In a 2019 interview with Rolling Stone, Robert Rodriguez and “Alita: Battle Angel” producer Jon Landau broke down the method of translating the minute features of Rosa Salazar’s efficiency into Weta’s computer-generated model of Alita. The method began with the digital scanning of Salazar’s face and the creation of a live-action model of Alita that captured the actor’s mannerisms whereas fine-tuning her look to go well with the character’s anime-inspired aesthetics. Landau clarified that the digital model was created solely after Salazar got here on board, as her “little idiosyncrasies” have been pivotal to who Alita is:

“We wished the entire little idiosyncrasies that she introduced in her audition to be there. We discovered from doing ‘Avatar’ that once you’re scanning somebody’s facial construction that it was essential to get the decrease a part of somebody’s face proper — it is tougher than you assume. As soon as we knew we might do this, we began working with them on the entire digital course of of making what Alita would appear like.”

Whereas Alita appears like an augmented CGI model of Salazar on the floor, the method of reaching the character’s ultimate look took 9 months, prompting Rodriguez and Landau to equate it to “the human birthing course of.” Landau additionally defined that Weta’s major problem was to push the envelope of facial and efficiency seize, which was achieved with the help of high-definition cameras and made the expertise extra “actor-centric.” The heavy use of digital tech and VFX in movies usually runs the danger of overshadowing its sensible features and human performances, however on this case, digital-Alita solely enhanced Salazar’s dynamic efficiency as an alternative of overwriting it. This was made potential with unbelievable attention-to-detail and the measured use of Salazar’s transferring rendition of the cyborg battle-angel because the blueprint for enhancements.