The Technologically Determined Decade: Robert Zemeckis,Andy Serkis,and the Promotion of Performance Capture

Introduction

In 2004, director Robert Zemeckis released The Polar Express, the first in a series of five films he would either direct or produce that utilized the technology known as performance capture. Zemeckis was already a major Hollywood director known for his effects-driven films – including Back to the Future (1985), Who Framed Roger Rabbit (1988), Death Becomes Her (1992), and Forrest Gump (1994) – when he began work with a technical team from Sony Pictures Imageworks on refining existing motion capture technologies for the production of The Polar Express. Motion capture relies on utilizing data of a performer’s movements – most often gathered through the use of skintight suits covered in reflective dots – to map those movements in a digital context. Performance capture, as refined by Zemeckis and his team, allowed for the simultaneous capture of both a performer’s bodily and facial movements, creating a more totalized and unified set of data for use in the film’s animation.

Over the next several years, Zemeckis would dedicate his career to performance capture. He would direct two more performance capture features, Beowulf (2007) and A Christmas Carol (2009), and produce two other films, Gil Kenan’s Monster House (2006) and Simon Wells’s Mars Needs Moms (2011), and in 2007, he and his business partners would enter into a deal with Disney to form ImageMovers Digital, an animation facility dedicated to performance capture films. In interviews and other promotional materials, Zemeckis spoke in glowing terms about the technology, utilizing lofty rhetoric to describe his vision for what performance capture was capable of and how it could revolutionize film acting.

Zemeckis’ enthusiasm for performance capture was not, however, shared by everyone. In her reviews of Flight (2012), Zemeckis’ first live action film since Cast Away (2000), New York Times reviewer Manohla Dargis writes, ‘To watch Mr. Zemeckis working fluidly in consort with Mr. [Denzel] Washington’s ferocious performance is to regret this director’s last, technologically determined decade’ (Dargis, 2012). Dargis speaks to a dominant perspective on Zemeckis’ career during what she describes as a ‘technologically determined decade’, the period of technological experimentation in digital filmmaking that spanned from The Polar Express through to the Zemeckis-produced Mars Needs Moms. This is a period, according to Dargis, to ‘regret’. Performance capture itself has become central to the special effects of contemporary Hollywood – used extensively in James Cameron’s Avatar (2009), Joss Whedon’s The Avengers (2012), and the many computer-generated roles of Andy Serkis, to name several big budget examples – but Zemeckis’ efforts to promote his take on performance capture fared poorly.

In his 2003 book, The Cinema of Robert Zemeckis, Norman Kagan sums up the director’s career with the observation, ‘Zemeckis has been one of America’s and the world’s foremost filmmakers, though perhaps shrewdly keeping a very low profile’ (p. 225). Kagan’s book was published the year before the release of The Polar Express; in it, Kagan stakes out several key aspects of Zemeckis’ directorial style. The most important of these, reiterated throughout the book, is ‘his obsession with technique and technology’ (p. 9). The centrality of technology and special effects in Zemeckis’ filmmaking is undeniable, although throughout his career much of Zemeckis’ digital technique purposefully keeps ‘a very low profile’. This, ultimately, is one of the key shortcomings of the ‘technologically determined decade’ according to critics. Performance capture is an extension of Zemeckis’ career-defining ‘obsession with technique and technology’, but has been criticized for how overt and frequently off-putting the effect is. Zemeckis would remain committed to the technology through the production of five films released between 2004 and 2011, and throughout these years Zemeckis was a vocal advocate for the technology. As a high-profile director of blockbuster films, Zemeckis used his position to champion performance capture.

This article will focus on the critical response to performance capture films, as well as the ways that these films and the technology itself have been promoted in interviews with production personnel. Although my main focus is on how critics reacted to Zemeckis’ performance capture films and how Zemeckis himself defined and defended the technology, I also examine how others connected with the technology have been similarly active as advocates for performance capture. Andy Serkis serves as an excellent counterpart to Zemeckis, having earned critical acclaim for his performance capture roles, including Gollum in Peter Jackson’s Lord of the Rings trilogy (2001, 2002, 2003), Kong in Jackson’s King Kong (2005), and Caesar in Rupert Wyatt’s Rise of the Planet of the Apes (2011). A common thread in both Zemeckis’ and Serkis’ promotion of performance capture has been the ways that both men frame the technology in terms of benefits it provides to film production, and, in particular, to film acting.

Focusing on how this technology is defined and discussed through the words of Zemeckis and Serkis – both famous above-the-line figures who present a privileged perspective on film production – is potentially fraught. As John Caldwell writes in Production Culture (2008: 318):

Studying the industry from the top-down – that is, from the vantage point of above-the-line executives in interviews – can be so useless. At least if one wants to get past the cultural flak of personal branding in order to examine actual production activities.

However, my interests lie precisely in ‘the cultural flak of personal branding’, exploring the ways that Zemeckis and Serkis come to represent, embody, and speak for the technological feats produced by hundreds of designers and technicians at companies like Sony Pictures Imageworks, ImageMovers Digital, and Weta Digital.

The impact of motion capture and performance capture technologies on filmic performance has been studied by several scholars, including Scott Balcerzak, Lisa Bode and Tanine Allison. These scholars highlight the importance of maintaining a critical skepticism toward the discourse put forth by industrial professionals. My project focuses on the promotional rhetoric of directors like Zemeckis and stars like Serkis, and it is vital not to take such statements about the objectives or implementation of these technologies at face value. Individuals like Zemeckis and Serkis can be a poor source of information on the state of the technology; the below-the-line labor of digital artists is often more important. Although in this article I chart a portion of the history of performance capture technology – particularly as it pertains to Zemeckis’ productions – the promotional rhetoric is my focus, not for its flawed representation of what the technology is or can be, but for how the technology is sold to audiences and critics.

A brief history of Robert Zemeckis

A full account of Zemeckis’ career is beyond the scope of this article, but some context is necessary to understand the significance of his years spent evangelizing for, and actively working with, performance capture. Understanding Zemeckis’ position in Hollywood before he entered this so-called ‘technology determined decade’, and the importance of technological experimentation to his directorial star image, helps to better illustrate how his promotion of performance capture was received.

