Media, VFX & Games

Laan Labs builds capture and rendering technology for film and television production, VFX studios, game developers and immersive media teams. Locations, sets, props and performers are captured with phones and ordinary cameras and reconstructed as Gaussian splats, radiance fields or meshes that go into a production pipeline. Laan Labs recorded volumetric "bullet time" video with 15 synchronized iPhones and NeRFs and wrote its own Gaussian splat trainer and renderer in Metal for 3D Splat App.

Production has always needed the real world in digital form: a location to plan shots in, a set to extend, a prop to rebuild, a performance to see from an angle no camera was at. Until recently that meant a LiDAR crew, a photogrammetry rig or a volumetric stage. Phones with depth sensors and radiance field reconstruction have changed the cost. A scout scans a location in minutes, a VFX supervisor records the set between takes, and an artist starts from a capture of the real object.

Laan Labs has worked with media companies including HBO, NBC, MTV and TBS, and has seen its scanning technology adopted across film, television and game production, from location planning and storyboarding to building game assets from scanned objects.

How Laan Labs helps

Location and set capture

LiDAR and photo scans of locations and sets at true scale, for scouting, shot planning, storyboarding and previs, and as a measured reference for set extension and match-moving in post-production.

Photorealistic environments

Gaussian splat and radiance field captures that reproduce a place with its light, reflections and fine detail, for virtual scouting, review with a director or client, background environments and immersive experiences. Quality depends on how thoroughly the place was photographed.

Props and game assets

Real objects, vehicles and environments scanned and delivered as textured meshes in OBJ, USD and glTF, giving artists an accurate starting point to retopologize and optimize, in place of modeling from photographs.

Volumetric and bullet-time capture

Synchronized multi-camera rigs built from phones, with time alignment, color matching and per-frame reconstruction, producing a moment or a short performance that can be viewed from the range of angles the cameras covered. Quality is best near the camera positions and falls off between and beyond them.

Faces and performers

Close-range scans of faces, hands and bodies from the iPhone TrueDepth sensor and from photographs, as reference for digital doubles, character work and fitting of costumes and prosthetics.

Custom trainers and renderers

Reconstruction and real-time rendering written for the hardware in use, including training on Apple Silicon without a cloud upload, compression to SPZ and SOG, and viewers for the web, iPhone, iPad and Mac.

Immersive playback

Captures shown at full scale on Apple Vision Pro, Meta Quest and other headsets, for location review, exhibitions and audience-facing experiences.

AI for video and content

Vision and language models that watch and describe video at scale, as in CreatorView, built with Glow Agency to assess creator and brand fit across social media content.

Where to start

A first project is usually one capture problem on one production: a location that has to be revisited after the crew has left, a prop that needs a digital copy, or a shot that calls for a frozen moment. Laan Labs builds and runs the capture, delivers the result in the formats the pipeline expects, and documents what coverage and conditions were needed to get it, so the method can be repeated by the production's own team.

Studios and tool makers that want the technology inside their own software can license it or commission components. See Gaussian splatting and radiance fields and spatial computing, AR and VR development.

Laan Labs products in this field

• 3D Splat App — Gaussian splat training and viewing on Mac, with no cloud upload
• RadianceView — Splats, scans and photographs shared as one link

Common questions