"Apple TV+'s Severance isn't just a story about splitting work and life; it's a masterclass in visual effects. Despite looking clean and simple, the show's pared-back offices are layered with hidden effects. The Emmy-winning series relies on ILM and Mango VFX to make its surreal offices, landscapes, and jaw-dropping sequences feel completely real. Visual effects technology has come a long way."
"Today, the challenge isn't just creating something that looks cool; it's making it indistinguishable from reality. Even tiny mistakes can pull viewers out of the story, so precision is everything. For Severance, that meant blending CGI and practical locations flawlessly, using the best 3D modelling software, whether it was an everyday office or a disorienting hallway that defies logic. VFX work started in season one with early tests on select shots, long before anyone knew how the show would land."
"By season two, the complexity skyrocketed. The team produced over 3,000 VFX shots, enough to make five feature films in the time it normally takes to complete one. Some shots were composites of ten separate elements, with three different studios each handling parts of the same shot. Coordination at this scale was possible thanks to Flow Production Tracking, keeping every studio and department in sync."
Severance achieves its unsettling realism through intensive visual effects work and tight cross-studio collaboration. ILM and Mango VFX combined CGI with practical locations to render surreal offices, landscapes, and disorienting architecture that remain believable on screen. The production prioritized precision to avoid small errors that would break immersion, conducting early VFX tests in season one and expanding to more complex work in season two. Over 3,000 VFX shots were produced, with some shots composed of ten elements handled by multiple studios. Flow Production Tracking and scalable tools like Autodesk Flame and 3ds Max enabled coordinated workflows and rapid processing across teams.
Read at Creative Bloq
Unable to calculate read time
Collection
[
|
...
]