Lord of the CGI

Published:  January 27, 2010
Lord of the CGI

Jake Lee has been in the VFX industry for 12 years, 11 of which have been spent in the good company of Weta Digital – probably best known for its work in The Lord of the Rings and King Kong.


Please tell me a bit about your education and background.
Hello, my name is Jake Lee and I’ve been in the VFX industry for approximately twelve years, almost eleven of those at Weta Digital. My education and background was a little odd. It was a mixture of photography and computer science. When I was about seven years old I was given an Agfamatic 2000 pocket camera and this was the beginning of a lifetime passion for photography. My university education was a mix of many subjects including Sociology, Psychology, Greek History, Astronomy, French Culture, Maths and Computer Science.

Fifteen years ago in New Zealand there weren’t any educational facilities that taught specific courses in VFX and CGI. Thankfully this is not the case today. I decided against art school or film school as neither had the specific courses I needed or wanted. I looked into programmes in the US and Canada, but didn’t have a spare $30,000 under the pillow that it would have taken. At the time the Vancouver Film School VFX course was ~$19,000 Canadian.

So plan B. I approached the bank for a loan to purchase hardware and software so I could set up a business and study CGI for Film. “No”, was the response, “You have no equity, and what a ridiculous idea for a business”. Ha. But with a little coaxing and my mother as guarantor (thanks Mum) the bank lent me the money I needed. So I embarked on a year of research into colour theory, model building, cinematography, lighting and all things vfx related. To learn more about animation I parked myself in the window of a Christchurch cafe for hours watching people outside walk, talk, argue, move, hug, sit, skate and occasionally fight. This taught me a lot about body language and movement and it was here that I developed an addiction for coffee, which later in my career became invaluable for sustaining consciousness many times at 3am. All this was funded by working from 6am-9am in a café, and 6pm-12pm in a restaurant five days a week. The hours in between were spent research, learning, watching, reading and creating VFX at a small studio a friend and I had established.


Can you tell me about your role at Weta Digital and how you came to work there?
I’ve had many roles at Weta but currently I’m supervising the new hires introduction to our tools, pipeline and approach to lighting. How did I get here? Some say hard work, I say luck.
I first heard that Weta was looking for Artists in 1997 through a casual conversation with my computer hardware supplier. He mentioned that they were looking for people with their own equipment, who were proficient with 3DStudio Max, which as it happens is the software I’d just spent eight months learning.

This chance conversation only occurred as a result of our studio being burgled, and I was in the process of buying new equipment with the insurance. He gave me the number of the people at 3Foot6, the LOTR production company. I rang asking about a position, they informed that they were looking for artists to do previz on the yet to be green lit Lord of the Rings. I wasn’t sure what previz was (it was a relatively new term for the industry at that time), but it sounded fun. They asked me to send a show reel, which I said I didn’t have. They said make one. I said I will.

I spent a week creating a two minute reel and a week rendering the frames. Finally I posted it and to my surprise a few days later I was on a plane to Wellington for an interview. I met with Rick Porras and Randy Cook who both interviewed me. I thought the interview went appallingly as I had no experience in any of the things we discussed. But at the conclusion of the interview, they asked when I could start. I said, “when do you need someone?”. They said how soon could I be here. Two weeks later I had packed up my life and business and was sleeping on a friend’s couch in Wellington ready for previz.

I do feel lucky, and am very grateful to Rick and taking a chance on me. I thought my show reel was pretty average at the time and now I can’t even look at it, it’s atrocious, but I keep it around as a reminder of how I started.

I began creating previz for LOTR, then after a year, moved into the Technical Breakdown Department at Weta Digital. That Department no longer exists but back then we needed a way to covert the previz imagery into real world data. This information was used on set to drive the motion control rigs, map out camera tracks and drive the filming of the VFX footage. I have spent many hours on Set, on the miniatures stage and on the Blue screen stages, basically anywhere VFX was filmed.

Following principal photography I moved into the Camera department, learning 3D camera tracking and where I began writing code and tools. This was an invaluable experience. About a year later I transferred into the lighting department focusing primarily on Environments, then into general lighting. I progressed from lighting TD (technical director), to sequence supervisor, 3D supervisor and finally I was a CG supervisor on a small show. It has taken about nine years to get to where I am today, but I have very much enjoyed the diverse experience along the way.


What drew you to the realm of CGI and VFX?
I remember the exact place and vaguely the time. I was traveling through Canada in autumn 1994 and had been in Vancouver for a month when I had an epiphany, although I forget what sparked it. I guess I must have become aware of how I could marry the hobbies I enjoy the most, Photography, Film and Computers into a career. I realised that with technology heading in the direction that it was, the possibilities of what could be created were limitless. This appealed to my overactive imagination. I booked flights back to New Zealand to pursue a career in VFX despite not having any idea about how to get started.