In terms of box office numbers, Zemeckis is one of the most successful directors in Hollywood. By overall domestic gross of all of his films, he is the third highest grossing director, with a combined total of approximately $2 billion (USD); he is only slightly behind Michael Bay at $2.1 billion, with both Zemeckis and Bay dwarfed by Steven Spielberg at $4.2 billion (The Numbers, nd, a). On a global scale, Zemeckis is the sixth highest grossing director with approximately $4.1 billion (The Numbers, nd, b). Over the course of two decades, his output has often been consistently among the best performing films of their years. Between 1984 and 2004, 4 of the 11 films he released had either the number one (Back to the Future and Forrest Gump) or number two (Who Frame Roger Rabbit and Cast Away) domestic gross for their years, and only 3 of the 11 (Back to the Future Part III, Death Becomes Her, and Contact) did not appear in the top 10 films (Box Office Mojo, nd).

Zemeckis’ films have not only been successful at the box office but many of his films are also noteworthy for their use of special effects technology. The Back to the Future movies are obvious examples; when one thinks of special effects driven filmmaking, it is usually in the context of science fiction and action genre films. Back to the Future Part II was nominated for the Academy Award for Best Visual Effects, losing to James Cameron’s The Abyss (1989). Three other Zemeckis films did win that award: Who Framed Roger Rabbit, Death Becomes Her, and Forrest Gump.

Forrest Gump is one of Zemeckis’ most successful and acclaimed films; alongside its Best Visual Effects win, it also won Academy Awards for Best Directing, Best Actor in a Leading Role, Best Adapted Screenplay, and Best Film Editing. Like Zemeckis’ other Academy Award winning effects films, it features a range of technically impressive effects. The most overt of these come in the form of sequences where Forrest Gump (Tom Hanks) is digitally composited into archival footage, allowing the film to depict the character interacting with Presidents Kennedy, Johnson, and Nixon, as well as other major figures of the mid-20th century, like George Wallace and John Lennon. Wheeler Winston Dixon observes that historical context helps signal the artificiality of these scenes: ‘the cultural/spatial displacement of Tom Hanks’ persona in juxtaposition with thirty-year old newsreel footage immediately signals to most audiences the intentionally duplicitous construction of these scenes.’ On the other hand, Dixon also notes ‘the numerous scenes in Forrest Gump which seem to be straightforwardly photographed, but which in fact involve many hours of digital effects work’ (Dixon, 1995/1996: 56). The digital compositing used to show Hanks receiving a medal from President Johnson is clearly marked as a special effect, but many of the other digital achievements of the film are more subtle (see Figure 1).

OPEN IN VIEWER

Figure 1.

Forrest Gump (1994) – Forrest (Tom Hanks) receives a Medal of Honor from President Johnson. Hanks is integrated into archival footage throughout the film. Screen grab from DVD (Forrest Gump, Robert Zemeckis).

It is easy to overlook the other digital manipulations in Forrest Gump. To cite one example: when the film was entered into the National Registry in 2011, the Library of Congress cited ‘its technological innovations (the digital insertion of Gump seamlessly into vintage archival footage)’ (The Library of Congress, 2011) with no acknowledgment of other techniques utilized in the film. Kagan (2003: 153) marvels at ‘Forrest Gump’s special effects, which are so understated and yet ever present that they subtlety [sic] create a universe as understated as Roger Rabbit.’ Effects like the recurring motif of the floating feather, the digital erasure of Lt Dan’s (Gary Sinise’s) amputated legs, and computer-generated crowd scenes in football stadiums and at the National Mall demonstrate the potential for digital effects that do not announce their presence as explicitly as the manipulated archival footage elsewhere in Forrest Gump, or those seen in other Zemeckis films like Who Framed Roger Rabbit or Death Becomes Her.

The subtlety of the effects utilized in Forrest Gump was a cause of concern for scholars like Dixon or Jon Lewis. Lewis, in his essay ‘The Perfect Money Machine(s): George Lucas, Steven Spielberg, and Auteurism in the New Hollywood’ (2007), strikes an alarmist tone when he writes about Forrest Gump:

Spielberg (who produced the film) reveals what he and Lucas (who postproduced it) have been up to all along: what they’re really interested in is the possibility of making movies without human actors – that is, in simulating the production phase entirely in a postproduction studio. (p. 70)

Lewis’s anxieties are especially prescient given the performance capture technology that Zemeckis would embrace in the following decade. In promoting and explaining the technology, figures like Zemeckis and Serkis often address the concerns that Lewis voices here. Such digital technologies, they both argue, do not supplant or replace humans, like the dystopian filmmaking future that Lewis envisions; instead, the technology provides new opportunities to augment and expand the range of possibilities for human actors to practice their craft.

In 1997, Zemeckis founded ImageMovers with Jack Rapke and Steve Starkey, who served as second unit director and associate producer on many of Zemeckis’ films. The studio would go on to release two films in 2000, both highly successful: What Lies Beneath, a supernatural thriller starring Harrison Ford and Michelle Pfeiffer, and Cast Away, with Zemeckis working again with Tom Hanks. Like the above examples in Forrest Gump, Cast Away was a film that utilized a large amount of digital effects in less overt ways. In an interview with Jeffrey Ressner, Zemeckis acknowledges:

Oh, we used a lot on Cast Away. Huge. Way more [than Forrest Gump]. It’s like everything was a digital shot. We had whole fleets of ships that we had to remove. We had other entire islands that were in the shots that we had to remove. (Ressner, 2013)

Though invisibility of digital effects would be a recurring motif in Zemeckis’ promotion of many of his films, during his performance capture period, the technology would become almost unavoidably visible, and for many critics, this would be the ultimate failing of this time in Zemeckis’ career.

Next stop: The ‘uncanny valley’

Tom Hanks, star of Forrest Gump and Cast Away, wanted to adapt Chris Van Allsberg’s The Polar Express, a Christmas themed children’s book, with the intention of playing the titular train’s conductor. Zemeckis was interested in the project, but had an ambitious vision for the film’s style: he wanted ‘to replicate as closely as possible the look and impressionistic artistry of the original short book … while relying entirely on live performances from the actors’ (Creative Planet Network, 2012). Zemeckis worked with his long-time collaborator Ken Ralston, the Academy Award winning visual effects supervisor for many of Zemeckis’ films, who was now president of Sony Pictures Imageworks. Started in 1992, Sony Pictures Imageworks had worked previously with Zemeckis on Contact and Cast Away, and won an Academy Award in 2005 for their work on Sam Raimi’s Spider-Man 2 (2004), a film that utilizes a digital ‘actor’ for many of the action set-pieces. Zemeckis and Ralston began experimenting with technologies for achieving the desired effect of utilizing ‘live performances from the actors’ in animation. They settled on motion capture (also known as mo-cap) as the best option for the aesthetic Zemeckis and his team wanted, though they also set out to make some refinements to the technology.

