Constructing a photorealistic digital human has long been considered the ‘holy grail’ of digital visual effects. While there are many components to constructing a digital human, the face is among the most important. The face serves as the body's most direct marker of identity and the place where the self is communicated to others. But the face is also one of the most difficult parts of the human body to adequately simulate in computer-generated imagery. Humans evolved to recognise faces and spot the most minute facial expressions, as well as the ways that they might be unnatural. The face is made up of many moving parts, including forty-three specific muscles connecting to fourteen different bones. These muscles and bones – along with skin, lips and eyes – move differently in every person, thus creating the particular facial tics, quirks and expressions that become recognisable as part of a person's identity. Recreating all of these elements in a computer program, with sufficient detail and subtlety to fool the human eye, is a daunting challenge.

Efforts to create a photorealistic digital human have been aided by performance capture, a set of techniques that record the facial and bodily movements of a human performer and apply them to a digital character. By involving an actor – often a well-known actor in a highly publicised role – performance capture places the efforts of a human performer at the centre of digital character construction, rather than the work of programmers and digital effects artists. However, performance capture is a compromise between an ideal process that would seamlessly transfer a human performance to a digital character and a technological solution that crafts a performance solely out of the ones and zeroes of a computer. Despite promotional rhetoric to the contrary, performance capture does not record every nuance of a performance and transfer it magically to a digital being. While the process does record elements of an actor's performance, it typically combines that fragmentary and incomplete data with software programs, digital facial models, and animation to create the final computer-generated (CG) performance. Thus, performance capture participates in a representational binary. On the one hand, cinematic technologies aim to directly record a facial performance; on the other hand, digital technologies allow for the complete fabrication of a face.