Submitted by jpressman on Sun, 06/29/2014 - 16:19

Audio Localization from Films to Games

dub-blog
June 29, 2014

By Jacques Barreau, Dean of Dubbing | VP, Media & Interactive Entertainment Transperfect Media

 

A few years ago, at the very beginning of my involvement in game localization, I met with an executive from a major game company. He asked me: “How do I make my games sound like a theatrical movie?” I answered immediately, almost without thinking: “record it like a theatrical movie!”

My group at Warner Bros. (WB) had been dubbing movies in more than 25 languages for more than 15 years when we began game localization. This led us to approach this new challenge from a different angle than the rest of the industry. It appeared to me that applying our expertise in film dubbing to game localization was the most logical thing to do and, most importantly, it was what everyone in the game industry wanted at the time: high-quality dubbing for games. This may seem simple, but the film world and the game world were (and still are) far away in the localization universe, and this is where this new gig became really challenging.

Games appear more and more like movies. This is obvious when you look at the picture, but what about the sound? Currently, the sound for games is not at the film level, but it has made huge progress, especially in terms of sound effects. The main difference is the way the dialogue is recorded and how the audio is mixed and integrated to the game.

One can watch a localized movie three different ways: original dialogue with local subtitles, dubbed into a local language or voiceover, which still can be found in a few countries of Eastern Europe, mainly Russia and Poland. A game may have the spoken lines translated and appearing like subtitles, or the spoken lines can be rerecorded locally, which would be equivalent to a dubbed movie. The demand for dubbed movies across the world has been growing, and a similar demand for games is growing fast. This can at least partially be explained by the fact that young kids don’t read or don’t read well enough to follow a movie or game. Having to read subtitles and play at the same time would be too distracting.

The game world is growing very fast but like in the film world, before Dolby 5.1 and now 7.1, the sound has often been behind the picture in terms of quality and process. The media on which we watch movies or play games is usually not the same. Although games can be played on big screen TVs and big sound systems, the majority of them, at least for the kids, are played on smaller screens with one or two small speakers. The games will follow the same trend as TV: kids now want to watch what they want, where they want and when they want, and it’s unlikely that the majority of them will plug their consoles or iPads into a separate sound system. Most kids play on their way to school, in the bus or in the car. This is actually one of the big issues with the sound for games, just as it was for TV for a long time. On these small devices, even a great mix done in a theatrical mixing room will miss all the extreme audio ranges in terms of bass and high frequencies, and keep only the midrange audible. The fact that the human voice is located in this range (300 Hz to 3000 Hz) changed the way we have listened to movies over the years. In a theater, we can hear the full mix, but the same film watched on TV will miss most of the basses and highs. We became accustomed to this for many years, as what we want to hear is primarily the dialogue. Our brain can recreate sound effects and extreme frequencies as long as we can still hear even a very small part of these sounds.

The same way Star Wars created a huge jump in sound quality for movies at the time by using Dolby surround sound in theaters, audio for games has to do its own jump now that the visual effects have arrived at an amazing level of quality, detail and resemblance to a movie. The fact that players don’t play in theaters limits the sound, but home systems are becoming so sophisticated these days that the sound of a game now compares to the sound of a Blue-ray DVD.

When we started to localize audio for our games, we had two choices. The first one was to make the game studios try to record like a theatrical dubbing studio and the second one was to have our theatrical dubbing studios record games, which was a new adventure for everyone. We did both, but for the characters that already made it to the big screen, such as Superman, Batman, Harry Potter and so on, we recorded at our theatrical dubbing studios.

When you enter a theatrical dubbing studio, the first thing you notice is a big recording room with a big screen. It is much bigger than any recording room in a game studio. The reason is simple — when you record a film, you match what you see on screen. A big room with a big screen helps the engineer and the actor to not only record the voice but also helps capture the perspective they see on screen. One can argue that you can fix a lot of things in the mix, but perspective is very hard to recreate if it has not been recorded properly. When you record an instrument or a voice you need to record it with some space around it. If you feel that the voice is up close, right in your face, the actor was probably too close to the microphone and subsequently created what we call a proximity effect. A good recording should give you a sense of perspective in the space created by the speakers around the screen. I have spent (and I am still spending) a fair amount of time explaining to engineers how to record a voice, but it is true that sometimes the recording rooms are so small that you cannot entirely achieve this.

The game studios haven’t needed big rooms until now, as the actors had no idea what the perspective would be when his character speaks in the game. For this reason, most game studios have small recording rooms and record neutral voices without any perspective. Not only does the actor not know where the line will be, but the same line could be used in various parts of the game and consequently with different mixes and situations.

The first challenge in recording great lines for games compared to what an actor does when recording for a movie is that the game actor doesn’t see much. He or she looks at a computer screen showing lines in English and in the local language. To me, this is closer to reading than acting. How can you really perform and give life to a character by looking at a list of lines? We always guide the actors by telling them what’s going on in the game as we receive information from the producer. But being able to watch the character you dub on the screen makes a big difference when recording.

As previously mentioned, the recording for games is almost always done in small rooms. The actors follow the script on a Pro Tools session that has all the original lines. The actor listens to the original line and records the line in his or her own language. At this point, audio reference is almost a luxury since we don’t always receive it. The length of the recorded line is very important and must be the same in the domestic and the localized version, even if the actor doesn’t have to sync his or her lines to the screen. An actor dubbing a movie will not think about the length of the recorded line the same way, as he will watch his character on screen and lip sync his lines to the original ones. If the lip sync is good, the length of the recorded line will be the same as the original one.