Alberto Menache, who worked on The Polar Express, defines the process in his book Understanding Motion Capture for Computer Animation (2011):

Motion capture is the process of recording a live-motion event and translating it into usable mathematical terms by tracking a number of key points in space over time and combining them to obtain a single three-dimensional representation of the performance. (p. 2)

As Scott Balcerzak explains in his chapter ‘Andy Serkis as Actor, Body and Gorilla: Motion Capture and the Presence of Performance’, ‘mo-cap takes one essential externalised aspect of performance (realistic movement), separates it from the body, and uses it to guide the special effect’ (Balcerzak, 2009: 196). Even in this early promotional discourse, describing the motivations that led the team behind The Polar Express to work with performance capture, we see evidence of potential distortions in the technology’s purpose. As Balcerzak writes, these technologies are

not about a digitisation of the human, but the humanisation of the digital through the addition of supposedly real movement. It is a process developed to make the special effect perform realistically as opposed to, as suggested by many, digitally enhance the actor. (p. 196)

Zemeckis is among those who see the technology this latter way; the narrative put forth to explain performance capture’s genesis seeks to naturalize and humanize the technology by foregrounding the potential opportunities it presents for filmmakers and actors.

Although mo-cap technology seemed well suited to The Polar Express, the major hurdle that Ralston and his team faced was integrating all components of the actor’s performance. Motion capture is generally performed in special bodysuits, covered in markers that are used to capture the movements. As Balcerzak explains:

Multiple lights reflect off these markers and, then, these reflections are captured and fed into a computer as the performer moves around the capture stage. All of this information is applied to a virtual 3-D body, which is then instantly mapped onto a kind of digital puppet. (p. 195)

According to Jerome Chen, senior visual effects supervisor on The Polar Express, motion capture at that time was generally split between capturing the performer’s bodily movements and facial expression in two separate processes that had to be combined after the fact. This approach proved unsatisfactory, because

it was too difficult to get the performance to be seamless. That’s when we realized we had to get it in a single pass. This had never been done before to the fidelity that we wanted for this project. That is where the moniker of ‘performance capture’ began. (Creative Planet Network, 2012)

In theory, the performance capture in The Polar Express would preserve the entire range of the actor’s performance, capturing what Ralston describes as ‘that serendipitous, accidental, sudden thought of an actor, the subtlety of what we do when we speak and move’ (Creative Planet Network, 2012). Although the filmmakers hailed the technology for its fidelity to the actor’s performance, it also allowed the production to move beyond the limitations of Hanks’s body and appearance. Hanks, filtered through the digital intermediaries of performance capture, plays The Conductor in the film, as he had originally intended; the resulting computer-generated character bears a striking resemblance to Hanks. But the technology also allowed Hanks to take on additional roles, playing characters that do not look like him. Hanks ultimately plays many of the major roles in the film, including the young child protagonist (see Figure 2). The technology also allowed for a compositing of performances; although most of the performance capture that produces the young boy is Hanks’, additional capture work was performed by a young Josh Hutcherson, and the character’s voice is provided by Daryl Sabara.

OPEN IN VIEWER

Figure 2.

The Polar Express (2004) – The Conductor (Tom Hanks) face-to-face with the Hero Boy (also Hanks, voiced by Daryl Sabara). Performance capture allowed Hanks to play most of the major roles in the film. Screen grab from DVD (The Polar Express, Robert Zemeckis).

This fragmentation of acting is inherent to motion capture and performance capture work, and has been explored by several scholars, who examine how it problematizes film acting. Balcerzak (2009) acknowledges that the fragmentation of acting has long been a key component of film production, though he argues that mo-cap goes far further than previously possible; he writes:

As a performer forced to emote in fragments, the film actor has always had less power in determining onscreen performance than what is promoted in the popular Stanislavski-influenced discourse on acting. With mo-cap, we see this discrepancy widen as the actor is literally stripped of his physical body to exist as pure kinesis – a marker cloud to be employed as a tool by the filmmaker. (pp. 210–211)

Tanine Allison, in her essay ‘Blackface, Happy Feet’ (2015), discusses the sort of composite performances present in The Polar Express (and which would be utilized in Zemeckis’ later performance capture productions as well). She writes

Motion capture allows performances to be divided from the star, with multiple performers creating only fragments of the whole. This piecemeal characterisation flies in the face of those – like Serkis, Cameron or director Peter Jackson – who would like to see the digital character as a direct extension of the human performer, transmitting every nuance of expression and intention. (p. 123)

Zemeckis’ promotion of performance capture focuses on the exciting potential benefits for actors while seeking, as Allison notes, to downplay the piecemeal and extremely mediated nature of performance capture acting.

Zemeckis and the team at Sony Pictures Imageworks were ambitious about what performance capture could accomplish, but the aesthetics achieved in The Polar Express were not well received by film critics. The ‘uncanny valley’ theory provides a compelling explanation for this. The theory originates in a 1970 paper by Japanese robotics researcher Mori Masahiro. ¹ Mori suggested that as robot design developed more toward the appearance of life-like humans, there would be would be a range where certain appearances – those that were close to, but not quite at the level of life-like – would be fundamentally off-putting. In effect, an almost perfect human representation is less desirable than more abstracted designs (Mori, 2012: 98–100). Dan North provides a succinct summary of the underlying psychological mechanisms central to Mori’s ‘uncanny valley’: ‘The digital construction of a face delivers an image which is fascinating precisely because it is “not quite right.” The human face sets off a series of perceptual reflexes in the viewer every time it comes into view’ (North, 2008: 153). These ‘perceptual reflexes’ struggle with figures like the human characters of The Polar Express, who straddle the line between visibly artificial and visibly human; this effect is Mori’s ‘uncanny valley’.

