Presentation
The Making of The Great Galactic War: Directing Generative Video in 3D
DescriptionThe Great Galactic War, a short-form streaming series from Nick Montori, Co-Founder and Creative Director of AI production house Anima Studios, has surpassed 120 million views. This session shows how a small team created a consistent, cinematic world at streaming scale, and the virtual cinematography tool that made it possible.
Like many teams working with generative video, Anima faced a core challenge: control and consistency. Foundation models can produce striking frames, but camera movement, composition, lighting, and scene continuity can shift unpredictably between renders, making it difficult to direct a coherent series.
Using real production footage, Pinar Seyhan Demirdağ and Dade Orgeron will show how the team used EVRN, a 3D directing layer for generative video, to solve that problem. They will demonstrate how a flat image becomes an editable 3D scene where characters, props, and environments become independent objects with accurate spatial relationships. The team could then place cameras, light scenes, compose shots, and lock layouts before sending them to the generative model, replacing prompt-driven guesswork with deterministic references for repeatable renders.
Attendees will learn how spatial direction fits into existing AI production pipelines and preview EVRN’s upcoming support for USD, GLB, OBJ, Unreal Engine, Unity, Blender, and Houdini.
Like many teams working with generative video, Anima faced a core challenge: control and consistency. Foundation models can produce striking frames, but camera movement, composition, lighting, and scene continuity can shift unpredictably between renders, making it difficult to direct a coherent series.
Using real production footage, Pinar Seyhan Demirdağ and Dade Orgeron will show how the team used EVRN, a 3D directing layer for generative video, to solve that problem. They will demonstrate how a flat image becomes an editable 3D scene where characters, props, and environments become independent objects with accurate spatial relationships. The team could then place cameras, light scenes, compose shots, and lock layouts before sending them to the generative model, replacing prompt-driven guesswork with deterministic references for repeatable renders.
Attendees will learn how spatial direction fits into existing AI production pipelines and preview EVRN’s upcoming support for USD, GLB, OBJ, Unreal Engine, Unity, Blender, and Houdini.
Event Type
Industry Session
TimeThursday, 23 July 20269:00am - 10:00am PDT
LocationRoom 502A
Production & Animation
Animation
Artificial Intelligence/Machine Learning
Generative AI
Pipeline Tools and Work
Real-Time
Rendering
Full Conference Supporter
Full Conference
Experience
Discover
