On this episode of the fxpodcast, we have an exclusive behind-the-scenes look at the visual effects work for Amazon’s Étoile, featuring an in-depth discussion with David Gaddie from Afterparty VFX. The team undertook the challenging face replacement work for the show’s ballet dancers, employing cutting-edge AI technology to seamlessly blend lead actors’ faces onto the performances of professional dance doubles. Étoile, the latest series from Amy Sherman-Palladino, renowned creator of The Marvelous Mrs. Maisel and Gilmore Girls, centers on elite ballet dancers. To realize this vision, the show’s VFX supervisor, Lesley Robson Foster, engaged Afterparty VFX to research and develop innovative AI-driven solutions capable of handling the demanding visual complexities inherent in ballet sequences. This task was notably difficult, as traditional deepfake tools are typically optimized for straightforward, frontal shots. Ballet, however, involves rapid spinning, flips, significant motion blur, and hair frequently obscuring faces, creating immense technical hurdles. Additionally, director Amy Sherman-Palladino preferred long, uninterrupted takes, some lasting nearly a full minute, eliminating conventional editing methods and cheats used to mask transitions between the actors and their dance doubles. Crucially, the final shots needed to authentically reflect the actors’ performances, rather than simply showcasing their doubles. In this episode, we explore in-depth how David Gaddie and the Afterparty VFX team developed proprietary solutions tailored specifically to these unique challenges. Their process combined computer-generated imagery, advanced AI, meticulous data segmentation, extensive manual refinements, and significant artistic skill to achieve convincing, performance-driven visual effects. #fxpodcast #chefs #data #etoile #machine