New concepts in voice testing for dubbing

At the beginning of the dubbing process, voice casting is one of our most delicate and interesting activities. People generally think that casting is just a matter of selecting the auditioning actor who can best re-create a given character. The reality is that the casting process is not that simple, at least not here at Warner Bros. Rather, casting is a multistep process. First, we must find an actor with the right voice potential. Then we must help that actor get as close as possible to the original voice, acquire the character’s attitude and deliver the best performance possible when recording in the studio. Finally, we must create a set of written guidelines to help the actor consistently find and deliver the right voice and attitude from session to session or episode to episode. These guidelines are based on a specific vocabulary designed to give a common reference to actors worldwide, teaching them the voice placement techniques that are so important in the re-creation of our animated characters.

Voice dubbing in the animation world differs greatly from dubbing for live action. A dubbed voice rarely matches the original actor’s voice, but for live action, this doesn’t really matter. The voice actor will speak with his or her own voice and will strive to ensure that the character’s attitude comes across and that the delivery is as clean as possible. This is the way live action dubbing has been done for years. Most animated characters, on the other hand, don’t have common “everyday-sounding” voices. They generally have funny, quirky voices that immediately become the character’s trademark. Everyone can recognize Bugs Bunny, Daffy Duck or the Powerpuff Girls after the first line. Not only will the actor dubbing an animated character have to re-create a trademark animated voice, the character’s attitude tends toward the extreme as well. The actor will have to follow these ever-changing moods, going from the calmest to the craziest. Thus, the actors have to concentrate on three critical points during the recording: animated voice, character’s attitude and delivery.

Because voice matches are rarely perfect — often far from it — the testing and selection of a voice actor involve coaching. I recall an instance in Japan during the early days of our dubbing group. We were looking for a voice for Dexter of the Dexter’s Lab cartoon series. I’d just finished listening to an impressive number of voice tests for Dexter, but nothing I’d heard was close enough. Everyone was getting nervous as we had to start recording very soon in order to deliver on time for broadcast. It was around 8 PM, and almost everyone was gone. From one of the windows of our little room, one could see the cafeteria where the actors usually wait between sessions. A girl was sitting there, alone. I asked the director who she was, and he answered that she was a new actress and didn’t have too much experience. I said, “Let’s try. At this point, we have nothing to lose.” The girl’s regular voice was a far cry from Dexter’s character voice. For almost an hour, which is a really long time for a voice test, she imitated what I was doing. She paced her voice like I did, compressed it in the way I showed her, lowered her pitch — and she finally got it. Immediately, I explained to her exactly what she had been doing with her voice so that she would be able to find that same voice again for the recording. She was amazed and so excited to get her first big role. Since that day, she has been our Japanese Dexter and a very good one. Realizing how easy it is to miss great voices, the studio — one of the biggest in Japan — decided to create a school for training new actors.

Our Japanese Dexter was a great and unexpected catch, and by working with her directly in the studio, I was able to get a great performance from her. But as I can’t be everywhere, I had to find a different way of helping actors, and so I devised written comments that offer them advice and techniques for reaching a particular voice. These voice placement techniques — developed in “Defining a New Vocabulary for Voice Characterization” on Animation World Network — represent the basis of our guidance at distance for actors. To write these comments, we must start with the actor’s voice analysis, as we’ll have to tailor our instructions to provide effective guidance. After that, we’ll concentrate on the actor’s rendition of the character’s attitude and on overall performance.

Let’s now go step-by-step through our casting process: the selection of the actor; the analysis of the voice placement to understand how he or she created the original character; and the creation of comments about the voice characterization, the interpretation of the character’s attitude and the performance.

The selection of the actor

I can hear which actor is the closest to the original and also who has the best potential to work on his or her voice. I rarely find a perfect match by just listening to the first test. For this reason, I trained my ear to recognize voices that have potential — actors who would probably be able to match the original but for some reason didn’t do it during the test. The causes can be various. Perhaps the actor didn’t listen to the original voice long enough or was not able to analyze it and consequently didn’t know how to reproduce it. Or perhaps the actor was not properly directed during the voice test session. I often find inexperienced actors who have great voices but lack the ability to use them to full potential. After many years of casting experience I developed an ability to detect a prospective actor for any given character, even if the actor actually tested for a different role. There is always something to be found in an audition and we might easily lose excellent actors just because their voice director misjudged them or didn’t test them for the right role. If we can adequately guide an actor to do a different character better, we can utilize a great voice, even in a role the actor didn’t audition for.

