Episode II has a certain melancholy about it . . . It appears to be a nice movie with a somewhat happy ending, at least with Anakin and Padmé. It’s only when you put it in the context of the bigger story that you see the handwriting on the wall. You notice the flaws in Anakin’s personality that are going to sink him in the end.
George Lucas
George Lucas began writing Episode II before the first film in the prequel trilogy had even been released. He had a rough outline completed before production began on The Phantom Menace, but in June 1998 he started drafting the actual screenplay. ‘I couldn’t have done Episode I without knowing the complete story,’ said Lucas. ‘It was all mapped out – I just had to write specific scenes. It was still quite a challenge, because Episode II required that I write a love story in the middle of a Star Wars movie. It’s a love story, with the Sith’s relentless drive to take over the universe in the background. The challenge was to balance those two things.’
As ever, writing did not come easily to Lucas. By September 1999 (following a two-month European holiday after The Phantom Menace) Lucas had enough of the spine of the script ready to begin feeding material to the design department. ‘I worked through the first draft as quickly as possible,’ said Lucas, believing it was better to complete a swift draft rather than continually to rewrite the opening twenty or thirty pages. Much of the design work done for Episode I would carry over into the second movie, but there was almost as much again that would have to be created from scratch.
The Skywalker Ranch art department had started work on designing aspects of Episode II in June 1999, a mere month after the release of Episode I. It wasn’t until December that year that they kicked into a higher gear thanks to a detailed scene breakdown from Lucas.
Producer Rick McCallum had set off on his initial location-scouting trip for the second prequel the day after the first opened in US cinemas. He and production designer Gavin Bocquet spent three months touring Tunisia, Italy, Portugal and Spain looking for easy-to-shoot locations that matched environments Lucas had described. McCallum had made several visits to Australia during the final year of production on Episode I to check out the Fox Studios facility in Sydney. The production of the next two Star Wars films would be based there, moving from Leavesden in the UK. The new deal would involve a lease that allowed core sets to be left standing for a substantial period of time – although the Fox facility was in increasing demand.
The biggest change, though, would come in the method of filming. Lucas had been experimenting for a while with the idea of shooting movies digitally, foregoing 35 mm film in favour of high definition (HD) video. This approach had several advantages, prime among them a simplification of the post-production process, especially in the integration of digital visual effects. The 24-P high definition digital camera developed by Sony and Panavision was selected as it was designed to record images at the traditional standard film rate of twenty-four frames per second. The tests convinced Lucas that the digital process retained the look and feel of film, finally deciding in April 2000 to commit. ‘Every single shot we do has a digital effect in it,’ noted McCallum. ‘There’s no point shooting on film [for us]. [Digital] is the way we want to go – it’s just easier and more economical, and the results are fabulous.’
In November 1999 Robin Gurland returned to her role as casting director on the prequels. As shooting was to take place in Australia, she began by filling many of the smaller roles with local talent, drawing on actors from Sydney and Perth. Joel Edgerton (from Australian TV series The Secret Life of Us) was signed up as Owen Lars (a younger version of Luke’s ‘uncle’ Owen, played by Phil Brown in Episode IV) and Bonnie Piesse as Beru Whitesun (again, a younger version of Luke’s ‘aunt’ Beru, played by Shelagh Fraser). As they were playing characters already established within the Star Wars universe, Gurland was looking as much as possible for a physical match as well as for actors who had the skills to inhabit the roles.
Gurland also tapped actors from New Zealand. Episode II would delve into the back-story of one of the most enigmatic (and popular) Star Wars characters: bounty hunter Boba Fett (seen in Episode V and Episode VI previously). Temuera Morrison was cast as Boba’s father, Jango Fett, while Daniel Logan played the younger version of Boba, shaped by tragedy into the character familiar to fans. They were joined by Leeanna Walsman as Zam Wesell, a bounty-hunter who stages an attempt on the life of Amidala.
