Honda Brings Celebrity Yearbook Photos to Life in Super Bowl 51 Ad

Honda’s “Yearbooks” Super Bowl 51 ad brings celebrity yearbook photos to life with a message to always chase your dreams and never give up. Created by agency RPA for the automaker’s long-running “Power of Dreams” campaign and the 20th anniversary of the Honda CR-V, the 60-second spot was directed by Elastic’s Angus Wall — with a52 VFX supervisor Urs Furrer, CG lead Adam Newman, and Rock Paper Scissors editor Zoe Mougin — and aired in the first ad break of the game’s second quarter. Younger versions of nine A-list stars spring from the pages of their high school yearbooks: Steve Carell, Tina Fey, Robert Redford, Amy Adams, Magic Johnson, Stan Lee, Jimmy Kimmel, Missy Elliott, and Viola Davis.

SCANable captured detailed 3D head scans and facial expression scans of all nine celebrities for the ad, performing photogrammetric capture sessions in New York, Los Angeles, Atlanta, Tampa, and San Francisco to give the VFX team the data needed to convincingly animate each yearbook photo.

Nine calendars and a fixed air date

An advertising campaign built around well-known people does not get to set the schedule. Each subject has their own window, in their own city, often measured in tens of minutes between other commitments, and the broadcast date does not move.

Capture therefore has to be portable, fast to set up, and identical everywhere. Nine sessions run to nine slightly different standards produce nine assets that do not match, and the inconsistency shows the moment they are cut together in one film.

Our Mobile CyberSCAN™ Studios exist for exactly this pattern of work. The rig travels, deploys in a hotel suite or a stage corner, and produces the same data in Tampa as it does in Los Angeles.

A photograph is not a model

Animating a still image means inventing everything the photograph never recorded — the shape of the head behind the visible surface, how the cheek moves, where the light would fall if the face turned.

A scan supplies that missing dimension as measurement rather than interpretation. The subject’s real skull proportions, the real set of the features, and the real surface detail all become available, so the animated version stays recognisably the same person rather than becoming an illustration of them.

An expression set has to be systematic

Capturing a range of expressions is not a matter of collecting interesting faces. The useful deliverable is a structured set — a defined neutral plus poses that isolate specific muscle groups — because that is what a facial rig can actually blend between.

Consistency across the set matters more than any single pose. Every capture must share the same head position, the same array geometry and the same instantaneous exposure, which is the reason CyberSCAN fires all cameras on one frame rather than sweeping.

The results ship as render-ready facial geometry, aligned to a common base so a rigger can build blendshapes without first solving for alignment.

Browse more Super Bowl and campaign work, or see everything we offer.