Smile 2

Smile 2 is Paramount Pictures’ 2024 psychological supernatural horror sequel starring Naomi Scott, written and directed by Parker Finn. The follow-up to Smile (2022) — greenlit after the first film’s commercial success and announced in April 2023 — stars Scott as a pop singer who experiences a series of increasingly disturbing events as she is about to go on tour. Rosemarie DeWitt, Lukas Gage, Miles Gutierrez-Riley, Peter Jacobson, Raúl Castillo, Dylan Gelula, and Ray Nicholson round out the supporting cast, with Kyle Gallner reprising his role as Joel. Principal photography ran from January to March 2024 across New York locations, and the film opened in the United States on October 18, 2024.

SCANable provided 3D scanning services for the production, capturing talent, props, and environments to support the film’s visual effects, including camera tracking and CG background work.

You need a room before you can break one

Psychological horror works by making a familiar space behave incorrectly — a corridor that runs too long, a ceiling that is not where it was, a camera that rotates through a room in a way rooms do not allow.

Every one of those effects is a deviation from a correct baseline, which means the correct baseline has to exist first. A distortion applied to a guessed geometry is just a wrong image; a distortion applied to measured geometry reads as the world itself misbehaving.

That is the specific value of point-cloud capture on films with no visible creature work. The dataset describes the room truthfully, so the film can lie about it precisely. Our set and location scanning supplies it.

A lead in almost every frame

When a single performer carries the film, their scan becomes one of the most reused assets on the production. It is consulted for face work, for background placement, for a stunt beat, for a reflection, for a shot where the schedule made the real coverage impossible.

An asset with that much traffic has to be right at capture. Reworking it later means reworking every shot that has already been built on top of it.

Simultaneous-array photogrammetric capture is what makes the initial record dependable — every camera exposes at the same instant, so there is no drift from breathing or micro-movement to be cleaned up by hand and no likeness lost in the cleanup.

Overhead structure nobody photographs

Performance venues carry a second environment above the one the story happens in: trusses, lighting rigs, cabling, and rigging points, all of it structural, most of it never framed directly.

It still matters. Overhead hardware defines where light comes from, what reflects in a polished floor, and what a camera passes beneath in a crane move. It is also awkward to capture, sitting well above every convenient scanner position and largely self-occluding.

Recording it during production is the only realistic option, because it is the first thing struck and the last thing anyone photographs for reference. Meshes built from that data go out after cleanup and render-ready preparation, together with any digital doubles the production has ordered.

See more feature-film capture work, or browse everything we offer.