Although Mori’s theory dates back to 1970, in recent years the concept has become more widely discussed, probably as the technology to produce lifelike representations of humans has advanced greatly. Frank Pollick, in his essay ‘In Search of the Uncanny Valley’ (2009), suggests that the term’s popularity in the context of computer-generated animation stems from Final Fantasy: The Spirits Within (2000, dir. Sakaguchi Hironobu), a computer-generated film loosely based on the long-running series of Japanese role-playing games. Pollick writes:

The uncanny valley entered the popular lexicon not long after the full-length feature film Final Fantasy appeared … The audience response was lukewarm and a general consensus began to evolve that it failed due to falling into the uncanny valley. (p. 71)

North (2008: 154) offers a similar assessment of Final Fantasy’s failings:

The factors inhibiting the viewer’s acceptance of the characters in Final Fantasy, are, I believe, due to their uncanny humanness; occasionally the movements of the characters are so ‘true’ that the mind alternates, as in Todorov’s conception of the Fantastic, between belief and disbelief in their reality, or rather, their indexicality … The balance between the visibly false and the partially realistic has been upset.

Grounding his analysis in Tzvetan Todorov’s theorizations on the ‘Fantastic’, North explains that the characters of Final Fantasy were unsettling for many viewers because of how we perceive representations of the human form.

Although the aesthetics of Final Fantasy were an issue for many viewers, it would be a mistake to blame the film’s failure solely on the uncanny valley. Critical response to the film was varied; several critics, including Roger Ebert, praised the visuals of the film, which he described as the primary appeal of the film: ‘The reason to see this movie is simply, gloriously, to look at it.’ Ebert’s main criticisms are leveled at the muddled plot; trying to understand how the film’s aliens operate, he writes ‘either I got confused on that point, or the movie did’ (Ebert 2001). Though the ‘uncanny valley’ is not solely responsible for the film’s poor performance, those involved in the production of Final Fantasy: The Spirits Within have acknowledged the peculiar and often unsettling aesthetic of the film; Andrew R Jones, who worked as the film’s animation director, reflected that ‘it can get eerie. As you push further and further, it begins to get grotesque. You start to feel like you’re puppeteering a corpse’ (Weschler, 2002).

Unfortunately for Zemeckis, The Polar Express has become one of the canonical examples of the ‘uncanny valley’. When the concept needs explaining, Zemeckis’ film is frequently cited; for example, in the 30 Rock episode ‘Succession’ (Season 2, Episode 13), Frank (Judah Friedlander) holds up Mori’s diagram of the ‘uncanny valley’ and explains the nadir as ‘Tom Hanks in The Polar Express’. Though the term itself does not appear in the critical reception of the time, the language used to describe the performance capture aesthetic resonates with Mori’s theory. Several of the terms that Jones uses in describing his experiences from Final Fantasy appear repeatedly in the critical response to The Polar Express’s aesthetics. ‘Eerie’ – a synonym for ‘uncanny’ – appears in Manohla Dargis’ review in The New York Times – ‘the eerie listlessness of those characters’ faces’ (Dargis, 2004) – as well as Kurt Loder’s review for MTV – ‘ upon close contemplation of the characters here, with their smooth, doll-like skin and count-every-strand hair, another word creeps into mind: eerie’ (Loder, 2004). Roger Ebert speaks highly of the film in his four-star review, but still deploys much of the same language, describing the aesthetics as ‘a little creepy. Not creepy in an unpleasant way, but in that sneaky, teasing way that lets you know eerie things could happen’ (Ebert, 2004).

Jones’s ‘puppeteering a corpse’ is echoed in animator Ward Jenkins’s (2004) critique of the film when he asks why the characters ‘look so dead and puppet-like’. Similar corpse and zombie comparisons are seen in Peter Travers’s (2004) review for Rolling Stone – ‘the characters … have a glazed look that is almost spooky in an Invasion of the Body Snatchers kind of way’ – and Paul Clinton’s (2004) review for CNN, which states that the film ‘should be subtitled The Night of the Living Dead’. Jessica Aldred, in her essay ‘All Aboard The Polar Express’ (2006), states that, ‘the critical response to The Polar Express … suggests that this viewer fluctuation between belief and decipherment, wonder and scrutiny, partly defined the film’s reception’ (p. 158). Not all critics saw Zemeckis’ first performance capture film as a failure – Ebert’s four-star review is quite laudatory, for instance – but others had difficulty with the aesthetics of the film, with the human characters produced through performance capture being uncannily off-putting.

Defending and refining the technology

Although the critics were unconvinced by his performance capture aesthetic, Zemeckis would continue with, and actively advocate for, the technology. It was not the first time he had faced skepticism; as Kim Masters (2012) notes, ‘When he invented a new way to meld live-action and cartoon characters in Who Framed Roger Rabbit, the animation community predicted the picture would be “the Ishtar of animation.” That technology now is standard.’ Throughout his performance capture period, Zemeckis served as a vocal advocate for the technology in interviews and promotional materials for his films. Zemeckis framed much of his praise for performance capture around the ways that it could enhance the craft of film acting, which follows from the technology being employed and refined with the aim of preserving the human performances of The Polar Express.

In an interview with Masters for The Hollywood Reporter, Zemeckis boasts,

You still get the magic and emotion of the human performance, and the actor gets to do things that he or she never gets to do in movies. First of all, they get to do the scene in continuity, from beginning to end, with all the other actors. They get to pace the scene just like they were doing it on the stage. They’re running the scene, feeding off each other. (Masters, 2012)

Zemeckis further enumerates changes in film recording technologies brought about by digital film production:

The traditional, hundred-year-old optic, chemical, mechanical way in which we record movie images is changing … It will be a language influenced by the artistry of video games and the internet – a whole new way of how we use images to communicate. (Aldred, 2006: 154)

In an article published by The New York Times before the release of The Polar Express, Zemeckis enumerates a long list of limitations in traditional filmmaking, limitations that are not present in his digital approach:

Without the tyranny of hitting marks and leading the lights and worrying about the boom shadow and your makeup and your wig and the line on your wig and all that horrendous stuff that stifles an actor’s performance. Or when they do the greatest take ever and they miss the focus. (Kehr, 2004)

The language Zemeckis uses when discussing performance capture is consistently grandiose. In an interview with Harry Knowles, Zemeckis states, ‘My goal here is to present this art form to tell stories that we never had a way to do before … We are never going to replace actors, we actually liberate actors’ (Knowles, 2009). To Zemeckis, performance capture represents its own ‘art form’. He speaks of liberation and tyranny; in a fairly literal sense, he identifies this technology as revolutionary. Simultaneously, he also works to assuage concerns about the technology’s impact; performance capture as defined by Zemeckis in these interviews and promotional materials offers exciting new possibilities for both film production and film acting.

