For the illiterate amongst you, the Lord of the Rings is a trilogy of books written by John Ronald Real Tolkein. The first installment in the trilogy was called “The Fellowship of the Ring” which was first published in 1953. This was closely followed by “The Two Towers” in 1954 and finally “The Return of the King” in 1955. The trilogy is without a doubt one of the finest works ever to grace the shelves and few novels have come close to its brilliance. It has been heralded as the book of the 20th century by a Time magazine poll and is known as “The greatest story ever told”.

As The London Sunday Times aptly put it, “The English world is divided into two parts. Those who have read the Lord of the Rings and those who haven’t”. Despite the series being over 50 years old, the books still hold large appeal to readers and critics still continue to marvel at Tolkein’s ingenuity. However, despite these works influencing fantasy movies for the past three decades and all fantasy authors owing a debt to Tolkein’s vision, up until now there had never been a film remake of the trilogy.

There was an animated feature film made in the early eighties but it was poorly made and it did not justify the Lord of the Rings at all. I found this very disconcerting and somewhat puzzling as Hollywood is never far behind with a film version of any remotely popular novel. I asked myself, why has nobody realized the potential of releasing a film based on the trilogy? I finally came to the conclusion that, up until now, we did not have the technology in order to depict an entire new world with magical creatures onto film. While films such as Star Wars and Jurassic Park were able to achieve this, it had never been done on as large a scale as the Lord of the Rings before.

To effectively reproduce the trilogy, would require thousands of fictitious characters to be put together on screen at the same time, in order to create the epic battles the Lord of the Rings is famous for. To create a fantastical world from scratch is a daunting task for anybody. Few are brave enough to take up the challenge but a certain Peter Jackson was. The Beginning When Peter Jackson and Miramax Studios announced that they had secured rights to produce movie adaptations of Tolkein’s first saga in the trilogy back in 1998, everyone was taken aback. Jackson, a New Zealand producer, was attempting to do what nobody had done before. In truth, he had his critics because his previous films, including works such “Bad Taste” and “Brain dead” had clearly pushed the boundaries of, well let’s face it, bad taste.

Jackson, however, was unhappy as he planned to film two movies instead of Miramax’s intended one. He eventually dropped out and signed a three-movie deal with New-Line Cinema and began filming. Jackson was determined to film the entire movie in New Zealand in order to maintain better control and to minimize costs. This meant that a massive ramp up of staff and technology. His studio, Weta, began recruiting from across the globe and began to develop a number of propriety tools. The stage was set and it was obvious to Jackson that more digital artists were required.

By mid ’97 they had 29 strong but they expected that they might grow to as many as 80 digital artists during the project. In fact, at the peak of filming on the first film, Weta had 164 digital artists. Similarly, the caliber and amount of processors used for modeling and rendering quickly expanded, from 32 to 400 for the “Fellowship of the Ring” and a further upgrade for “The Two Towers”. Despite the gargantuan challenge placed before Jackson and his crew, they astounded the world by completing the principal filming on all three films in just 15 months before December 2000 and then doing some re-shoots for the second film in 2002. During that time, more than 350 sets were built and in excess of four million feet of film shot. Weta Digital produced more than 570 digital effects for the film but the enormous workload led to Peter Jackson calling on American studio Digital Domain to provide some assistance.

Weta claims to have created in incredible 1600 Computer Generated shots for the entire trilogy. What makes it even more astonishing is that this does not include the massive amount of live action effects and props handled by Weta Workshop division. Not bad for a previously unknown studio. ‘ When the first film “The Fellowship of the Ring” was released in December 2001, it was an instant hit, making $47 million in its opening weekend. The Fellowship of the Ring was nominated for 13 Academy Awards, including Best Director, and won four. The film also received the American Film Institute’s prestigious Film Award and was nominated for 12 awards from the British Academy of Film and Television Arts (BAFTA), winning awards for Best Film and garnering Jackson the David Lean Award for direction.