Who are some of your most admired VFX/ CGI artists and why?
The Artists I admire the most are a small group of very talented people at Weta Digital that I have worked with over the past 8 years. They have, and still do, produce stunning effects and work. Their tireless commitment, without complaint, to constantly push the boundaries of what is possible is incredibly inspiring.

How have you seen CGI/ VFX change over the years?
Yes CGI has changed over the years and I guess the most obvious change for me is to do with the speed and power of the technology. It’s faster, smaller and cheaper. This has allowed us to push what is possible and now we are able to create imagery at the “Imagination level”. With LOTR, it was a project that was made possible with the advancement in CG techniques; we couldn’t have done justice to Tolkien’s world without these changes.

One of the main changes is in the improvement in realism or perceived realism. The integration of CG and live action is almost seamless, it is difficult tell where the real stops and digital starts.


Why do you think it has become such a huge/ popular medium?
Primarily, I think that it is now very accessible to anyone. The hardware is getting cheaper and if one can’t afford the high-end software there are plenty of demo and educational versions. In addition there are many free and open source packages to support any budget. It is also a medium that people can produce interesting imagery with no prior knowledge or artist background. CG is a medium that marries the creative with the technical, science and art have never been more closely linked.

When you are first approached to add the CGI/ VFX to a movie what are the initial steps you take in addressing the brief?
Everything we do in VFX is geared towards producing imagery that realises the Director’s vision. Initially there are meetings with the director or the VFX supervisor to discuss the project. On most big projects and hopefully on smaller budget films concept art is produced. This indicates an overall tone of what the director is expecting to see. The art combined with the previz (previzualisation) becomes a blue print on how to proceed. Each sequence is then broken down into its individual components, ie; Characters, Background Sets, Effects and Environment to name a few. Each of these assets is then further dissected and distributed to the various departments. In the case of a character like Kong, the models department begins building the digital model, the textures department start creating the surface textures, and the Creatures department begins construction the skeleton, muscle system and animation puppet. This pipeline is similar for most assets and is an iterative process. It may last for months until we get final approval from the director.


What processes would you then follow in completing the brief?
From a lighting perspective my workflow for creating a rendered shot is to start with the concept art work and any notes from the VFX Supervisor or director. Then I investigate the surrounding shots to see if there has already been a lighting look or tone set. Once I have a rough idea of the palette I then explore similar shots that have finaled. This is to ensure continuity.

I begin with a basic Key, Fill, Rim three light rig. It’s important to match the direction, colour and intensity of the Key light from the live action. Then I balance the Fill and Rims, adding more detail where needed. Again, it’s an iterative process and when I feel it integrates with the live action I submit it for review.

What are some common problems you would encounter when working with VXF/ CGI and how would you overcome these?
A couple of major problems we always face is that of time constraints and the growing need for more complex CG. A way overcome these is to constantly push the limits of technology and Weta Digital is always looking to develop new tools and techniques. We have recently developed new lighting techniques,that not only decrease the time needed to light a shot but greatly improves render times.

What would be the main differences be between adding VXF and creating a whole scene/ background?
The main difference is the time it takes. The process is still the same if we add one CG character to live action plate or build the whole scene. We apply the same attention to detail to everything whether it is one hundred background buildings or one foreground character.

The lighting process differs a little different when adding a single CG elements or creating an entire scene. When adding CG to a live action plate our main objective is to match the lighting that was filmed and to integrate the CG seamlessly so there is no distinguishable difference. When creating a scene from scratch all aspects of the frame are created and we become responsible for the lighting direction. This does allow for a little more lighting freedom, but we still need to adhere to the brief and Directors vision.

The compositing process is simpler when creating an entire scene. The rendered images can be broken up into as many or few layers as required and recombined in any order. The need to Rotoscope a live action plate (to layer the CG between) and the keying of blue-screens is unnecessary.

What do you think is important to keep in mind when creating CGI/ VFX – especially if you don’t want them to ‘date’ quickly?
I think the biggest thing to remember is that VFX should be created to help tell a story, not be the story. IF CG pulls you out of the film experience and breaks the ‘suspension of disbelief’ then it has failed. Knowing which CG will date is hard to gauge. We have the technology and tools today to create almost anything we can imagine and as long as they are used within the context of the film then CG should not date.

Tron is a good example, created in 1982 very early in the development of CG. It may lack the complexity and detail of modern techniques but the CG still holds up within the context of the Film.

Filmmaking is about telling a story, evoking emotion, creating empathy, and if CG can aid in this process and remain true to the context and tone of the film then I don’t believe it will date.