Zemeckis’ next work with the ‘art form’ was Monster House (2006), directed by Gil Kenan, with Zemeckis and Spielberg serving as executive producers. The performance capture in Kenan’s film was better received, with several critics positively comparing the film’s more stylized aesthetic to The Polar Express. Ian Freer writes in Empire:

Whereas the photorealism of Polar Express created a soulless, almost dead-eyed quality to its characters, the obviously animated heroes of this movie, all oversized features and Play-Doh hair, are ironically more believable than many live-action child stars. (Freer, 2006)

AO Scott’s (2006) review in The New York Times offers a similar comparison:

The people in the movie look a little like molded-plastic figurines (the ones in Polar Express looked more like porcelain dolls), but their gestures are uncannily fluid and unpredictable, making you appreciate the quality of the acting more than you generally do in animated films.

In Scott’s review, we see language familiar from the critical commentary on The Polar Express, although his use of words like ‘uncannily’ is less negative than in the reviews of Zemeckis’ earlier film. This distinction that reviewers like Freer and Scott make in contrasting the aesthetics and stylization of The Polar Express and Monster House reflects the central role that the uncanny plays in the reception of these performance capture films. Lisa Bode, in her work on the uncanny effect of digitally-achieved posthumous performances, focuses on how the uncanny was theorized by Ernst Jentsch and Sigmund Freud; she writes: ‘Most of the phenomena that both Jentsch and Freud agree are likely to trigger the uncanny are things that elicit a momentary uncertainty in the beholder as to whether they are animate or inanimate’ (Bode, 2010: 59). The Polar Express falls within this liminal space of indeterminate animacy, Mori’s uncanny valley, because of its (attempted) photorealistic aesthetic; Kenan’s Monster House features a set of more stylized characters, which do not generate the same sense of dissonance as Zemeckis’ earlier film (see Figure 3).

OPEN IN VIEWER

Figure 3.

Monster House (2006) – DJ (Mitchel Musso), Jenny (Spencer Locke), and Chowder (Sam Lerner) exploring the haunted house. The more stylized character designs helped to keep Monster House out of the uncanny valley. Screen grab from DVD (Monster House, Gil Kenan).

Although the critical response suggested that Zemeckis’ approach to performance capture had been improved upon by Kenan, the praise that Monster House received also provided support for Zemeckis’ stated goals for performance capture and how it could benefit film acting. Scott’s appreciation for the visibility of the actors’ performances is, according to Zemeckis, the intent of the technology. The performance capture technology developed for The Polar Express was intended to preserve the live actors’ performances; reviewers like Scott explicitly call attention to how the acting that serves as the foundation for the animation can be appreciated in the final product.

Zemeckis would return to directing performance capture the following year with Beowulf (2007); the critics remained split on Zemeckis’ aesthetic choices, although many noted that refinements to performance capture technology produced better results than The Polar Express. One example of a positive assessment was Ty Burr in The Boston Globe (2007):

In The Polar Express, the result was a cast of zombie children and a creepy, soulless Tom Hanks. The good news is that the technology has improved and that the cast of Beowulf merely looks like they have the squints.

On the other side, Justin Chang (2007) criticizes

the way Zemeckis prioritizes spectacle over human engagement, in his reliance on a medium that allows for enormous range and fluidity in its visual effects yet reduces his characters to 3-D automatons. While the technology has improved since 2004’s Polar Express (particularly in the characters’ more lifelike eyes), the actors still don’t seem entirely there.

Beowulf also further developed the composite digital bodies seen in Hanks’ roles in The Polar Express, offering new ways to ‘liberate’ (to borrow Zemeckis’ rhetoric) actors from certain pro-filmic realities. In interviews, Zemeckis speaks of the theoretical applications and advantages of performance capture, which provide ways to overcome limitations of film production. Lisa Bode (2010) points to another example of this sort of discourse, in this case employed by one of the film’s stars; she notes, ‘John Malkovich has spoken glowingly in interviews of how, “like doing a play,” the process allowed him to sustain his characterization rather than offering it in pieces for traditional camera setups’ (p. 64). In Beowulf, the production possibilities opened up by performance capture allow for certain practical limitations of film acting to be overcome, with the most dramatic limitation being the very bodies of the film’s stars. Yacov Freedman (2012: 40) explains that in Beowulf, ‘the title character was an amalgam of Ray Winstone’s head atop a bodybuilder’s physique, and Angelina Jolie, who was pregnant during the film’s production, had her performance grafted onto a digital model of her prepregnancy body’ (see Figure 4).

OPEN IN VIEWER

Figure 4.

Beowulf (2007) – Beowulf (Ray Winstone) exploring Grendel’s cave. The character’s body was an amalgamation of Winstone and a bodybuilder model. Screen grab from DVD (Beowulf, Robert Zemeckis).

This digital synthesis of the body raises questions about the effects of this technology; despite Zemeckis’ assurances in interviews that performance capture is a boon to actors, the technology represents a significant disruption to established practices. Lisa Bode, for example, has written extensively on one particularly provocative use of the digital recreation of actor’s bodies: the resurrection of deceased stars. Even more than the ‘uncanny valley’ effect of Zemeckis’ performance capture films, such resurrected performances are often disquieting or outright disturbing for viewers and critics; Bode (2010: 52) highlights several recurring terms in the discourse surrounding such performances – ‘“Frankensteinian,” “ghouls,” and “dead-eyed zombie”’ – that will be familiar from the discourse deployed by critics in responding to The Polar Express. Bode (2007: 37) also notes that such digital resurrection potentially hurts the market value of still-living actors: ‘screen actors are pretty resistant to the idea of digital resurrection, as their labour market is competitive enough without having to jostle with the dead for roles.’ But the performance capture films of Zemeckis highlight that such a threat to actors – despite Zemeckis’ claims to the contrary – are not limited merely to resurrecting the stars of yesteryear. The Polar Express feels at times like a one-man show starring Tom Hanks and, as Barry King notes in ‘Articulating Digital Stardom’ (2011), Beowulf star Ray Winstone ‘acquired a toned, taller and youthful body to play a role that would have gone to a younger actor in a non-digital environment.’ King goes on to note the costs of the digital ‘liberation’ that established stars like Hanks or Winstone can enjoy from this form of production:

Nor is the opportunity to inhabit a cool digital body entirely an expression of freedom when viewed from the perspective of young, up-and-coming actors, who rely on their natural physical assets to be cast in a particular kind of role. (p. 258)

Despite serious concerns about performance capture and its potential to harm the value of acting labor, Zemeckis remained focused on promoting and defending the technology, and would enter into a business arrangement that allowed him to continue with performance capture productions.

