Fallout

Fallout is a post-apocalyptic drama series developed by Geneva Robertson-Dworet and Graham Wagner, and executive produced by Jonathan Nolan and Lisa Joy, for Amazon Prime Video, which premiered in 2024. It is based on the celebrated role-playing video game franchise created by Interplay Entertainment and now owned by Bethesda Softworks, bringing the games’ retro-futuristic wasteland to live-action television.

SCANable provided ground and aerial 3D laser scanning of props, sets, characters, and locations for the series. Combining terrestrial LiDAR, aerial photogrammetry, and 3D body scanning, our team delivered the accurate environment data and digital-double-ready captures that supported the show’s extensive visual effects and set-extension work.

Assets that have to hold across a season

A feature has one continuous production. A series has eight episodes, multiple directors, several units, and a post schedule where episode one is being finished while episode seven is still shooting.

An asset captured for episode two will be used in episode six. If its scale, orientation, or naming drifts between those points, the world stops being one world.

Episodic work therefore treats the scan library as infrastructure rather than as a series of deliveries. Every capture is registered into the same conventions, so a location scanned once behaves identically wherever it reappears across the season.

Making the set-extension seam disappear

Practical construction stops where the budget stops. Everything above the top of the built wall, beyond the end of the built street, or past the edge of the built interior is added later.

The join between the two is the only part the audience can catch. It works when the digital half is modeled outward from the practical half’s measured geometry — not matched to it by eye, but derived from the same point cloud.

Extensions built that way share the set’s exact dimensions, so the seam falls where the geometry already agrees. Our set and location scanning is what supplies those measurements while the build still exists.

Ground and air in one registered dataset

Terrestrial scanning is precise but ground-bound. It cannot see roof surfaces, upper structures, or the shape of terrain beyond a few hundred metres of clear line of sight.

Aerial photogrammetric capture covers exactly what the tripod cannot, and stops exactly where fine detail begins.

Neither is sufficient alone. The useful product is both, registered into a single coordinate space so a VFX team receives one continuous environment rather than two datasets that need reconciling. Ruined and irregular architecture makes this harder than a clean modern build, because there are few flat planes and straight edges to register against — which is precisely why the measurement has to be right at capture time.

More television work is in the portfolio; the services page covers how it is captured.