| |
Interview with Brett Feeney
Digital Supervisor on Happy Feet, Brett Feeney talks about the production process of the animated feature, animating penguins, travelling to Antarctica and the future of animated films at Animal Logic.
December, 15th, 2006, by Raffael Dickreuter
 |  | Brett Feeney, Digital Supervisor
on "Happy Feet" at Animal Logic.
| | | |
| |
| | | |
|
After having done cg for many years, what aspect of it do you still enjoy the most?
The most enjoyable part is the problem solving for me. I really enjoy being met by large challenges and difficult projects. Breaking the problem down into digestible pieces that are more easily solved, and working with R&D to find the optimal solution.
You worked for the past few years at Animal Logic. Explain a bit the responsibilities you have had a the company
Most of the work I have done at Animal Logic over the years is in pipeline design and building. I have worked on many feature films here, where I am responsible for coming up with processes and toolsets to achieve complex visual effects and animation tasks. I have been the technical lead on a lot of the visual effects films during my time here, and most recently, the Digital Supervisor of Happy Feet.
To what extent did the Animal Logic pipeline suit an animated feature production, the company so far more known for its visual effects work?
Animal Logic was primarily an effects house at the beginning of our involvement with Happy Feet. We had to immediately change our way of thinking, as even the largest of VFX jobs we had done to date, before Happy Feet started, was no where near the scale of what we had to do to create a full animation studio capability. We had to create a lot of in-house tools to help keep track of the data and workflow, as well as come up with a language to describe all of this to each other. Where we had people skilled in different disciplines, working across VFX projects in the past, we had to break out into whole new departments. So really our problem solving methodology was one thing that was suited to the animation pipe, and the most change came from having to partition up into specialized groups/disciplines, as well as scale up the capabilities.
How many people all in all worked on the cg side on the film Happy Feet?
We started with a very small group of people actually laying out the proposed methodology, and then slowly started to fill the various departments out as we went. At the height of production we had somewhere near 240 people. There were probably more people than this over the life of the project, but some departments that finished earlier, and some started later. At a guess I would say that there were around 320+ people involved in the digital side of Happy Feet. It was broken out into about 15-20 departments, some of which had/were sub departments under 12 supervisors.
How early on were you involved in the production?
I was involved from the very beginning of the project. There was a very small team of us that did the initial planning and talking with the director about what he wanted the film to look like, how he wanted the performance to look/feel, and what level he wanted all the environments to sit at. So my initial involvement was sitting in a small focus group, with George, to hear how he wanted to make the movie, through to helping document the process of achieving this, and explaining back to George how we were going to go about it.
What references and inspirations were used to create the look and feel of the movie?
Some of the original look and story ideas came to George after watching the Richard Attenborough series “Life in The Freezer”. He was taken by the harshness of the environment that the Emperor Penguins endured, and thought that it showed incredible courage and strength of community for these birds to survive in this most inhospitable of places. So we were all required to watch at least key sections of this documentary. Combined with George meeting several Antarctic expeditioners and looking at the amazing photography they all had from their time there. He created a colour and look board from some of his favorite pictures.
Tell us about the production process of Happy Feet and your involvement in the production
My role in the production process was to design the pipeline and processes for making Happy Feet, from top to bottom. As Digital Supervisor I had to work with all departments to help shape the pipe and solve problems at any/every turn. I worked with an Associate Digital Supervisor and a team of 12 other supervisors for this side of things. I worked closely with the Directorial team to help facilitate the story process, and through the Art Direction teams and performance departments to inform our pipeline choices, as well as feed back to all these areas any wins/loses we could come up against while doing this.
