Lexus December to Remember Commercial – “Racetrack”

Lexus’ “Racetrack” spot from the 2015 December to Remember holiday campaign was created by agency Team One and production company Elastic, directed by Pat Murphy. Visual effects studio a52 brought the toy racetrack world to life, with VFX supervisor Jesse Monsour, head of 3D Kirk Shintani, and CG supervisor Adam Newman leading the CG work, and editor David Brodie of Rock Paper Scissors cutting the spot.

SCANable provided 3D scans of the toy cars used in the racetrack shots, delivering accurate photogrammetric prop geometry for a52’s CG artists to animate and integrate seamlessly into the commercial.

There is no angle the camera will not find

Most prop scanning produces supporting data — something to track against, or to help a modeller get proportions right. On a spot built entirely around the objects themselves, coverage stops being a question of what the shot needs and becomes a question of what the shot might need.

Every surface has to be captured, including the ones a real toy hides: the underside, the interior of a wheel arch, the seam where two moulded halves meet. Once the object is animated and the camera is free, all of it eventually faces frame.

Incomplete coverage is only discovered later, in a shot that is already approved for everything else.

Small subjects, close cameras

Scanning a hand-sized object is not a scaled-down version of scanning a room. The working distances collapse, depth of field becomes the limiting factor, and the number of camera positions needed to cover a compact form with overlapping detail goes up rather than down.

Turntable and array capture solve this by moving the subject through a controlled set of orientations instead of walking a scanner around it. Lighting stays constant, coverage is systematic, and the underside — the part a handheld pass always misses — is captured properly.

The result is a mesh with even detail density across the whole surface, rather than one that is sharp where the operator stood and vague everywhere else.

Geometry that has to move

An object destined for animation has requirements a static scan does not. Raw photogrammetric output is a dense, uniform, unstructured mesh: accurate, but unusable for rigging, deformation or efficient rendering.

Making it animatable means retopology — rebuilding the surface with clean, deliberate edge flow that follows the form, sized so a rig can bend it where it should bend and hold it rigid where it should not. Then a UV layout that lets texture sit correctly, and correct real-world scale so physics and camera lensing behave.

That is the work our modeling and asset preparation exists to do, and it is what separates a scan from a shot-ready asset. The wider service list covers where the capture itself fits.

See more brand and spot work.