While many of the key cast – among them Ewan McGregor, Natalie Portman and Ian McDiarmid – would be returning from The Phantom Menace, there was one major, challenging role to fill. Episode II was set a decade on from the first movie, so Anakin Skywalker was now in his late-teens, meaning Jake Lloyd could no longer play the role. The hunt was on for an actor who could take Anakin through the next two movies that would see pivotal changes in his character, primarily his fall to the dark side and his conversion into the formidable form of Darth Vader.
Finding the new Anakin was to be the biggest challenge for Gurland. The search had been ongoing while she filled the other key roles, with an open casting call bringing in around 1,500 submissions from young actors and their agents, all of whom knew the role could be career-making. Gurland viewed some 400 videotape tests, narrowing those down to a selection of around thirty actors whom she thought were serious contenders. At this point Lucas got involved, but neither the writer-director nor the casting director felt they saw the new Anakin among the selection – even though the actors who were auditioned included Ryan Phillippe, Colin Hanks (son of Tom Hanks), Jonathan Brandis (from TV series seaQuest, who would commit suicide in 2003, aged twenty-seven) and later Fast and Furious (2001) star Paul Walker. Lucas and Gurland even met with Titanic (1997) star Leonardo DiCaprio to discuss the role of Vader, but the actor proved ‘unavailable’.
Time was running out and Gurland feared she would never find the right person when she came across Canadian actor Hayden Christensen. Initially struck by his physical presence, she then viewed his work, primarily the Canadian drama Family Passions (which Christensen appeared in when he was twelve) and several mid to late 1990s movies, including The Virgin Suicides (directed by Francis Ford Coppola’s daughter Sofia). Of particular interest was his role in Higher Ground (2000), a Fox Family Channel TV drama in which he played a sexually molested teenager who turned to drug taking. ‘Hayden Christensen convinced us all that not only was he good – he was bad,’ said McCallum of the future Darth Vader. Christensen was officially announced as the new face of Anakin Skywalker in May 2000, just one month before filming was due to begin in Australia.
The new Fox Studios facility in Sydney had been converted from old show grounds at Moore Park into a state-of-the-art facility. The thirty-two-acre site would eventually encompass eight soundstages, production offices, workshops and a service community of up to sixty independent businesses. Several movies and TV series had used the facility before the Star Wars crew moved in, including The Matrix (1999) and the Ewan McGregor-starring musical Moulin Rouge! (2001). Here over sixty sets would be built for Episode II (many more than had been constructed for the previous movie – despite the fact that Lucas had promised his crew that Episode II would be a ‘smaller’ movie).
The administrative team for the film set up a central production coordination office in the hexagonal-shaped, low-level building in the middle of the complex. From here the comings and goings of cast and crew would be arranged, and each morning would begin with the issuing of that day’s ‘call sheets’ showing who and what was required and which scenes were to be filmed. As always on a Star Wars movie, secrecy was of the utmost importance. Windows overlooking the soundstages from the next-door Fox Studios Tour (open to the public) were blacked out. That didn’t stop some enterprising paparazzi gaining access to a next-door sports stadium that also overlooked the studio complex in the hope of grabbing a few shots of the new Darth Vader in action.
Although he had been working on it for a while, Lucas’s final shooting script for the movie was incredibly late, much to the frustration of McCallum. Although much of the pre-production design work and set building had been possible based on story and scene outlines, Lucas’s inability to finish the screenplay caused major problems. British screenwriter Jonathan Hales – who had written for British TV, American soap-opera Dallas and The Young Indiana Jones Chronicles – was drafted in to help wrap up the script for Episode II. The production draft was available less than a week before the start of filming, after the pair of writers spent May and June on further revisions: evidently the romance aspect of the storyline continued to give Lucas problems. ‘I wanted to tell the love story in a style that was extremely old-fashioned, and frankly I didn’t know if I was going to be able to pull it off,’ admitted Lucas. ‘In many ways, it is much more like a movie from the 1930s than any of the others had been, with a slightly over-the-top poetic style . . . I knew people might not buy it.’