In addition to four Golden Globe nominations, the film also received numerous distinctions and awards around the world. How it was done When you sat in the cinema and marveled at the wonderful special effects which managed to entertain you for three solid hours, did you ever actually wonder how it was done. Cave Trolls, elves, dwarves and hobbits don’t simply render themselves (although it would be nice if they did); no, they were made through the painstaking efforts of the digital artists at Weta Digital. How did they do it? They used a wide variety of techniques, programs and effects to create what we see on screen.

Easily one the most impressive tools utilized by Weta was their very own proprietary software, Massive. Various methods of digital crowd replication have been produced at visual effects facilities around the globe in recent years, but few have come close to the power and level of sophistication that Massive has. Developed as a battle simulation system specifically for the Lord of the Rings trilogy, the software is capable of churning out thousands of orcs across the screen and it was at the heart of the tumultuous battle sequences in the Two Towers. It was designed by Stephen Regelous who was asked by Jackson to put together a crowd system for what was then to be, the Hobbit. He originally turned down the idea, thinking that it did not have much scope for anything exciting.

He then reconsidered and decided to give the 3d characters their own AI (artificial intelligence) instead of just attaching animation cycles to each one. Massive is flexible enough to be used for more than just the crowd scenes. A production version of this program debuted earlier this year. The current price tag for a floating license is set at $40 000 which could be money well spent. Massive enabled Weta to create any number of computer generated autonomous agents (agent is the term for a Massive warrior), each made up of a body and a brain. With this program, it was possible to render, thousands of agents on screen at once, each with their own “personality”.

But how exactly were these autonomous agents made? Building an army To get the actions that the agents needed to perform, Weta relied heavily on its very own motion capture or “mo-cap” which was based in Wellington. The Lord of the Rings mo-cap stage is one of the world’s largest. The motion capture actors were filmed by 24 cameras simultaneously on a closed stage and had their movements motion captured and then put into massive. These actors wore suits covered by reflector balls, which were then tracked digitally by the cameras. This captures a wide range of movements that can then be used for the Massive program.

The actors are then recreated digitally. What the stunt actors look like is not important because only their wire frame was captured. This real-time transposing of human mo-cap data to the non-human skeletons of the various creatures helped drive the armies created on the Massive system. The brain of the agents is basically a hierarchical AI network that breaks up instructions, very basically, along the lines of, if an enemy is within 10 meters of you, run at it while raising weapon. Assemble enough of these behavioral patterns, have each instruction trigger a small animation loop – typically generated by motion capture and then blend the animation together to form a CG character. Unfortunately, this has to be done thousands of times for crowded scenes.

Massive made 80000 agents for prologue scene in the first film. Characteristic actions were also designed and recorded for orc’s, elves and men. In this way the digital characters balance, breathe and generally behave in their own distinct ways. For battle scenes, Weta Digital had to visualize the action of every character so they used motion tree designers (a detailed map of every agent’s actions) and stunt men and battle things out, to see if they could possibly predict what would happen in a fight. The actors hit each other with swords or any particular weapon that they might be using and each of these has to be taken into account. Each agent will then draw from a set of more than 300 motion capture cycles.

So while you are looking at the screen, you ” re not thinking to yourself, all the characters look alike. Each one of these ten thousand characters has their own “personality”. All of the agents have the ability to loco mote, that is, they are able to walk, run, charge or flee and fight very convincingly. There are 30 different Master Agents, each with a specific race and fighting character i.e. Elven swordsman, Orc Pikes man, Uruk Berserker etc. Every base agent will be able to come into a battle and be able to defend itself successfully or die. Every agent’s “Brain” continually chooses from all the action cycles available using the motion tree and can therefore adapt to its circumstances.

Each agent is capable of making up to 24 decisions per second and it will then perform one of the thousands of actions recorded by the motion capture team. These actions must be carefully rehearsed to make sure that they will all blend in to each other when the characters are put together. How to make a motion tree Weta broke up the film, scene by scene and decided what Massive agents will be required. The motion capture actors then perform the actions required using the weapons that they believed were necessary.