The production process was quite a unique one for an animated film, in part due to working with a live action director, and in part due to the nature of the look and the project. This was an interesting challenge for us, as we had not done a full feature before and had based some original assumptions on a traditional animation pipeline approach to getting the project done. Early on we saw that the traditional animation approach was not going to be the best course of action for this movie and we quickly started to learn together with George, what we had to do to make Happy Feet. Where this augmented our approach, as compared to the traditional approach was that it kind of changed the ratios/dynamics of storyboard, animatic, previs and editorial. Because we had a large body of motion capture and animation to get through, George favored working with the mo-cap end of it first, as this allowed him to shape things a lot faster, and get an early preview of how things were shaping up, story wise. So our previs was more like setting the spatial relationships and location scouting, then on to motion capture to get the base for the performance down. From there, we went into Lensing (setting cameras) the performances, and all the while maintaining a tight loop through the editorial department. We had to invent ways of providing an editorial look to what we called “Motion Selects”, where George could take a little of ‘this’ from one performance take, and a little of ‘that’ from another, all blended in real time so that it was fast and could be burned out to sequences for the edit rooms, or sent through in an instant to the Lensing team. So this triangle of Edit, Motion Selects and Lensing was what we referred to as the Hub, and this is where the story really started to solidify for George.
What do you consider the biggest challenges of Happy Feet you had to solve?
Some of the biggest challenges were dancing crowds, character and environment surfacing, and generally creating enough flexibility in the pipe to allow for fast interaction and reaction to change. We had kilometers of snow and ice to make look real, and an average of 2-4 million feathers per bird to handle per frame. Some of our best performance results were achieved by caching and baking of the results of some of the heavier processes, such as subsurface scattering, ambient occlusion, and the birds feathers and fluff. There were many more challenges as you could imagine, and we seemed to meet most of them fairly gracefully. The dancing crowds were one that we could not solve with standard crowd systems due to the random nature of the way crowd systems generally move characters about. We came up with a quite simple approach to it in the end, with a tool we wrote called “Horde” which enabled us to place dancing groups of penguins about the environment in either an ordered/patterned fashion, or random. It also allowed us to mess with their timing in an organic fashion as well, rather than an offset that stuck for the duration of the motion. We were able to warp the offset over time, so that no characters were the same offset for the duration of the motion. This meant that a char that was late in one step, could be early on the next, and on time the one after.
Explain us a bit the pipeline and how it was setup for the production
The pipeline was set up gradually as we started the project. For things we thought were going to be the biggest challenge, we started early, like the animation tools and the feather, ice and snow surfacing pipelines. Mostly we stuck to a central layout-focused pipeline, with editorial woven in at every step. It became evident early on that maintaining the story pipe was the most important, which is why it was very Layout and Edit centric. The story and the rest of the pipe was built in layers, initial prep work was done in advance of when we would need it, then we would make more automated tools for each process as we all learned more about how it would fit into the whole.
Rumor has it that animators actually went to Antarctica to study Penguins, is that true?
Almost… I was sent to Antarctica to gather reference for the environment and as much of the wildlife as possible. Being that a lot of the wildlife down there is protected, you have to get authorization to be able to get close enough to observe them in detail. So for most of the trip we were concerned with the Environment, except for where we had the authority to hang out with some of the penguins, and would spend all day photographing their every move and videoing them negotiating there way around. I was sent down there for two trips over three months. It is an awe-inspiring place and a once in a lifetime opportunity. I will always be grateful for being chosen to go down there. The two trips comprised of a trip to Scott Base on Ross Island, the New Zealand Antarctic Base, as well as a boat trip round the Antarctic Peninsula. The two areas were wildly different and chosen because they would best represent the two main areas of the movie. Emperor Land was very much the Ross Island part of the trip, as we were mostly walking kilometers out on the frozen sea to find trapped bergs from the winter that got frozen in. The Peninsula part of the trip was definitely Adelie Land, lots of exposed rocks where they would gather pebbles to build nests. We even got to see abandoned bases and whale stations on that leg of the expedition, which was similar to Mumble’s journey to the forbidden shore.
The film also blends in human characters. Was this considered a risk and what was done to make it work?
It was always planned to have the live action humans in Happy Feet, as a stylistic and time-saving measure. It would have been far too time expensive to have to build and rig the humans in those shots since they don’t play an active part in the story for any longer than two scenes over a 70 scene movie. The stylistic side of it came from altering the grade and comp of the human elements to make it similar to the CG world. George tried where possible to keep them from having things that the other parts of the movie did not, by way of matting all their hair down so no one had wild flowing hair, or making it look sculpted, keeping the clothing and lighting simple and primary so it could fit in with the CG world a bit more. They were also treated a bit in post to take some of the detail out of them.