The delays with the script were not allowed to slow the work on building sets needed by the production. ‘At that stage, [Episode II] felt like a “virtual film” because we got the script only days before we started shooting,’ recalled McCallum. ‘We had to build these sets to a script that didn’t exist.’ The screenplay would continue to be revised throughout production as Lucas reworked his dialogue.
McCallum was the one operating the clapperboard on the first day of principal photography on Episode II on 26 June 2000, exactly three years to the day since the start of filming on Episode I. For the next three months at Fox Studios and on locations around Europe and beyond, the Star Wars cast and crew would toil on the new movie. Even when they had finished shooting, a further year-and-a-half of detailed post-production work would follow. Much of the work completed on this first day of shooting would be cut from the final film (as on Return of the Jedi). Scenes featuring Ian McDiarmid as Chancellor Palpatine announcing the erroneous news of the assassination of Senator Amidala to the Republic Senate were shot, with Palpatine’s podium against a vast blue screen.
Although he expected problems with the digital HD cameras, McCallum wasn’t prepared for a power failure on the first day. Accepting the teething troubles, Lucas and McCallum decided to view the first few weeks of principal photography as an extended learning period as everyone figured out the best way to use the new technology. While the actors were nervous, especially McDiarmid, who had a lot of dialogue to deliver and no other actors to play against, the camera operators were particularly worried about how the brand new high definition cameras would perform. The decision to go digital was not without controversy (although the majority of major Hollywood movies now shoot this way). ‘People asked, “Why are you doing this?”’ recalled Lucas. ‘The real question was ‘‘Why not?” It was vastly superior in every way, and it was cheaper. You’d have to be nuts not to shoot in this way. As far as I was concerned, we should have been shooting digital cinema twenty years ago.’
Coming to grips with older technology on that first day’s shooting was Anthony Daniels, who suited up as C-3PO for the first time in thirteen years, having only voiced the incomplete skeletal C-3PO puppet in Episode I. Now the droid was kitted out in his familiar gold livery and working as a diplomatic aide to Amidala. Nothing had changed for the actor as he still faced hours confined in an uncomfortable, heavy, armoured suit in which he was virtually unable to walk. ‘I’d forgotten how lonely it is in here,’ admitted Daniels.
Although much of the film would be created digitally, the production of Episode II physically expanded to take up all available stage space at Fox Studios. ‘We were trying to fit all the sets into every square inch of the studio,’ said McCallum. ‘[It] was about half the size of Leavesden. We took over the entire studio, and still it was a tight fit.’ The production’s time at the studio was tightly scheduled into a period of nine weeks from June 2000. The majority of the studio filming would have to be completed during this period, as once the crew left to move on to shoot the locations, a return to Fox Studios would be impossible as other productions were booked to move in. Unlike on Episode I, where the crew were able to return to Leavesden to conduct more filming, any additional requirements for Episode II would have to be managed within the scheduled pick-up filming that would take place back in Britain (maintaining the traditional British connection to the Star Wars saga).
The second week of filming saw the shooting of the opening scene of the movie: the attempted assassination of Senator Amidala on the landing platform at Coruscant. This was one of Natalie Portman’s first shots, working on a minimal set consisting of little more than the spaceship ramp and the immediate landing area, with the rest of the location to be created later digitally. Originally Lucas wanted to shoot the entire scene against blue screen, but was persuaded that having some ‘real’ elements within the scene would help sell the illusion, an important lesson that would go on to inform much of the approach to this movie and Episode III. ‘George only builds what he needs,’ said McCallum of his director’s economical approach. Many of the Coruscant scenes were shot on non-existent or partially realized sets. This allowed for the creation of new scenes – such as the discussion between Anakin and Obi-Wan in an elevator (shot much later at Elstree in November 2001) or a meeting between assassin Zam Wesell and bounty-hunter Jango Fett (shot at Ealing in London in March 2001) – to be created and seamlessly added to earlier material.
