Benjamin Button's digital head needed three different bodies

The shot

The Curious Case of Benjamin Button (2008) required Brad Pitt's character to age backward from 87 to 63 on screen while his emotional performance stayed constant. Digital Domain's own case study of the project describes director David Fincher challenging visual effects supervisor Eric Barba and the studio to create 'what was considered the Holy Grail of visual effects: create a completely believable computer generated human being.'

What the documents show

Digital Domain's account is a verified production record from the studio that did the work: it states the team 'created and integrated CG heads with three different body actors, and drove the character's every move with Pitt's performance and facial expressions,' delivering the digital character across '52 minutes of the movie, over 325 shots.' The Academy's own record of the 81st Academy Awards lists the film's Visual Effects Oscar as going to Eric Barba, Steve Preeg, Burt Dalton and Craig Barron, matching the names Digital Domain's own case study credits for the character work. Neither document describes the technique as a machine-learning system; both frame it as a modeling, capture and compositing pipeline built specifically for this production.

The workflow

Digital Domain's case study describes a pipeline that started with Pitt's own facial performance, captured and used to drive a digital head model, which was then composited onto one of three different body actors selected to match Benjamin's age and build in a given scene. The studio's account also notes its work 'encompassed on-set supervision for all visual effects work, along with pre-vis, compositing and matte painting,' indicating the digital-head sequences were planned and supervised during production, not assembled afterward from unplanned footage.

What the tool does not change

The system did not generate a performance; Digital Domain's own language is explicit that the digital head was driven by 'Pitt's performance and facial expressions,' meaning every expression the audience saw originated with the actor, not with the software. Selecting which body actor to composite for a given scene, and judging whether the seam between a digital head and a physical body actor's performance was convincing in extreme close-up, remained decisions the production made shot by shot.

• Whose recorded performance is a claimed de-aging or digital-double effect actually driven by, and is that clearly stated?
• Does a case study describe a pipeline built for one production's specific shots, or a general-purpose capability?
• How many minutes or shots does the studio's own account credit to the technique, against the film's full runtime?

The film's own production record sets a specific, checkable bar: a named actor's captured performance, driving a digital head, composited onto physical bodies, supervised on set, credited to four named individuals by the Academy. That specificity is what separates it from a promotional claim about de-aging in general.