A certain amount of pre-visualization is necessary in order to perform the right actions. The mo-cap team included many stunt, fight and bladed-weapons experts. Once all the necessary actions have been captured digitally and a capabilities list has been compiled, then the actions are sorted with a program called “Tree-planner” which divides the actions of a character up into specific “transition points”. For example, an elf might be in a static position and then suddenly it lunges forward. This is then a transition point and the movement from one transition point to another is the action that took place. A typical agent has 10 to 30 cycles for walking, running, climbing, falling and fighting as well as hundreds of cycles for smaller “blending” movements.

Each move will be motion-captured and then be blended together by an agent’s brain. The agent is then able to predict which move to make, depending on its circumstances. Weta also used motion capture to develop different fighting styles for elves, orcs and humans. These actions are then assimilated into one big “motion tree”. Massive was used in conjunction with Grunt, Weta’s own in house rendering solution. Using a standard professional renderer would have taken up to 30 hours to make a single frame of animation if a battle scene with 10000 warriors needed to be created.

Grunt was specially tailored for crowd scenes and so it manages to do this in half an hour. For non-battle CG, the studio made use of Renderman, which was developed by Pixar and used for its own movies. Renderman has been seen by many as the world’s best rendering system to date. Also in Weta’s 3 D pipeline was Alias / Wavefront’s Maya, which was used for character modeling, rigging, and animation.

Maya’s built in renderer isn’t highly regarded but its wide range of features and open-ended nature have made it the tool of choice for movies such as the Matrix, the Grinch, Final Fantasy and Episode One: The Phantom Menace. Weta also used Shake and Flame The Beasts of Middle Earth While the orcs and trolls in the films were often simply human actors in prosthetic suits, it was also necessary to utilize CG to create purely digital creations. Our first glimpse of Weta’s Digital CG creature capabilities was in the Mines of Moria sequence when Frodo and the rest of the Fellowship came face to face with a hideous cave troll. The troll was built using NURBS a technology that draws models using curved surfaces, instead of plain polygons although Weta returned to polygons for the Two Towers.

However, this time round, they used a technique known as subdivision surfaces which adds more detail to the basic triangle model. Further detail was added to the skin by using model deformation and bump mapping to add ridges of scales. Under this skin layer was a muscle and bone structure although the skeletal frame used looked very different from the average man on the street. By spending time rigging up the inner workings of the characters, Weta could animate by using only the skeleton. They would then take into account the effect of the skeleton on the muscles and in turn, the muscles on the skin. The end product would then appear lifelike in its movements.

One of the main challenges that the workshop faced was that of incorporating the creatures effectively into the scene. On set survey data provided crucial information as to how the creature needs to be coloured and lit by providing an instant reflection map of the environment. Standard procedure is then to keep camera movements to a minimum. With the cave troll scene, however, hand-held camera work (done by Jackson himself) shows the battle between the heroes and the troll up close and personal. This required the use of programs such as 3 D Equalizer and Shake that were designed to pinpoint elements in a live scene and to track their movements so as to provide virtual 3 D camera views of the troll. Another infamous creature from the first film was the Balrog.

When it was decided that the creature’s demonic appearance would be shrouded and a cape of black smoke and a ten-foot high mane of fire, some precautions were necessary as digital flames were notorious for looking artificial. Weta’s solution was to create the flame from 50000 2 D sprites, each one being a shot of real fire layered onto the Balrog. And let’s not forget Gollum, who won an MTV movie award for his performance in the Two Towers. Motion capture was also used for Gollum and for doubles of the film’s main characters. In fact, there’s talk that Gollum might set a whole standard for synthetic thespians.

He’s and incredibly complex creation and many have noticed that he’s entirely CG and he is a character with a purpose and direction unlike his predecessor, Jar-Jar Binks. The Return of the King I suppose that we will all just have to wait and see what happens in the final installment of the trilogy but the final film has got a lot to live up to as it’s older siblings were truly amazing, not just in the effects side of things but also in the telling of the story. Jackson really pulled all the stops out in his retelling of the greatest story ever told and I commend him for his efforts, well done Peter Jackson!

<