How long would you normally be given to work on a film? How does this affect the process of your work?
This is a very difficult question to answer. All films have varying postproduction schedules and the amount of time we work on a project depends on whether we are contracted early, at concept time, or brought on when the VFX work is well under way. A project that we are the primary facility on may take eighteen months or if we are needed for only a small number of shots, we may complete our work in 4-6 months.

The main difference is the time allocated to R&D and development. With a limited schedule, there is less time devoted to all aspects of the pipeline but the process is still the same. The attention to detail is not sacrificed but the amount of time spent tweaking, changing and reworking the assets and images is less. This isn’t always a bad thing; new technology and creative solutions are often born out of time constraints.

What equipment/ software/ hardware do you use?
The desktop machines are HPs with 16GB of Ram, running a Weta build of Linux. The primary 3D software is Maya with custom plug-ins, scripts and tools. The primary render software is Pixar’s Photo Realistic RenderMan (PRMan). We have developed our own lighting and RIB (the format required by PRMan) generation tools.

How much time would you normally leave aside for rendering and what sort of equipment do you use to meet this massive amount of data input?
Once again this is difficult to quantify. Every shot I have worked on has required varying render times, some shots take an hour to render, some take a day. This is further complicated by the approval process. Every shot is first reviewed by the VFX supervisor then approved by the Director. There may be many iterations and late nights, before the Director’s vision is achieved, so any individual shot may take a few days, a few weeks or, in some very difficult cases, months.

Weta Digital has a massive amount of storage and processing power; here is a quick look at the machine room.

We have 3,800 render blades and currently have 450TB of high performance storage and are generating approximately 10TB of data a day at the moment.

This produces an enormous amount of heat so the Site is cooled by a highly efficient water-cooling systems using Wellington’s naturally nippy climate, This means that most of the time we don’t need to use any power to cool the site.

Where do you see visual effects and CGI heading in the future?
Ahhhh, to predict the future question. It’s crystal ball gazing really; the changes in the last five years have been massive. What the next ten years will bring is any one’s guess but one of the areas of CG that has been slow to catch up is in facial realism, particularly with facial motion. There is so much complexity within the human face and to simulate these nuances digitally has, in the past, been extremely difficult. But with modern techniques it won’t be long before we are creating realistic digital people that are indistinguishable from real people. But is this a good idea? I can see the benefits in creating digital stunt people and adding crowds, but replacing an actor? It’s the can we? should we? question.

Hardware and software will continue to get faster and cheaper, as will CG. This will in turn increase the number of productions adding VFX. My only hope is that CG will be a value added tool and implemented wisely.

Do you have any advice for budding VFX/ CGI artists out there?
My advice for people keen to explore VFX is gain a broad knowledge of CG but focus on an area you would like to specialise in. Its important to have a general understanding of the full process but most jobs in large companies will advertise for a specific role, including Modellers, Creature TDs, Animators, Shader Writers, Lighters, Effects artists and Compositors to name a few. For smaller facilities however you will need a broader skill base.

The things we look for when hiring new graduates, (and these apply to most departments) are; a good work ethic, be proactive, have an eye for detail and being meticulous. It’s important to have the ability to take direction and criticism, and to be philosophical about your work. Like most films, a lot of footage ends up on the cutting room floor.

A technical ability for code or scripting is very useful and python is a good place to start. Being able to use a 3D package like Maya, 3DSMax, Lightwave is a basic requirement and most companies create proprietary software so understanding the theory and process is more important.

It’s easy to teach an artist to use a computer or software but its very difficult to teach the computer how to use an artist.

Additional skills relating to a career in lighting would be; Acquire a good understanding of cinematography, the lighting tools are digital (and not as heavy) but the theory is identical. Gain some understanding of lighting and colour theory and watch a lot of film. Practise by taking objects, fruit, bottles, your sister for example and place lights around them and explore how it reacts with the surfaces. Then take a photograph and try to replicate in CG. Most importantly understand how it happens and why.

So in summing up the ideal VFX artist is someone who is partly technical, partly creative, is adaptable, proactive, patient and dedicated. Would I recommend this as a career? Definitely! But it requires dedication, long hours and hard work.

Is there anything you wish to add?
VFX doesn’t need to be expensive and with proper planning and a good understanding of the process you can create high quality images on a very low budget. The hardware and software available to the home user is relatively inexpensive and being able to produce memorable images just takes a little time. Good Luck and remember to have fun, sometimes we forget this is the entertainment business and the process of creating the imagery should also be enjoyable and entertaining.


Leave a Reply

Your email address will not be published. Required fields are marked *