For a film, after recording the dialogue, a local version will be mixed in the theatrical mixing room. There will be a big screen, big speakers and a lot of space to accurately recreate what we see on screen such as perspectives, panoramic moves, reverbs and so on. The foreign versions of games are not mixed the same way. The local dialogue lines are inserted into the game, which is not really a final product at this point but more a succession of scenes where it is very difficult to create continuity in terms of actors’ performance and audio mix. This is probably the biggest challenge in the localization of games. The way we dub games makes consistency difficult to achieve.

So, will we ever be able to control game mixes the same way we control film mixes? Some software developers are trying to address this issue and starting to propose systems creating mix states that can be triggered from within the game to totally change the environment in real time. I think this will always be the main difference between films and games in terms of audio. A film is a finished product and the mix is married to the picture. A game, on the other hand, is an interactive product and the mix will change according to the situations created by the player. If we really want the audio for games to match the quality of audio for movies, we would have to think about each situation the player can create and have different recordings and mixes for each of them.

 

Birth of a Brand

In the film and TV world, some characters such as the Looney Tunes with Bugs and Daffy, or DC Comics with Batman and Superman, go beyond their appearance. They have an attitude, a very recognizable voice (characterization) and some memorable key phrases, such as “What’s up, Doc?” They are used by marketing and merchandizing, they live in theme parks, they act in films and are now in games. At this point, they become brands.

In recent years, as the majors started to really push the big brands, we saw a lot of games being a continuation of blockbusters. More and more, films and games using the same brand are done in parallel and help each other in terms of sales and marketing.

From their birth in 1934, DC Universe characters now total over 5,300. Superman started in 1938 and Batman followed the next year. In 1990, Batmanthe video game was already following the plot of the 1989 movie. In 2005, Batman Begins was following a side plot of the movie released the same year. The game and film even shared the same title.

The first Harry Potter novel, one of our biggest franchises, was released in 1997. Within 14 years it was translated into 67 languages. Eight of the Harry Potter games correspond directly to the books and films, and were released to coincide with the films’ debut.

Our Lord of the Rings: Aragorn’s Quest from 2010 was a family game focusing on Aragorn. A year later, War in the North used both literary and movie influences from Lord of the Rings

Finally, what a career for Batman, who as an animated character became a huge movie star in the 1990s and started a new career in games with the very successful Arkham Asylum. Like our superstar Bugs Bunny, Batman is everywhere, from films to TV to games. As The Dark Knight helped Arkham Asylum to become a AAA game, it is also true that a new generation of players wanted to go to see the Batman movies.

Thus, localization goes beyond film or game. We localize an entire universe. BatmanHarry Potter, the Looney Tunes and the Lord of the Rings, to name a few, are obviously big brands for WB and all of them made it from the big screen to the game. So, for us, the localization has to go beyond the film or the game. The player, who was previously a spectator, has to continue the same experience he started with the movie in a more interactive way. As a passive spectator in the theater or an active player in the game, the experience must be the same. We need to ensure a smooth transition from the film to the game.

How do we do that? We start compiling glossaries with character names, locations, objects, vehicles, weapons and everything else that needs to be translated. We do that in 40+ languages using existing books, TV shows, movies and so on, but this is just the beginning of the process.

How would a player react if the voice of Batman is not the same in the film and in the game? If we want to have the film and the game in the same universe, the voices must match. This is my group’s expertise, but the voice is only one aspect of the recorded performance. We recreate the same ecosystem that was in place during the film dubbing by using the same directors and the same recording studios. On one of our Lord of the Rings games, we even used some audio files from the movie. In this case, matching the voices was not only a creative choice but a technical requirement. This is a complicated process, and it can also be costly, but we achieve fantastic results and the posts in magazines and blogs are unanimous: using the same cast and doing a theatrical-quality recording dramatically enhances the player’s experience. It is reassuring to know that the end users will see the difference and appreciate our efforts.

In some special cases, voices have to be processed. Finishing a new Batman game, we are processing the voice of Bane the same way we did in the movie. For the theatrical release of Batman, we centralized the foreign mixes and used the same voice treatment (a series of Pro Tools plug-ins) for all the different versions. We had to optimize the amount and color of the effect to the different actors’ voices (French, Japanese or Spanish languages don’t have the same dynamic) but the effect was the same and all these voices sounded similar. We are now doing the same thing for the game. As explained before, we apply the voice treatment without seeing what is really going on, but this is the best we can do to be as close as possible to the movie character.

Game localization vs Film Dubbing

From a strict line count standpoint, game localization can be (and often is) bigger than film dubbing. The reason is simple. A movie is a finished product that runs between 90 to 110 minutes on average. A script has approximately 120 pages with 60 lines per page. This represents roughly 7,000 lines. This is what we will dub. A game line count is much more complicated to evaluate, as you don’t know how many lines you’ll watch until you start playing. Your playing will trigger certain lines but not others. This doesn’t really matter for us, as we have to localize everything. This is why we have games with 20,000 or even 50,000 lines, numbers that you almost never see in a movie script.

Our big movies, especially the family entertainment ones, are dubbed in 25+ languages. The games are not at the same level yet and we’ll see in the near future if the number of languages increases — presently there are around eight.

Films are obviously seen by more people and are available in more locations than games are. For a variety of reasons, its difficult to compare film locations and game platforms, but we’ll have to reevaluate this in a few years. With interactive trailers now available in selected theaters, we can easily imagine that the game world will expand in the near future. How and where? This is the question!