A Journey Through' Kaze, Ghost Warrior' With Timothy Albee

If you have the Kaze, Ghost Warrior.

You have heard it before. Stories of: “If they won’t let me do a movie, I’ll go do it myself!” are part of the Hollywood legend. It is well known that if you have talent, “they” will give you a chance — or will they? For some people, waiting for their lucky day is not enough: they need to make their own chances and force a breakthrough on their own. Sam Raimi’s homemade Evil Dead (1981), Peter Jackson and his four years-in-the-making Bad Taste (1987) or Robert Rodriguez’s can-you-believe-he-shot-it-for-$7,000 El Mariachi (1992) are here to remind every wannabe filmmaker that it is possible.

Ask Timothy Albee, a former animator who last year embarked on an amazing journey: creating a stylish CG-animated short film to raise funds for a feature film project. And the result, Kaze, Ghost Warrior, is, by any standard, impressive. Albee resisted producing a mere technical showcase à la Final Fantasy and created instead a unique world populated by intriguing characters. Kaze is much more than a “look-what-I-can-do” film. It is a real piece of art in which substance and style — for once — transcend technology. Most impressive of all, Albee created Kaze alone, on two PC workstations, working in a cabin in the wilderness of Alaska…VFXWorld recently spoke to Albee about his Alaskan journey to make Kaze, Ghost Warrior.

Alain Bielik: How did this all start?

Timothy Albee: I’ve always been an artist. I’m a painter and a musician (I even trained in opera), but what I really wanted to do was cel animation. I quickly delved into CG animation and worked on video games, commercials, TV series (Babylon V) before joining Walt Disney Feature Animation, credited as an assistant animator on Dinosaur (2000). I then ran my own animation company called Exile Films for a while. It all changed the day I saw Fight Club (1999). There was a scene in which Brad Pitt’s character drives very dangerously and asks his passengers: “If you were to die now, what do you wish you had done in your life?” An answer popped up into my mind instantly: go live in a cabin in Alaska… And I did it.

AB: Really!

TA: I quit everything and on May 1, 2001, I settled down in a cabin lost in the wilderness. I was planning on a sabbatical leave of one or two years. To tell you the truth, I’ve always had a passion for mushing and settling down in Alaska allowed me to live this passion in a very intense way. This said, the conditions are pretty extreme out there. I almost died on four occasions…

AB: How did you sustain yourself during this sabbatical?

TA: I traded paintings against food and supplies. I also wrote books on CGI, filmmaking and animation, one of which now replaces the manual for LightWave 3D. LightWave is my favorite animation package, along with Softimage. I’ve had the opportunity to work with Maya and it is a very powerful tool. However, when I delved into producing CG animation, I looked for a package that was an extension of my own creativity. LightWave supports this. With Maya, you first need to write a bunch of scripts in order to be able get it to do what you see on its demo reel — basically re-programming the thing. It is billed as “Teamware,” and you end up with as many programmers as animators. It didn’t work for me. Plus, LightWave comes with the fastest renderer, that looks beautiful, right out of the box, while it’s assumed that with Maya, you need to acquire RenderMan or mental ray for a decent looking render, which requires even more programming as shaders must be written. To me, Maya is like buying a very powerful car in which you still have to build the transmission. With LightWave, one person can do it all. And that’s exactly what I had to do.

AB: How did you get this idea?

TA: The story of Kaze has been with me since adolescence. When I was younger, whenever I read a book or watched a movie, I could always see ways it could have been made even better. French director Jean-Luc Godard once said: “The best way to criticize a film is to make another film.” So, I shopped my Kaze story to all the studios. That was even before I went to work for Disney. They all thought the idea and characters were fantastic, but they didn’t believe it could be made for the $10 million I was asking for. They actually told me: “Add $20 million to your budget and we’ll do it!” But making it for only $10 million was the whole point.

AB: Why was that?

TA: When you work on a big budget production, it becomes committee work. The executives, whose job it is to be adept in the ways of money, begin to think they know better than the people they’ve hired to be creative, and they put pressure on the director to make the movie they think will sell. If the director wants to keep his job, he must capitulate. In the end, the movie is marketing, not storytelling. With a $10 million budget, the studio leaves you alone — hopefully! You get total control and your integrity is preserved. Plus, with even moderate advertising, the studio can make its investment back on the opening weekend alone!

AB: Were you actually certain you could make an entire CG-animated feature film for $10 million?

TA: Yes, although nobody around me believed it. I had friends who could only dream about their feature film project because they figured they needed 300 people to do it. I needed to prove that it was possible to do, and I adopted a new strategy. Instead of trying to sell my concept for a Kaze feature, I set out to direct a short that would be a pilot for the project. It was to be done for as low a cost as I could manage. Plus, in order to make a point, I elected to work in the most extreme conditions: I would do it alone, in six months, for $5,000, and the film would include many technical challenges such as character animation, lip-synch, animal fur, a deep forest and “realistic” characters.

AB: Why did you want to do it in six months especially?

TA: I wanted to prove that a Kaze feature could be entirely done in 18 months: if I could direct a 23-minute demo in six months on my own, a team could definitely do a 90-minute movie in 18 months… That was the point.