An early set-piece sequence in Episode II illustrated this then-new approach to filmmaking. The speeder chase through the skies and skyscrapers of Coruscant, in which the two Jedi pursue Zam Wesell, was the latest indulgence by Lucas of his early days as a boy-racer back in Modesto. The speeders themselves were even tricked out to look like futuristic versions of the 1950s or 1960s hotrods seen in American Graffiti. The script only loosely described the scene, omitting many details. It was more fully developed by the pre-production team working in pre-visualization at the Main House on Skywalker Ranch. Ben Burtt’s ‘videomatics’ drew on the process Lucas had initially developed on the original Star Wars trilogy, using vintage movie clips, newsreels and stock footage, as well as specially shot models to create a primitive visualization of the action. ‘It was a Saturday morning cartoon version of the speeder chase that illustrated what they needed to shoot on stage,’ said Burtt. This in turn was used to create more sophisticated animated storyboards by pre-visualization and effects supervisor David Dozoretz and his four-strong team.
Two months before any live action footage was shot, director Lucas could view this action sequence in some detail and make adjustments. That meant that when it came time to film the live-action components of the scene over two-and-a-half days of blue-screen filming in Sydney, Lucas knew exactly what he needed to shoot. With the animatics available, McGregor and Christensen were also able to get a good idea of what the completed sequence would look like, allowing them to tailor their performances. McGregor noted that the usefulness of the process had progressed immensely. ‘The animatics for the speeder chase were very well developed,’ he said. ‘On Episode I, they were just basic shapes that gave us a rough idea of what was going on, but the animatics for Episode II looked great.’ In addition to the moving animatics, the cast and crew also had a detailed map of the chase worked up by the art department and a series of impressive, colourful concept paintings by artists Ryan Church and Erik Tiemens that showed the different districts (high rent, warehouse, old city and entertainment district) through which the chase would progress.
Shooting the scene in Sydney saw the actors clambering in and out of a full-scale yellow speeder prop mounted on a gimbal that gave them a series of realistic jolts, shocks and an ever-present rocking motion. ‘It actually made us feel rather sick after a while,’ noted McGregor, comparing the experience to spending days on a funfair ride. With the pre-visualization animatics and other illustrative material fresh in their minds, both actors were able to ignore the blue screen surrounding them and put themselves in the middle of the desperate chase through the skies of Coruscant.
Those few days of live-action shooting were not the end of work on that sequence. Pick-up shots filmed much later at Elstree saw McGregor perform the opening moments of the chase, as he leaps from Amidala’s apartment window and clings to an assassin droid before dropping into the speeder piloted by Anakin. Months later again, the artists at ILM would take the edited footage and add in the digitally created backgrounds, other vehicles, creatures, people and all the other hundreds of individual elements needed to complete the scenes. Ben Burtt and Matthew Wood then added the soundscape, creating unique noises for the speeders, other traffic and various other atmosphere elements, before the music by John Williams helped to add further excitement to the chase. The sequence ends in a seedy entertainment district where the speeders crash and the Jedi finally catch up with Wesell. This was filmed on set, with around 150 extras wearing practical alien masks created the traditional way by creature supervisor Jason Baird. Some of the alien heads had even been pulled from the Lucasfilm archive, including a few seen in the infamous cantina scene in the original Star Wars.
This flexible, all-inclusive approach to filmmaking was developed by Lucas while making Episode I, refined on Episode II and perfected on Episode III. His mantra was to use whatever was most practical and economic for the scene, whether it be real-world, live-action monster masks, practical effects, well-understood simple techniques or the latest in computer-generated imagery developed by ILM. This one chase sequence involved all of those techniques and brought all of Lucas’s skills as a filmmaker to bear on telling his story. Other locations – the water world of Kamino (with its ‘Grey’ alien-style inhabitants, where the clone troopers that prefigure the stormtroopers of Episode IV are created) and the desert planet of Geonosis (populated by insect-like creatures) among them – were realized this way. Since the early 2000s, when these films were in production, this approach to filmmaking has become the Hollywood standard, especially for big-budget, high-concept blockbusters.
