A computer-generated actor who is as good as the real thing? Don’t be surprised

The recent release of Rogue One (2016) has sparked an unexpected controversy. The film features Peter Cushing, a familiar face from the original Star Wars: IV A New Hope (1977), reprising his role as Grand Moff Tarkin. Cushing appears in new scenes and interacts with fresh characters, despite having died in 1994. More than 20 years later, an actor has been digitally resurrected.

Responses have been mixed, to say the least. Some hate the digital version of Cushing because they perceive it to be artificial and distracting. Others have ethical concerns about the use of a deceased actor’s image (although Cushing’s estate gave permission for this). Other viewers just assumed they were watching a live actor.

As a visual effects artist, I found the work in Rogue One impressive but not totally convincing. There is still, for me at least, an element of the uncanny valley, that oddly disconcerting gap between the artificial and the real that is yet to be completely overcome in this area. If the computer generated version of Cushing had been used more sparingly the audience would have had less time to study every detail and search for flaws. (Of course, those of us who knew Cushing was dead seem to have been the only ones distracted.)

The CG version of young Princess Leia in Rogue One was altogether more successful with audiences, though I confess to finding her smooth-skinned youthful face less believably real than digital Cushing’s craggy visage. The acceptance of CG Leia was likely due to the character’s limited screen time – before you had time to fully process what you were seeing the scene was over. Carrie Fisher herself is said to have approved of the move. Since her death, however, Disney has emphasised that it will not be creating a digital Leia for future episodes.

Regardless of the success or otherwise of these examples, I suspect we are very close to a genuinely believable CG actor who will fool even experienced professionals. As an artist I find the prospect exciting, though as a human being I have twinges of concern that the technology may not always be used wisely.

Indeed the ability to digitally replicate actors as photoreal versions of themselves at any age is a tool of such powerful storytelling potential that regardless of public opinion, directors will demand the technology be continually improved.

This has been a long time coming. For at least 15 years, there have been various attempts at recreating photoreal digital versions of living actors.

The efforts range from the ridiculous – such as when Dwayne Johnson’s likeness was mangled to create the Scorpion King in The Mummy Returns (2001) – to a digitally rejuvenated Arnold Schwarzenegger who first appeared in Terminator Salvation (2009), before various versions of Arnold at different ages appeared in Terminator Genisys (2015).

Robert Zemeckis, with his trilogy of motion captured CG films, The Polar Express (2004), Beowulf (2007) and A Christmas Carol (2009), has arguably done more than any single director to push the boundaries of fully CG digital actors.

The characters from these films are oddly unnerving to watch and more than one commentator has called them “creepy”. However, they represent key moments in the development of an emerging technology.

Digital scans and body doubles

A more common current use for digital doubles is to portray living actors in situations of great danger. There are a couple of options in this scenario. One is an entirely digital character created from a full body scan of the actor. The other option is to digitally paste an actor’s face onto the body of a stunt performer.

Disney has recently denied a rumour that for the next Avengers film it plans to paste Benedict Cumberbatch’s face on to a body double to facilitate shooting around his busy schedule.

However, it is very likely that Cumberbatch will shoot closeups for performance when he is available and digitally altered body double shots will be used for his action sequences. Extensive digital doubles were already used for Doctor Strange (2016) so there is no reason to assume Disney is about to change its working methodology.

Less well known is the use of this technology for “beautification”. Body doubles for nude scenes have been used in Hollywood for decades. Now, an actor’s face can be pasted right on top of the body double for a seamless effect, though no actor wants to admit to it and visual effects companies are required to sign non-disclosure agreements.

It is generally only when things go terribly wrong – such as when viewers noticed obvious signs that Lena Headey’s nude Game of Thrones Walk of Shame involved Headey’s face pasted on to another actress’s body – that it’s noticed.

There is an evolving debate about the ethics of actors being made to appear younger or more attractive. However the “person” on screen has always been a fictional character that has only ever been partially brought to life by a performer. Makeup, wardrobe, lighting, all contribute significantly to their look. Digital effects are just one more tool giving filmmakers added flexibility in the depiction of the character.

Digitally archiving for beauty

It is now standard practice for any up and coming young actor who draws the attention of a major studio to have a full body scan completed so that there will always be a digitally archived version of them in their “prime”.

These archives can be accessed for dangerous stunt work, or in the case of unexpected injury or death that may prevent the actor from completing filming, such as when Paul Walker died before completing Furious 7.

However it is becoming more common for scans such as these to be used as the basis for beauty work.

Harrison Ford famously refused to dye his hair for Indiana Jones and the Kingdom of the Crystal Skull (2008) and again for Star Wars: the Force Awakens (2015). However, not all actors may have the confidence or the industry clout to resist when there is a digitally perfect version of their 25 or 30-year-old self readily available to be used.

Peter Allen, Lecturer in Film and Television, Victorian College of the Arts, University of Melbourne.

This article first appeared on The Conversation.