Everytime I go to the movies, it feels like work. I catch myself trying to study the animation and the special effects.
It is almost impossible nowadays to see a movie that does not require some background computer animation. In Alita Battle Angel, I set myself up to some headache.
The movie is based on a manga by Yukito Kishiro. In 2563, a cyborg is rescued by a doctor and nursed back to health while trying to remember her past. The action is set in Iron City, a trash city under the glorious Zelma, a flying city no one can get up to. The whole movie features robotic bodies and weird human machinery, all this in a futuristic world.
The most surprising is not how freakishly real those robotic/human bodies look, it is the performance of the main character Alita. The contradiction is maximal. An actress performs and is then rendered into a 3D character that needs to look real. Her eyes are bigger, but that is basically the only thing you would notice in her human face.
I watched some reviews and behind the scenes videos of this huge production, produced by James Cameron and in partnership with Weta.
Alita’s human face had about 5000 iterations to make her emotions look real and the animators had to meet with plastic surgeons to fully understand what was happening behind a face to then re translate it in animation. Alita was all about motion capture and performance capture meaning that not only her body but also her face were closely recorded and translated in 3D the way it was many years ago for Gollum in Lord of the Rings and for Avatar. The difference here is we needed for her to almost look like a human, making her reactions as close as Rosa’s. The main focus was faithfully re translating the nuances of human acting.
Once again the advancement of computerized animation go beyond my belief.