A few years ago in Mexico City I was recasting the character of Harry for the third Harry Potter movie. The previously approved actor’s voice was sounding too old for the character, and we didn’t have time to organize another round of voice tests. I asked the engineer if I could listen to all of the tests that were done for the entire movie. Listening to the actors state their names in their natural voices, I clicked on an actor doing the test for Malfoy. After listening to him a few times, I could hear in my head this voice becoming a perfect Harry, with a bit of work and coaching. The next day, we called this young actor, and because he was very flexible, he did a great job as the new Spanish-speaking Harry. This example, although pertaining to a live action movie, illustrates the importance of listening to the actor’s natural speaking voice. This is why we always ask the actors to use their own voices when stating their names on their audition tapes.

The analysis of the voice

Actors are rarely perfect in their first test. A point-by-point analysis is the only way to understand how the actors are placing their voices and to give them some precise instruction for getting closer to our character voices. It’s also a means of helping the actors understand what they are doing with their voices, thereby teaching them how to control everything better. I’m reminded of a mathematics course I took as a kid. Our teacher would base our grade more on how we reached the answer than on the answer itself. Now I realize that if you understand how you did a given thing, you will be able to do it again. If you create a voice but you don’t know which placement you used, it will be difficult if not impossible to find that voice again. Intuition must be supported by technique. A famous French singer once said, “Talent without technique is just a bad habit.” We want to believe that our actors will soon have both. They have the talent; now they must understand and develop the technique.

Our voice analysis aims to help them achieve that. During the auditions, we look at voice placement (such as where in the throat the voice comes from), pitch, projection (as opposed to volume), special effects (such as the nasal quality of the voice or a lisp), attitude and global performance. All of these are things you can talk about with the actor as long as you’re explaining what you mean by pitch, for example. These parameters are independent and can be adjusted by the actor, one at a time. If, for example, an actor gets the pitch right but cannot get the voice placement right for a given character, we can explain how to correct placement without moving pitch. This is why a thorough analysis of the voice test is so important, and the more detailed the analysis, the more accurate our guidance will be.

Creation of comments

The written comments we send to the international studios are our only means of communicating with the actors. The goal of this practice is to be able to talk to the actors about things we cannot show them in person. The comments have to be clear enough to allow them to perceive and work with something they could not define before. I first used this method some years ago as a student of electroacoustic music at the Conservatory in Marseille. We were creating contemporary music with sounds, noises and electronic processing. Just like the generation before us, we had to create a vocabulary to define and classify these sounds.

Comments may be particularly useful when we encounter too much similarity among different characters. This is normal, as bizarre characterizations are difficult to do. If you can easily recognize an actress doing a characterization, you will notice when her voice is used in another show, even if she is portraying a different character. We avoid this problem by giving the actress a precise explanation of how to create a voice that will be unrecognizable — or at least less recognizable — to the audience. In countries where the pool of actors is small, voice characterization training is essential in order to guarantee enough diversity from that small pool of voice talent. This is especially critical in regions such as Eastern Europe or Scandinavia where many dubbing actors are stage actors. They always deliver a great performance, but they can easily miss the character’s attitude — try to imagine Daffy Duck sounding like a Shakespearian character. As mentioned earlier, the voice is often the character’s trademark and has a very important interaction with the character’s attitude.

As an example of all of this, comments that I sent to the actor doing the Hungarian version of Bugs Bunny included these: “OK voice, but he is a little too linear — he needs to vary his delivery and voice a little more. His attitude needs to sound more arrogant and self-confident. His voice pitch could be a little lower, and his voice needs to be more nasal. The voice should be placed in the mask, which is the area between the nose and the upper throat. He should sound almost a little nasal but avoid sounding congested, which is easy to do if the voice is placed only in the nose. Keep the voice thin by not projecting too much and not forcing. Avoid creating gravel; if there is gravel in the voice, then the projection is too strong. The placement should be somewhat even, but add variation and contrast to the delivery to avoid sounding monotone. The attitude needs to be casual and relaxed, while keeping the performance animated and attitude a bit cynical. Be sure Bugs differentiates from Daffy’s placement and attitude.”

We can see how characterization, attitude and performance interact, and how they are equally important for the actor to achieve the right re-creation. After analyzing these three parameters, we can write our comments to create guidelines for the actor and voice director. This is the only “guidance at distance” we can provide, and, ideally, it will help the actor give the best performance possible. Then, and only then, can we say we have completed our casting.