Uncanny Valley
The unsettled feeling triggered by computer-generated humans that are nearly, but not quite, photoreal.
In depth
As a character becomes more unrealistic, we find it cute (like Wall-E). As it becomes *almost* human but not quite perfect (dead eyes, stiff movement), our brain rejects it as a 'corpse' or 'sick', causing revulsion.
Example
“The animated characters in 'The Polar Express' are often cited as prime examples of falling into the Uncanny Valley.”
Origin and history
The concept was named in 1970 by Japanese roboticist Masahiro Mori in his essay 'Bukimi no Tani Genshō' ('The Uncanny Valley'). Mori was describing humanoid robots, but the term migrated to computer animation in the 2000s as productions like 'The Polar Express' (2004) attempted photorealistic human characters and audiences responded with discomfort. The phenomenon is now a central concern of any production attempting digital humans.
Why filmmakers use it
The uncanny valley exists because human faces are the single thing audiences are most expert at reading. Tiny micro-expressions and eye-light cues are processed unconsciously, and any mismatch reads as 'wrong' even when viewers cannot articulate why. Productions either embrace stylization to avoid the valley (Pixar) or invest enormously in performance capture and lighting realism to cross it (Avatar, Alita: Battle Angel). The middle ground is unsafe.
More examples in cinema
- 'The Irishman' used digital de-aging on Robert De Niro and Al Pacino, drawing mixed responses about whether the result crossed the valley or fell into it.
- Marvel's CGI faces during the Disney+ era have been frequent subjects of uncanny-valley debate among audiences and critics.