Disney and the ImageMovers Digital deal

Performance capture received, at best, a mixed reaction from the critics, but Zemeckis had some prominent supporters in the industry. The most significant backing came early in 2007, prior to the release of Beowulf, when Dick Cook, chairman of Walt Disney Studios, established a joint venture between Disney and Zemeckis’ ImageMovers. The new animation facility, ImageMovers Digital, was dedicated to producing performance capture films.

Disney’s support of Zemeckis was part of the company’s efforts to reorient toward computer-generated animation. After the industry dominance Disney enjoyed in the late 1980s and ‘90s, the 2000s was a significantly less profitable period for the company’s hand-drawn features. CG animated films like Pixar’s Monsters, Inc. (2001, dir. Pete Docter), DreamWorks’ Shrek (2001, dir. Andrew Adamson and Vicky Jenson), and Fox’s Ice Age (2002, dir. Chris Wedge and Carlos Saldanha) significantly outperformed Disney’s Atlantis: The Lost Empire (2001, dir. Gary Trousdale and Kirk Wise), Lilo and Stitch (2002, dir. Chris Sanders and Dean DeBlois) and Treasure Planet (2002, dir. Ron Clements and John Musker). At the time of the ImageMovers Digital deal, Disney was struggling with its first forays into producing their own computer animated films, though the 2006 purchase of Pixar would help significantly. In hindsight, the Pixar deal might seem more than sufficient, but in 2007 Disney was still adjusting to computer animation production, and so taking a chance on Zemeckis’ performance capture technology was part of Disney’s diversification strategy.

With support from Disney, ImageMovers Digital’s first production was A Christmas Carol (2009), starring Jim Carrey, Gary Oldman, Bob Hoskins, and Cary Elwes, all of whom played multiple roles in the film. This would be followed by Mars Needs Moms, which would be both ImageMovers Digital’s last production and the final film of Zemeckis’ technologically determined decade. Like Monster House, Zemeckis served as executive producer on Mars Needs Moms. The film, based on a children’s book by Bloom County cartoonist Berkeley Breathed, was directed by Simon Wells, who had served as supervising animator on Who Framed Roger Rabbit, and had previously worked for Amblimation, the short-lived animation production arm of Spielberg’s Amblin Entertainment, and Dreamworks.

Mars Needs Moms would be ImageMover Digital’s final production, and (so far) the final performance capture film for Zemeckis. One year before, Disney pulled its support of ImageMovers Digital, due in large part to new management at Disney. In September 2009, two months before the release of A Christmas Carol, Cook stepped down as chairman, to be replaced by Rich Ross, former president of entertainment at the Disney Channel. By the time of Mars Needs Moms’ release, Disney was also on much firmer footing in terms of its computer-generated animation production. The huge investment in Pixar had paid off, and Disney’s own computer animated films were starting to perform better; Tangled (2010, dir. Nathan Greno and Byron Howard) was the first non-Pixar animated hit Disney had had in over a decade. In 2007, Zemeckis’ performance capture was worth the gamble for Disney; by 2011, Disney cut its losses. ImageMovers Digital had plans for future films – a remake of Yellow Submarine (1968, dir. George Dunning) was far enough in pre-production to have actors attached – but these would be shelved as the company shut down. After the closure of ImageMovers Digital, Zemeckis returned to live action films, releasing Flight in 2012 and The Walk in 2015. Zemeckis’ performance capture period came to a close, but the technology would continue to be developed and refined by others; the vision for performance capture that Zemeckis enunciates in interviews may not have come to fruition for his own films, but the future is bright for the technology.

Comparing rhetoric: The case of Andy Serkis

So far, this article has focused on Robert Zemeckis’ performance capture films, charting both the discourse that others used in defending and defining the technology and the largely negative critical reception these films received. Although Zemeckis’ films were largely unsuccessful, the technology of performance capture has flourished in Hollywood. The technology has been embraced by such A-list directors as James Cameron in Avatar, Steven Spielberg in The Adventures of Tintin (2011), and Peter Jackson in the Lord of the Rings trilogy, King Kong, and the Hobbit trilogy (2012, 2013, 2014). It is because of how ubiquitous the technology has become in recent years that examining Zemeckis’ performance capture period is important.

Andy Serkis has become the most visible example of a performance capture star. As with my examination of Zemeckis, my focus here on Serkis should not be taken to occlude or downplay the work of the myriad digital artists and other technical personnel that made Serkis’ roles possible – for example, most of his acclaimed roles are the work of Weta Digital, one of the premiere digital effects companies in the industry today. My interest is less in the actual process and labor that go into Serkis’ performance capture roles, as in his promotional labor as he explains his relationship to the technology. Serkis serves as a useful counterpoint to Zemeckis, a figure whose performance capture work has garnered much attention, both popular and scholarly. If Zemeckis’ years spent developing and refining the technology are to be ‘regretted’ (as Dargis suggested), Serkis’ career charts a different trajectory. Like Zemeckis, Serkis is a vocal advocate for performance capture, and it is illustrative to compare how the technology is understood and discussed within the context of Serkis’ career, and the rhetoric that both men employ in defending and defining performance capture.

Although Serkis regularly acts in live-action roles, his fame stems primarily from his work with motion and performance capture. His breakout role was Gollum in Jackson’s Lord of the Rings films, performed using pre-Polar Express motion capture. As Serkis details on his website, separate facial capture was used only as a reference:

After much discussion about whether to go down the facial capture route it was decided that there wasn’t going to be enough time to get the system functioning quickly enough before the film’s deadline, and in the end it was not pursued after all and Gollum’s face would be key-frame animated. (Serkis, nd)

Performance capture technology is utilized for Serkis’ later roles, including Kong, Captain Haddock in The Adventures of Tintin, and Caesar in Rise of the Planet of the Apes.

