Interviews

Interview: Masanobu Tanaka Animation Director—The Last Guardian

The Last Guardian relied heavily on procedural animation to bring Trico to life. Of all the technologies developed for the character, what was the single most important or successful in terms of making the creature appear alive and believable?

The animation control for the fixed “look-at” was among the most important technologies we built. Initially we used rotation from the neck and then the head; however, the problem with this method was that it made Trico look robotic, almost like a toy snake. With guidance from our creative director, Fumito Ueda, we decided upon a method that first fixed the head position and rotation followed by the neck. Real animals have the ability to fix their head position while the body is moving, which was a source of inspiration. We believe that switching to this method allowed us to add more realism to Trico. We named this method shishimai or “lion-dance control” after traditional Japanese lion dancing.

What, if any, were the unique challenges of a giant quadruped character in terms of navigation around the environment, and what animal references did you use most of all for the many subtle animalistic details on display?

Controlling a quadruped animal is no easy task, which we made even more challenging with our desire to build a controller that allowed a huge creature like Trico to squeeze into tight spaces and walk around freely. Major development obstacles included slopes, stairs, corners, and jumping into an abyss. We took great pains to make the movement natural by combining hand-key and procedural animations. In creating Trico’s movements, we referenced cats and dogs, and incorporated adorable animal gestures that were familiar to us.
Often it feels like The Last Guardian is the culmination of the studio’s previous projects. What was learned from animating the many colossi in Shadow of The Colossus in terms of bringing weight to Trico?

Yes, the experience with Shadow of The Colossus helped us a lot in the creation of The Last Guardian’s animations. However, the colossi were not alive, as it were; thus their movement was stiff compared to our desire to make Trico’s more fluid and lovable. With Trico, we definitely incorporated our learning about expressing the colossi’s weight while adding a degree of softness and elegance.

Trico generated strong responses from players in terms of invoking both empathy and frustration (the latter because he wouldn’t simply obey orders), as real animals are wont to do. How important do you believe “disagreeable” AI is to rendering a convincing character?

While I was not in charge of Trico’s AI, my personal opinion is that it’s important to express the AI’s will, not to be disagreeable. However, I think Trico’s AI was not as technically elaborate as to allow for it to have its own will. Instead, we expressed its will with animations, staging, and so forth. With Trico, we emphasized simplicity of AI to allow for easier controls by artists.

The Last Guardian creative director, Fumito Ueda, notably has an animation background. Is there a shared animation philosophy that has developed over the course of the studio’s games, and if so, what might it boil down to?

We have so many philosophies that it is hard to express them all. Regarding The Last Guardian, we took great reference from Fumito Ueda’s pet cat. Personally, I used to have a dog but never a cat, so I frequently referenced the animations of his cat. Also, we were told to emphasize the expression of weight and eliminate popping between animation transitions, as those errors would spoil all the good parts of the game that we worked so hard to bring to life.

Interview: Bruno Velazquez Animation Director—God of War

Revitalizing the style of such a well-known character as Kratos after working with him for over a decade must have been a fun and challenging experience. Were there any animation rules or dos-and-don’ts you built up for him over the years?

Over the course of seven games, we kept an internal document called “Kratos Rules” that the team adhered to for all aspects of realizing Kratos. Everything from the way he looked to the choices he would make and of course the way that he would move and be animated. These rules included such things as “Kratos always maintains eye contact with the enemy unless there is a major rotation; otherwise he is dialed in and focused on the enemy at hand.” The hit frame should personify this idea so his hits feel deliberate and final.

For the new God of War, we looked over each rule and revised them accordingly to reflect an older, more measured Kratos now responsible for raising his son Atreus. For example, “Kratos never runs away or takes a step back from the enemy. He is always moving forward or making the player feel like he is about to move forward.”
While this is still true in the new game, we added the following to the rule: “When forced to fight, Kratos will advance and maintain a smart dominance of the situation. However, he is older and wiser and attempting to turn away from violence as a solution. He wants to change to show his son there is a better way to resolve a problem.”

This was the first entry to the series that heavily utilized motion-capture. What were some of the key challenges of shifting the team over from a keyframe approach, and how did you maintain consistent style between the humans and creatures?

At the beginning of the project I was concerned about getting a team primarily built for keyframe animation to transition to a motion capture–heavy project. The way that it ultimately worked out was to get the whole team to jump head-first into the process. We purchased a motion capture suit that did not require cameras or a complicated setup and encouraged the animation team members to try it themselves. We spent many hours taking turns at wearing the suit and recorded as much data as we could.
Once the animators got to apply their own motion to our various characters, they were able to quickly get past any misconceptions and concerns about the technology and became more comfortable with it. This process emphasized the use of motion capture as

one more tool available to us and by experiencing the process themselves from beginning to end it helped them to accept its use. In the end, we did hire professional actors and stunt performers to capture the motion, but having the animators understand the process from beginning to end was a big step for us.
Once we got comfortable using more realistic animation data we were able to determine how much we could push our fully keyframed characters. We always used Kratos’s and Atreus’s motions as a barometer to keep us consistent.

You state that achieving the balance between fun and responsive gameplay and natural movement is one of your key goals. Do gameplay requirements ever conflict with animation and how do you typically resolve them?

Gameplay requirements always seem to affect and conflict with our inner animation desires. However, our team understands that gameplay is king. This is why we worked very closely with the combat designers who are responsible for building Kratos, Atreus, bosses, and all enemy AI characters. As we figure out the gameplay and identify the timing of motions that will serve the gameplay best, we identify data that includes hit-frame numbers and the total amount of frames a combat move needs to feel responsive. The animation team then finds creative solutions to maximize the use of keyframes available by focusing on the spacing of our poses.
A good example of this is the motion of the Blades of Chaos, which are keyframed frame-by-frame because the spacing and positions of each frame had to work for gameplay as well as being smooth enough so the trail effects wouldn’t look linear. When working on the cinematic kills where Kratos dispatches his enemies, however, the animation team had much more freedom to let loose with the animation timing. This allowed for the animation team to reach a nice balance with the gameplay team.

A notable visual style choice in God of War was a complete omission of camera cuts in cinematics and action scenes. What was the reasoning behind this decision and what were the related challenges?

Early on in the project we decided that we wanted to make this God of War more grounded and unflinching. Our goal was for the player to feel like they are right there next to Kratos and Atreus, along for the entire duration of the adventure by only focusing on their perspective. Because of this we decided not to cut the camera to show a passage of time or a glimpse of what “the bad guy” was doing in his lair. This was all in aid of a more personal story, so not cutting helped us to make it feel more realistic and visceral than previous games.

We quickly discovered that it was going to be very challenging to shoot scenes, and planning was critical. The moment that things fell into place was when we changed our approach from shooting like a film to instead producing more of a stage play. Once the actors adapted to approaching scenes like this, it really made a big difference. We would spend at least one day rehearsing before shooting so the actors could do a table-read together, absorb all the previsualization we had prepared, and rehearse the staging and movement with the virtual camera. In the end, it all worked out as a well-coordinated dance between our actors and cinematographer.