As director, Lucas did much to keep the pace of filming brisk, hitting an average of thirty-six set-ups (individual shots) per day, an extraordinarily high amount for a film production, much more akin to that achieved in television. By the end of August the Star Wars crew had wrapped in Sydney and were on the move, taking in Lake Como in Italy (where Padmé’s retreat at the Naboo lake was shot), the Plaza de España in Seville, Spain (where Anakin and Padmé touch down on Naboo) and a week in Tunisia where they returned to many iconic locations from Episode I and the original Star Wars.
The location shooting in Spain was completed in a just a few hours during a single day, even though it had taken ten weeks to set up. Watched by thousands of locals who had come to see the new Star Wars film in production, Portman and Christensen played out a relatively simple scene against a superb practical location (later augmented by digital additions). This was another example of Lucas’s filmmaking approach in action – a visit to Spain, however brief, gave the scene a sense of reality that all-digital environments lacked.
While Spain provided the Naboo exteriors, Italy’s Palace of Caserta once more supplied the interiors of the Theed Palace. Following a day off after shooting in Spain, the cast and crew reassembled near Naples for another single day of shooting at the eighteenth-century palace. Then the Star Wars caravan moved on to Lake Como to film Amidala’s home at Naboo’s Lake Country. A popular retreat for aristocrats, the wealthy and movie stars, Lake Como has been occupied since Roman times and is the location of several extravagant villas. The Villa del Balbianello on the western shore was used as Amidala’s retreat, the location of her clandestine wedding to Anakin that climaxes the film. Built in 1787 on the site of a Franciscan monastery, the building had been the final home of explorer Guido Monzino and now contained a museum dedicated to his work. It would later also be used as a location for the James Bond reboot movie, Casino Royale (2006). Lucas had visited there the previous year on his European holiday, so had written scenes with the location in mind. ILM altered the exterior of the building digitally, as usual, to make it fit better with the established architectural style of Naboo. As was seemingly traditional on Star Wars movies, location shooting was affected by unexpected, extreme weather – this time torrential rain around Lake Como caused a standby interior scene to be added to the schedule while the crew waited for the weather to clear up.
September 2000 saw the arrival of the Star Wars crew in Tunisia, shooting for a day and a half on the locations for Mos Espa, which remained more or less intact from their use on Episode I. A touch up of the paint and a quick refurbishment of the street and the scene was ready to be shot. That was followed by a more extensive trip to the iconic Chott el-Djerid location used for Luke Skywalker’s home in the original Star Wars. As Episode II saw Anakin return to his home, these sets had to be recreated as nothing remained from the original shoot over twenty years before. Archived technical drawings (not always accurate compared with what had actually been built on location), photographs and even frame grabs from A New Hope were used to recreate the Skywalker homestead. It was important to Lucas – and to the fans of Star Wars – to get this right. ‘It was wonderful trying to recreate that set,’ said production designer Gavin Bocquet, well aware of its historic importance to many film fans. ‘The homestead was at the very heart of Star Wars.’ Only two people working on the current film – Lucas and Anthony Daniels – had been there in 1976 for the original, and the making of that film had not been as extensively chronicled as the later Star Wars movies. Their memories, alongside what material could be retrieved from the Lucasfilm archives, proved to be enough, along with locations that remained remarkably untouched, to recreate the Tatooine familiar to filmgoers and fans. Lucas found filming on these locations to be ‘a very nostalgic experience. It was odd to be back in one of the places we shot the original movie, especially since it hadn’t changed much. Last time I was there [in 1976] I was under a lot of pressure. I didn’t have any idea of where my life was going or where this movie was going. This time, it was a much more mellow experience.’