AB: How did you manage to create 23 minutes of CG character animation in six months?

TA: I knew I couldn’t spend hours fine-tuning facial animation or body language. The idea was to work on a shot up to the point where I felt the animation was carrying the performance across, no longer. Some will consider that this is “rough” animation, but I had 2,000 shots to animate in six months. That’s 28,000 frames! The only way I could pull this off was by devoting four hours, no more, to any single shot. The aim was not to create perfect shot after perfect shot. It was to produce a film that worked well enough as a whole to allow audiences to fall into the world, believe in the characters. It was to convince investors that the Kaze feature film was doable for the amount of money and in the amount of time that I was asking for.

AB: Was your knowledge of LightWave instrumental in pulling this off?

TA: It was definitely part of it. I pulled together a lot of seemingly unrelated technologies to accelerate the whole process. For example, the techniques and riggings I designed allowed me to deliver between 30 and 40 seconds of animation per week, from start to finish. When I was at Disney, we were producing a mere seven seconds of rough animation per animator per week! Also, I worked with a programmer to write a brand new lip-synch program from my ideas. The tool allowed me to do dialogue and facial performance much faster than ever before. I had an enormous amount of lip-synch animation to do on Kaze and I couldn’t afford to spend four days on a shot as we sometimes did on Dinosaur. A main idea was to incorporate traditional puppetry techniques in the animation process by turning the mouse into a motion capture tool. It was much faster to do lip-synch with this new technique than with the traditional sliders. I could animate in 20 minutes a shot that would have required two solid days to do the traditional CGI way. This lip-synch tool is a stand-alone program that will eventually support all major 3D software packages. It is slated to come out at Christmas.

AB: How did you handle the render time? You have fur, a bamboo forest… This typically takes a lot of time to render.

TA: I modified the traditional concepts of rendering to avoid spending hours on complex shots. For example, I strung seemingly unrelated tools together that, when used in this fashion, allowed the fur to be pre-baked. It allowed the color to be pulled from the fur as a texture. The render time went from 18 minutes per frame to 58 seconds!

AB: What was your gear when creating Kaze?

TA: I worked and rendered on a Dual Intel Xeon 2GHz with 2GB RAM and 200GB HD space. The third processor on the “render farm” was an AMD 1GHz Athlon with 720MB RAM. Both machines had a NVIDIA 4200Ti card. Modeling, animation and rendering were completed in LightWave 3D, while compositing was handled in Digital Fusion, another great tool. Incremental backups were done nightly on DVD+RW; full backups were done weekly.

AB: Tell us about the visual style of the film. Some shots of the landscape are visually arresting. Where did you get your inspiration?

TA: I drew my inspiration from various wilderness painters, and the figurative work of Jean-Leon Gerome (1824-1904). He worked in rich painterly strokes and always knew how to bring your eyes to where it needed to be. This is an approach that I used a lot on Kaze. By drawing the viewer’s attention to a specific area of the frame, less time and effort were required on the supporting areas of the image. Part of my tests for Kaze was to try reproducing realistic animals and landscapes in CG. I set out to create 3D wildlife paintings that would touch viewers with the same strength as traditional oils. I even tested my work by exhibiting canvas prints of these works in a local gallery. Based on the response, the experiment was a success!

AB: Now that Kaze is out, what has the response been?

TA: I’m pleased to say that I’ve signed a deal with a company to develop Kaze into several mediums. There will be two feature films and a TV series in between. The films are budgeted at around $10 million. They will be produced in Germany — I’m actually going to settle down there — and post-produced in Poland. We also have a Kaze video game in the works. The other project that I’m working on is the furthering of development on the facial animation package to work on the entire character. This is a technology that will make animation much easier for both professionals and “newbies.” “TA Facial Animation’s” features include 61 frames per second playback (on my 4200Ti), the ability to modify everything — even to swap head on a model — while the animation is still playing, and the whole program currently weighs in at only 420 KB. We built it from scratch because I wanted to have a complete control of the main code of our animation package. We will still rely on other software for modeling and render. Expect the first movie to come out in 2007.

AB: Finally, why did you write a book on the making of Kaze [ CGI Filmmaking: The Creation of Ghost Warrior from Wordware Publishing]?

TA: I knew there were many people out there who were in the same situation that I was in, but who didn’t know how to put their dreams into motion, or even more, who didn’t think their dream could ever become a reality. I wrote a book detailing my experience, every step of the way, in order to show others that they can do the same if they choose. It is not a mere technical manual for CG animation. The book explains how to do a whole movie, from concept to planning to dealing with studios and lawyers. The book also deals with the ups and downs of the creation process, the self-doubts that are so common to passionate, creative people. The book also gives tips on how to look for a distributor, which turned out to be the hardest thing. Surprisingly, in all this process, making the film was actually the easy part!

To learn more about this short, visit Kaze, Ghost Warrior.

Alain Bielik is the founder and special effects editor of renowned effects magazine S.F.X, published in France since 1991. He also contributes to various French publications and occasionally to Cinefex.