Whereas the reception to Zemeckis’ films shows that many critics struggled with accepting the technology, the response to Serkis’ acting is different. The technology is not rendered invisible, but in Serkis’ performances the line between the actor and the technological intervention is claimed to be blurrier. Balcerzak (2009) states, ‘The filmmakers and other supporters of [performance capture] have certainly tried to sell it as simply a new form of acting – a kind of inevitable evolution for the art of performance in cinema’ (pp. 195–196). The response to one of Serkis’ recent roles, Caesar in the new Planet of the Apes films, suggests that critics have accepted this argument. In his New Yorker review, David Denby (2011) explicitly rejects a distinction between performance capture and traditional acting:

Digitized acting (if that’s the right phrase) should be as warmly recognized as any other kind of acting … To register the moment, Serkis lengthens his jaw in sullen resolve, turns his back, and gives Caesar a regretful shudder – the scene is almost tragic.

Dargis, who was so critical of Zemeckis’ efforts, notes that Caesar is ‘given nuance through performance-capture technology … and the efforts of Andy Serkis … When Caesar scowls, as he increasingly does, you don’t see just digital wizardry at its most expressive; you also see a plausible, angry, thinking character’ (Dargis, 2011). Ebert (2011) demonstrates how easy it can be to blur the line between Serkis and the technology: ‘One never knows exactly where the human ends and the effects begin, but Serkis and/or Caesar gives the best performance in the movie’ (see Figure 5).

OPEN IN VIEWER

Figure 5.

Rise of the Planet of the Apes (2011) – Caesar (Andy Serkis) adjusting to his new environment after being placed in a primate shelter. Critics credited Serkis for the emotional expressiveness of Caesar. Screen grab from DVD (Rise of the Planet of the Apes, Rupert Wyatt).

As Lisa Bode notes in ‘Fleshing It Out’ (2015: 40), ‘many reviewers praise Caesar for his unprecedented levels of “expressivity,” in comparison to previous motion-captured performances.’ The critical reaction to Serkis-as-Caesar demonstrates the progress that performance capture technology had made since Zemeckis’ films, but it also suggests an overly simplified explanation for what makes Caesar successful. Bode notes, ‘if [Gollum] and Caesar are both so compelling on screen compared to performance-capture characters in Beowulf or Avatar, then Serkis, as the common denominator between the two, must be responsible’ (p. 40). Such a response is tempting, in part because it falls within prevailing discourses on film acting. But it also crucially ignores another common denominator: Weta Digital, the company behind both Gollum and Caesar. The critical response privileges Serkis, as master of the technology in ways previous stars like Hanks in The Polar Express or Winstone in Beowulf were not; this shifts the focus away from the technological developments and artistic labor of those at Weta Digital making characters like Kong and Caesar. The critics are not alone in this; Serkis, and others involved in the promotion of these films, reinforce a narrative that focuses on the craft of acting.

The critical attention that Serkis’ performances have garnered has also led to attempts to have his work recognized by the Academy. New Line Cinema launched a nomination campaign in 2003 for Serkis’ work in The Two Towers, ‘the first attempt to get a role driven by performance capture acknowledged at the Academy Awards’ (Leaver, 2014). More recently, 20th Century Fox lobbied for nomination for Serkis’ roles in Rise of the Planet of the Apes and its sequel Dawn of the Planet of the Apes (2014, dir. Matt Reeves) (Feinberg, 2014). Despite Zemeckis’ efforts, it is Serkis who has become one of the premier ambassadors for performance capture; the acclaim he has received for his acting helps to increase acceptance of the technology.

In interviews and other promotional materials, Serkis advocates for performance capture technology, explaining how it works and assuaging concerns some in the industries may have. In a short piece for Empire magazine which ran alongside an interview with Zemeckis, Spielberg, and Cameron, Serkis offers his perspective on the technology and what it means for acting. Like Zemeckis, Serkis defines the appeal of the technology largely in terms of what it offers to acting:

Love it or hate it, we have to acknowledge that performance capture is not going away, and that it will undeniably be remembered as a valid and important component in the evolution of the ancient art, craft and tradition of acting. (Serkis, 2010)

Both Zemeckis and Serkis frame their rationale in similar terms, advocating for what the technology offers actors. A distinction between the two is that Zemeckis often talks about performance capture as a radical departure from conventional acting, freed from familiar conditions and constraints. Zemeckis may see this as a positive, but for many in the industry the newness of performance capture, and its potential to realign the rules and priorities of film acting, is what they fear; Zemeckis’ rhetoric merely serves to further exacerbate these anxieties. Serkis, on the other hand, is more traditional in his rhetoric. Writing about the promotional production diaries filmed for King Kong, Tanine Allison (2011: 333) argues that, ‘Serkis, along with the rest of the crewmembers interviewed in the diaries, straddles the line between hyping the new and relying on older, more familiar discourses of cinema and acting.’

According to Balcerzak (2009: 198): Andy Serkis’s discussion of an ‘actor’s choices’ driving his performance as Kong is nothing new in the popular discourse of acting. The performer’s free will, to varying degrees, has been a mainstay in discussions of performance for over a century now, beginning with Constantin Stanislavksi’s experiments at the Moscow Art Theatre at the turn of the previous century.

By aligning his performance capture work with legitimated antecedents in acting, Serkis works to naturalize the technology and to neutralize its disruptive potential; in one example, he goes so far as to evoke centuries and millennia-old acting traditions: ‘Just as the Ancient Greeks wore masks, or the Japanese performed kabuki, so Zoe Saldana dons a motion-capture suit with markers’ (Serkis, 2010).

Like the mask analogy Serkis deploys here, a common point of comparison is prosthetics. Arguing that performance capture represents a form of ‘digital prosthesis’ serves to ground performance capture as an outgrowth of existing, accepted acting practice; as Bode (2015: 33) writes, ‘The term “digital prosthesis” has been pushed by actors and directors in the commercial film industry seeking to frame performance capture as a form of legitimate acting, by referencing selective examples from cinema’s analogue past.’ The prosthesis analogy, based on accepted and legitimated frameworks for understanding and critiquing an actor’s performance, is appealing for critics. Bode points to the ways that critics, ‘trying to understand performance capture in relation to what is familiar, locate Serkis on screen “underneath” Caesar or as providing Caesar’s “skeleton and soul”’ (p. 35).

Zemeckis, on the other hand, is a less successful advocate for performance capture because his films can be seen to confirm fears of what the technology will do to acting.

