October 2008 [CG]
Motion Capture & Facial Capture
Since the setting of Biohazard is in the States most of the actors are non-Asians. How did you decide on the actors?
We selected 7 people out of several Hollywood actors and actresses who came to the audition. They all flew to Japan for the shoot. During the 10 days stay in Tokyo, we used two studios (one for body and one for facial capture). Since the schedule was close to winter holiday season, we opened both studios so that we could work faster and finish before Christmas. First we collected data for the body movements then while we monitored the action, we gathered the facial data as well.
During facial capture I assume the actors had dialogues. Did you record actor’s elocution at the same time?
We used dubbing, which is also used in studios in LA. We used different people for motion capture and voice. For character Clair and Leon, we used the same voice as in the game (Clair: Alyson Court, Leon: Paul Mercier). With Alyson we held the recording at the studio in Canada. Due to tight schedule, we went online and recorded in Canada and LA.
This film was released in both English and Japanese. I believe that the timing for the two languages were different. Were the motions recorded to match English or Japanese language?
All timing for motion capture was based on English. Since actors were native English speakers we had no problem with the elocution but there were differences between the motion picture and dialogue so to eliminate the hassle of adjusting the duration of the film, we showed the film to the actors as they read their lines. Yet still we had to make few adjustments in the dialogue to match the motion picture.
There is a scene where Clair jumps into the water when the laboratory explodes. How did you go about filming this scene?
All actions were captured at the studio so for the water scene actor would stand on stage and act as if she was under water. Later, we fixed the motion by hand.
Angela and Leon Underwater
At the beginning there is a strike scene at the airport and we see realistic crowd of people. Is this all done at the capture studio or done using crowd simulation software?
For this film, it was all done at the studio. We took the data for each individual as well as in smaller groups and placed individual motion data onto the crowd data and adjusted the overall crowd movement.
Are there any points which you noticed during this character animation process that you would like to improve in the future?
For this project, we captured facial and body separately so next time we would like to do it simultaneously.
Shading and Rendering
One of the main differences between game and film is the size of the image. For game, user will watch on computer or television screens but CG movie is projected on screen size that is over 200 inches. Characters and background are seen larger so they need to be created with higher resolutions.
I assume theatre version have greater scale than the game version. Did you have difficulties when creating contes for airport and laboratory scenes? Did you have trouble rendering?
We had to consider the entire scale to create more drama. Although we were aware of this, we could have exaggerated even more.
I thought the airport was very realistic but did you use an actual airport as a model?
Originally we were thinking of an airport in one of the western part of the States but the director had an image of the airport that appears in the drama “24” so we went to the States to do some research. However, since the story takes place in 2005 at an industrial city with a population of 100 thousand people, we had to design cars, airplanes, and buildings that would match the setting. Also, there were copyright issues so we had to create original designs for everything including weapons. We spend some time creating the overall atmosphere.