer to it. I think that the most ideal composition for that is to start consulting from the beginning and to adjust one by one. We’ve been working that way, and we've got the right solution for that.
It can be a burden. It costs money. Of course, even if we say it’s less expensive, sometimes it may not be the case. As such, I advise you to give as much information as possible about your intention and context. As I said earlier, 1:1 matching of terms, concepts and words does not work so well. Then you must choose a different word according to the context or consider other expressions. The quality of localization depends on how much you understand the context. Of course, you might think: “What does that matter?” When you get a foreign language script, you may not readily see it.
But I want to say that it’s meaningful when it builds up and accumulates, and it creates world view and frame. Moreover, as the service gets longer, the context is different when it gives a sense of immersion to the user. If the consistency of the terms is broken, the users are confused.
I think you’ll face many problems during the localization process. But there are a number of solutions in the industry. If you approach this with a little more interest, I believe you’ll get satisfactory results.