Interviewing Zemeckis, Spielberg, and Cameron for Empire, Ian Freer prompts the three directors by saying, ‘One of the common misconceptions with the technology is that it will replace actors, or that you can take a performance and add in ten per cent Brando, 50 per cent Clooney’ (Freer, 2010). Cameron and Spielberg both challenge this as a misconception, but Zemeckis is silent on this point. Spielberg cites Zemeckis’ work – ‘In A Christmas Carol, it’s 100 per cent Jim Carrey’ (Freer, 2010) – but this is cherry picking. Other Zemeckis films demonstrate the potential of performance capture to both replace actors and generate composite performances; The Polar Express, with Tom Hanks performing in most of the film’s major roles, and Beowulf, with its digitally amalgamated characters, both clearly signal ways that performance capture can disrupt traditional norms of film acting.

Compared to Zemeckis’ films, there is a reassuring stability to Serkis’ performances. From a technical standpoint a wide array of digital processes make up an ‘Andy Serkis’ performance – as Balcerzak (2009: 201) notes, a performance like Kong is the product of a network of individuals working on the production, including ‘mo-cap editors, animators and, of course, Peter Jackson who oversees the entire process’. But in the promotional discourse that Serkis puts forth, such details are frequently glossed over in favor of a narrative that conforms to prior conceptions of the craft of acting. Audiences are still encouraged to marvel at the technological apparatus that allows such performance capture roles possible – effects, both practical and digital, have long served as appealing spectacles – but they are also primed to focus on Serkis as the creative force behind such performances.

Serkis’ role in performance capture is not only as an actor: in 2011, he and Jonathan Cavendish founded a performance capture company, the Imaginarium Studios. Combating anxiety about performance capture is one of the goals of the company; in an early announcement, Serkis described it as

a new performance capture facility he hopes will teach filmmakers not to be worried about the growing use of graphics … Part of the idea of the Imaginarium is to allay fears – across lots of different communities, not just the acting community – who have in the past wondered if they are going to be replaced by CG characters or robots or whatever. (Paste Magazine, 2010)

The company has helped advise on the performance capture utilized in Rise of the Planet of the Apes, Avengers: Age of Ultron (2015, dir. Joss Whedon), and Star Wars: The Force Awakens (2015, dir. JJ Abrams), all of which feature Serkis as an actor.

Barbara Creed (2000: 83–84) argues that

the presence of the synthespian in film is not meant to be perceived by the audience as a ‘special effect’ nor to draw attention to itself: the virtual or synthetic origins of the star will have to be rendered invisible by the text in order for the character to offer a convincing, believable performance.

The synthespian as Creed defines it here is still far off; as North (2008: 153) observes, ‘At present, though, the state in which the synthespian is made manifest is one of fearful but fascinated mythology, rather than an inevitable path towards a future of artificial entertainers.’ Performance capture, either in films directed or produced by Zemeckis or starring Serkis, has not yet produced a true synthespian, at least one that could meet Creed’s definition. The technology is too new and too overt to ‘be rendered invisible’ yet, and the promotional discourse put forth by Zemeckis, Serkis, and others consistently foregrounds the technological process (or a marketable version of the process). But the discourses that surround performance capture seem to be moving, incrementally, in that direction; the technology has come a long way from the response to The Polar Express and its glassy-eyed zombies to more contemporary reactions that marvel at the creative labor of Serkis and other actors behind performance capture roles. And though Zemeckis may not be involved in future productions, he played a major role in promoting and popularizing a technology that will continue to be central to Hollywood filmmaking for the foreseeable future.

Conclusion

Zemeckis’ performance capture period – what film critic Manohla Dargis (2012) derided as his ‘technologically determined decade’ – spanned from the release of The Polar Express in 2004 to Mars Needs Moms in 2011. Though the aesthetic that Zemeckis and his partners at Sony Pictures Imageworks and ImageMovers Digital developed was generally not well received, performance capture as a technology has been embraced within Hollywood filmmaking, a widely used tool in the creation and performance of digital characters. Kim Masters recently lauded Zemeckis for his seeming foresight: ‘When he became enamored with motion capture for The Polar Express, some derided the “dead eye” look, but the film grossed more than $300 million worldwide, and the technique came into common use’ (Masters, 2012). Zemeckis played a major role in the introduction and promotion of this technique, but his films were less successful than other productions that have employed performance capture since. Dargis’s assessment of this period in his career may prove to be the prevailing perspective, but understanding how Zemeckis understood, defined, and promoted performance capture is important in light of the technology’s ongoing prevalence in big budget Hollywood film production.

The films of the technologically determined decade are well suited to Zemeckis’ interest and fascination with cutting-edge filmmaking technologies; but they also stand apart from many of his other films. After returning to live-action filmmaking with Flight, Zemeckis claimed, ‘I work really hard to try to keep film technique invisible. You want to have the audience get lost in the movie, so they’re not actually paying attention to whether or not the camera is creeping in on someone’ (Ressner, 2013). Films like The Polar Express and Beowulf suggest a different set of priorities for Zemeckis; for better or for worse, the audience and the critics could not help but pay attention to the highly visible style and the new technology behind it. A few years later, however, Zemeckis now positions these films as important learning experiences that allow him to continue to create films with invisible technique. Speaking of Flight, he explains: ‘Everything that I’ve been doing in the digital cinema for the last 10 years actually allowed me to make this movie that no one considers to be digital for $31 million, which happens to have over 300 CG shots in it’ (Masters, 2012).

In his 1995/1996 essay on the digital effects of Forrest Gump, Dixon was troubled by the rising sophistication of these technologies. Presenting some forward-looking predictions of where digital imagery could lead, Dixon writes, ‘It would be nice to suggest that as digital imagery becomes cheaper, the process of creating digital imagery will find a more democratic leveling point, where everyone can make a film because no physical elements are required’ (p. 65). As a major Hollywood player, Zemeckis could never represent Dixon’s proposed democratization of film production, but Zemeckis’ vision for what performance capture and other digital cinema technologies could achieve in terms of altering and improving the production process does suggest some of the potential Dixon writes of. As Freedman (2012: 38) argues, to Zemeckis ‘motion capture represents a wholly new form of filmmaking, one that cannot and should not be limited by our previously held definitions of synthetic versus recorded cinema.’ According to his own rhetoric, Zemeckis’ goal was to develop a new filmmaking paradigm, and his time spent with performance capture – despite its shortcomings – stands as a fascinating period in the career of one of Hollywood’s biggest directors.