The scheduled pick-up studio filming followed in September 2000 at Elstree Studios in Borehamwood, the traditional home of Star Wars. Although much diminished from its 1970s heyday, the studio still boasted the specially built George Lucas Stage and was regarded by the director as a lucky charm when it came to the Star Wars movies (he and Steven Spielberg had also filmed parts of the Indiana Jones trilogy there).
Principal photography on Episode II wrapped in the UK on 20 September 2000. The only actor present was Ewan McGregor, who shot the opening to the speeder chase. This final shot of the production saw McCallum wield the clapper-board once more, just as he had done for the first shot back in Australia. Thousands of miles of travel, journeys to five countries, sixty-one days of shooting and a year of pre-production had led to this moment. The task of making Episode II was not yet complete, though. The film would not be released to cinemas until May 2002, and an intense eighteen months of digital post-production remained.
The focus passed to the team at Industrial Light and Magic, who would take the raw footage Lucas had shot and work their digital magic on it. Digital special effects would be present in just about every aspect of the new Star Wars movie, sometimes obviously in huge space battles, but often invisibly as real-life sets were extended or little fixes made to troublesome live-action footage.
George Lucas was still at heart an editor. From his earliest days, the part of the process of filmmaking he had enjoyed the most was editing, in stark contrast to writing, the process he enjoyed the least. Lucas viewed shooting the film as ‘gathering material’ that was then endlessly malleable thanks to his development of new filmmaking technology. ‘I like a film to be organic,’ he said. ‘I like to change it.’
Ben Burtt had initially assembled a very rough cut as production progressed, refining it as filming went on and more raw footage became available. Back at Skywalker Ranch after the shoot, Burtt and Lucas spent several months fine-tuning their first cut. That initial rough cut that ILM would start work on was completed in February 2001 and came in at two-and-a-half hours. This was a better outcome than the first cut of Episode I, which was almost three hours long, but still too long for a Star Wars movie (Lucas generally aimed for somewhere around the two-hour mark).
The first cut revealed the need for further additional shooting to clarify story points or improve on the visuals. ‘My reaction to those first viewings is always disappointment,’ admitted McCallum after he and Lucas first viewed the film in one sitting, rather than in bits and pieces as they had previously. ‘All I see are things that are wrong, the lost opportunities: why didn’t we do this or that?’ Lucas agreed with McCallum’s feeling, but the new way they were making films allowed for second (and third) chances to get things right. ‘I began looking for anything that was unclear or needed to be amplified,’ said Lucas. ‘At that point, I wrote new scenes to fill in those spots.’
During the final week of March and the first week of April 2001 the Star Wars crew were shooting once again, this time at the venerable Ealing Studios in London (where Alec Guinness had shot such classics as The Man in the White Suit and Kind Hearts and Coronets in the late 1940s and early 1950s). This was the first of a total of three post-production pick-up filming sessions. Key actors were reunited, including McGregor, Christensen, Samuel L. Jackson, Christopher Lee, Natalie Portman and Anthony Daniels. Much of the new footage was shot against blue screen as it would be digitally integrated with the existing movie. ‘Blue used to be my favourite colour,’ said Daniels, ‘but I think I’ve overdosed on it.’
Among the scenes shot at Ealing was an entirely new action sequence that sees C-3PO, R2-D2, Anakin and Padmé journey through a deadly droid factory on Geonosis. Lucas created the sequence to replace some rather static and talky scenes with one of action, replacing chat with excitement. There was another reason for the scene: ‘The droid factory scene also got Artoo and Threepio out of the ship and more engaged in the final act of the movie,’ admitted Lucas, who had always regarded the two droids as key to all six Star Wars movies.
This retooling of the film was an ongoing process for Lucas and his team from September 2000’s wrap of principal photography through to the end of 2001. ‘At that point ILM took over,’ said Lucas. ‘They created backgrounds and the final animation, but even then we were re-cutting and reassessing it all. I couldn’t do a final cut on the whole of the last reel until I had all the material gathered, which wasn’t until about March 2002. I couldn’t really tell if something was working until I had the final visual effects shots.’