Which area of the production required the largest amount of custom tools being developed?
Hard to say categorically, since every department on the project had major development of tools for them to complete their tasks. There were departments made around some of the developed tools/processes. The Character team certainly was the leader of the pack, in ‘first out of the gate’ terms because they had to feed the Mocap, Animation, Character Surfacing and Previs teams.
How was XSI used in this production?
XSI was initially selected for Character and Animation tasks, but ended up serving us in a few more areas as well. Modeling was the easy one - the environment guys used it almost exclusively. There was also a department that formed out of a suite of tools development out of necessity, called the Motion Såelects department. We were able to make XSI behave like an online 3D editing suite for the huge amount of takes of motion capture that were acquired for the film. This was made possible with a suite of tools that mimicked video edit suite controls. We also added the ability to write EDL’s for passing take information back and forth between the “Selects Tool” and editorial. Essentially the tool helped us make up performances from editing bits of takes together, and blending them in real time for instant feedback. It meant that we were able to only clean up what was necessary from the raw mocap data, and use the best from all takes, rather the best of the takes.
Which areas of the software do you like the most?
My favorite area is the mixer, with respect to Happy Feet. It was the part of XSI that originally got us looking in that direction for the project, as well as made it possible for tasks like the selection process spoken about earlier. We used it a lot for initial motion blending and manipulation, in animation. However the bulk of the anim work would have been done with the fcurve editor. Other areas that were vitally important were the netview, which allowed us to create a common interface for the Animation team, across all characters.
Which areas do you think should be improved after having used it extensively in production?
There could definitely be more work in the fcurve editor, as well as the referencing system. I know we spent a lot of time staring into the fcurve editor and scratching our heads about the referencing system. To their credit, Softimage did give us unbelievable support throughout the project. We most certainly would have had a lot tougher time without the Special Projects Team.
Where do you see cg software go in the next few years and which areas are still the most difficult to solve?
We are starting to see commercial software go from single tools that solve one problem to a suite of tools in one interface.
Director George Miller in a Mocap session (left). - Voice actor Elijah Wood.
There have been several movies now with talking animals. What other kind of animated feature would interest you a lot?
Something that is a bit more gritty and edgy, like Hard Boiled in 3D. Well, maybe not quite as gritty as that, but you get what I mean. People have been talking about doing something like that for years, but I would really like to see someone be bold enough to fund and distribute something like it.
What do you think of the cg industry in Australia, the talent pool it has and what kind of film projects get done by vfx companies there?
The talent pool is quite deep here in Australia, but a lot of them go off to seek their fortunes elsewhere. There are a lot of talented Aussies out there in the world of CG/filmmaking. The industry in Australia has enjoyed being reasonably healthy over the last 5-6 years and continues to grow. The main bodies of work done here are VFX in live action films, and a huge amount of TVC work for clients worldwide and local. We are starting to see more Aussie companies get their names attached to large FX films. Everyone is starting to get noticed for the work they are doing on overseas content, as well as local grown projects.
Is there anything you would like to say to the rest of the cg community?
If there is one thing that has been a big lesson out of all this, it’s about breaking your process down to its core components/steps before you start. Most of the very well R&D’d tasks go off without a hitch. The most time consuming tasks are generally the ones that are started from a small request that looks like it won’t need any special development, and ends up being refined and refined over and over to the point that it becomes almost impossible to see where you came from, and where you are aiming - as well as being time-consuming to achieve the desired outcome.
Might Animal Logic do more animated features?
Animal Logic will most certainly be doing more animated features. This has been a longtime goal for many people in the company. Now with the success of Happy Feet, it is more of a reality than ever before. It has helped us grow to a healthy capacity for the length of the project and really put our theories to the test as far as building a pipe to support it. Now that we have done all of this, it has shown us how to deal with growing and building our way through major projects.
Discuss Article at the forum
| |
|