One of the major concerns on Episode II was the commitment to an all-digital Yoda, after the use of an unsatisfactory traditional puppet version on Episode I. Lucas had experimented with a CG Yoda for a final scene in The Phantom Menace that briefly saw the character walk, but was unconvinced the technology was well enough developed to pull off an all-digital Yoda, preferring instead to concentrate his efforts on Jar Jar Binks. With Yoda, noted Lucas, ‘We weren’t in a place where we could do it in close-up and make him look absolutely real. ILM probably could have done it, but I didn’t want to lay that on them at that time.’
Now, on Episode II, the time was right and ILM were championing the idea of a fully CG Yoda. During the writing of the screenplay, Lucas had several conversations with animation director Rob Coleman and visual effects supervisor John Knoll. For Coleman, one of the attractions of working on Episode II was the chance to develop a digital version of such an important Star Wars character. To convince his director that it was possible and would be convincing, Coleman and his animation team carried out an experiment. They replaced the puppet Yoda in a few key scenes in The Empire Strikes Back with a digital Yoda and showed them to Lucas. Finally convinced that it would work, the writer-director took a new approach to writing for the character. ‘I said, “OK, then, I’m going to treat Yoda in the script as if he were digital.” It allowed me to write him in a very different way: he could walk around, he could fight with a lightsaber.’
That led to ILM’s work on one of the most challenging sequences in the entire movie: the climatic lightsaber duel between Yoda and his old Jedi Master, Count Dooku. In something of a casting coup, Lucas had secured Hammer Horror star Christopher Lee to play Dooku, the rogue Jedi turned Sith Lord and right-hand man to the evil Palpatine. His casting connected the new films with the first, in which Lee’s Hammer co-star Peter Cushing had played Grand Moff Tarkin, Vader’s henchman. There was among those involved in the production huge concern about how this action sequence could be achieved: in private many disparagingly called it the ‘fight between the frog and the old man’, and they were worried about their ability to pull it off convincingly.
Lucas had recognized the difficulties in achieving his vision with a less-than-agile actor (Lee was then approaching his eighties) and a non-existent digital character. ‘He’s one of the most important characters in the pantheon of Star Wars characters,’ said Coleman of Yoda, so he was determined to get him right. The other part of the equation was how to make Lee appear to be a convincing, agile swordsman. ‘I didn’t know how well it would work,’ admitted Lucas. ‘We set ourselves an impossible task and just hoped we could accomplish it.’
The aim was to show Yoda at the height of his powers as a capable, physical Jedi. Coleman, however, had long realized that there were other limitations to what could be done with Yoda: to stay true to the character as shown in the original Star Wars trilogy, the animators would have to recreate Frank Oz’s Yoda, complete with his idiosyncratic movements that were actually the result of being a hand-operated puppet. ‘It was difficult to visualize a fight between these two unlikely adversaries: an 80-year-old man who was about six foot five, and an 800-year-old creature who was three feet tall,’ said Coleman.
A mix of techniques would be brought to bear on achieving a solution – some high-tech, some very low-tech. Lee was well trained in swordcraft, long a requirement of most classically trained British actors of a certain age (along with horse-riding skills). However, although he was fairly fit, there were some shots where a younger stunt double was required. In postproduction, the double’s head was digitally replaced (not entirely successfully) with that of Lee. The CG Yoda was seen to leap around the spacecraft hangar in which the battle took place, making the height differential between the characters less of an issue. The scene was the biggest challenge for the animators at ILM, but was one they only achieved to 80 per cent of their satisfaction.
By January 2002 the film had gone through a total of around five major cuts and ILM had completed most of the effects work (with over 2,000 visual effects shots), allowing composer John Williams to work to an almost final version when recording his score with the London Symphony Orchestra at Abbey Road Studios. Over a two-week period, the score was recorded, with a mix of new material and iconic Star Wars themes. ‘The music is the glue that holds it altogether,’ said McCallum of the finished movie.
The final shot completed by ILM was delivered on 8 April 2002, allowing Lucas to complete his final cut of the movie two days later. A preview screening for cast and crew was held at Skywalker Ranch’s opulent Stag Theatre that same day, 10 April 2002. It was the first chance most involved had to see the finished film, complete with effects, music and sound effects. ‘The original Star Wars was a joke, technically,’ said Lucas looking back at the more primitive filmmaking methods available to him in the 1970s. ‘I had to cut corners and cheat and make it kinda fuzzy so you couldn’t see what was going on. Most of what we did on Star Wars had never been done before; it was all prototype stuff.’ Technology Lucas had developed using his own time and money over twenty years now made it easier than ever before for him to complete his story. It was down to rebel filmmaker Lucas to show the rest of Hollywood the way forward.
In terms of storytelling, the intention of Lucas to replay and echo elements of the earlier films in the new ones became clearer. ‘I’ve created scenes that were reminiscent of those in the first trilogy. Situations are the same, with slightly different circumstances. I compare it to a musical motif, where the same themes keep recurring.’ This resulting middle instalment in the second Star Wars trilogy deepened and complicated the universe, showing the steps that would lead Anakin Skywalker to ultimate darkness in the third film, another production that Lucas knew would hold a host of brand new challenges – in storytelling, technology, filmmaking and in meeting audience and fan expectations.
Star Wars: Episode II – Attack of the Clones was the first ever theatrically released film to have been shot entirely digitally. Although Lucas and McCallum had been campaigning for cinemas to convert to digital projection in time for the release on 16 May 2002, only around 120 theatres in the United States were equipped to screen the film digitally (out of the 3,000 screens playing the movie). For the rest, traditional 35 mm film prints had to be supplied.
Despite the fan anticipation, Episode II failed to emulate the box-office success of its predecessor. By no means a financial disappointment, the film took over $310 million in the United States and a further $339 million across the world. However, it was the first Star Wars film not to be the biggest grossing film of its year of release. A trio of franchise movies out-grossed Attack of the Clones in the United States: Spider-Man, Harry Potter and the Chamber of Secrets and The Lord of the Rings:The Two Towers. Episode II would be the lowest performing of the six live action Star Wars feature films at the US box office (when figures are adjusted for inflation).
Critically, the film had a tough time, too. Hayden Christensen was widely criticized for his portrayal of Anakin Skywalker, although he would later win praise for other roles (such as journalist Stephen Glass in Shattered Glass, 2003). The problem may have been with the role he was being asked to portray – a whiny teenage Jedi, rather than the proto-Darth Vader many fans wanted to see. Roger Ebert identified the problem as being with the script rather than the film’s impressive realization. ‘As someone who admired the freshness and energy of the earlier films’, he wrote in the Chicago Sun-Times, ‘I was amazed at the end of Episode II to realize that I had not heard one quotable line of dialogue.’ He also noted that the romantic relationship between Skywalker and Amidala had fallen flat on screen, whatever Lucas’s stated intentions about recreating an ‘old-fashioned’ 1930s movie romance. ‘There is not a romantic word they exchange that has not long been reduced to cliché,’ wrote Ebert.
As with most of the previous Star Wars movies, Attack of the Clones won many technical awards including an Oscar nomination for Best Visual Effects. The controversial clash between ‘the frog and the old man’ won the film a Best Fight prize at the MTV Movie Awards. However, it was in the Golden Raspberry Awards (dished out for the worst films of the year) that Attack of the Clones triumphed, with seven nominations and two wins, one for Lucas for Worst Screenplay and another for Christensen for Worst Supporting Actor.
George Lucas had made life difficult for himself in choosing to tell the story of the middle Star Wars prequel in the way he did, but he felt he had no choice as he was following a thirty-year-old template. Attack of the Clones was no The Empire Strikes Back, but it did offer many images and moments Star Wars fans had never seen before. It was also hugely successful in one vital aspect: it set the stage for the climatic events of Episode III.