Kingdom of the Planet of the Apes marks the tenth instalment in the hit franchise, which debuted more than 50 years ago. Director Wes Ball called on the team at Weta FX to create the film’s 1,500 VFX shots. The feature is Oscar nominated in the Visual Effects category, with honors shared between VFX supervisors Erik Winquist and Stephen Unterfranz, senior animation supervisor Paul Story and special effects supervisor Rodney Burke.
According to Winquist, Kingdom marks the fourth time the studio has contributed to the Planet of the Apes franchise, with director Ball taking things in an interesting direction this time around.
“As the timeline jumped forward many generations, it meant we needed to conceive of a world where nature had long since taken over the abandoned remains of human civilisation, and our apes were all using speech as their primary form of communication,” Winquist explains. “To provide the foundation for our distant future landscapes, we chose locations in and around Sydney, Australia, to shoot in this time around, which provided the sun-drenched environments with rich vegetation and overgrowth, coastal areas and a sizeable backlot where about a dozen different exterior sets were constructed. We digitally augmented the photography in post to extend partial set builds and add ancient ruins of our present human infrastructure. Sometimes this was barely noticeable hints of building remnants. Sometimes it was substantial or complete replacement of the landscapes with digital skyscrapers buried in millions of vegetation assets.”
As with the series’ predecessors, the VFX team leveraged on-set performances captured during principal photography.
“This took the form of calibrated exterior mocap volumes in our larger sets, and leaned on what we call ‘fauxcap’ — using an array of phase-locked global shutter witness cameras to record their movements — in more remote or challenging environments, such as a narrow forest road, or the shallow creek that features in a human-hut sequence in the second act.”
In both cases, Winquist says the team used a stereo pair of head-mounted cameras to capture facial performances.
“These provided us with depth information of our actors' faces that fed into a Facial Deep Learning Solver (FDLS), which we further improved for this show. It was leveraging the FDLS that made for a more efficient facial animation process on this picture, which was important, as there is more spoken dialogue among our 12 new hero-level characters in this film than there was in the previous trilogy combined.”
The first baseline pass at the apes' facial performance from the FDLS provided a consistent starting point for the facial animators, who could then spend their time focusing on translating all of the nuance of the actors' choices on the day, and producing an ape performance that affected the audience in the same way that the actors would have.
“But this story also called upon us to deliver a number of sequences that had significant effects components to them,” Winquist notes, “from a village-attack scene early in the film, where everything was on fire, to substantial water simulation components between half a mile of river rapids, pounding waves along a cliff, and a torrent of sea water, which floods our primary location for the film's third act.”
These sequences typically incorporated a physical effects component on-set as well, where the production team had aspects of those effects in-camera to augment and extend. A network of real LPG gas flames in the attack provided light and behaviour reference for digital fire simulations, and a shallow tank with rushing current gave the cast something to struggle in for the river scene. Large dump tanks provided crashing waves during the cliff walk.
“The river scene in particular was among our most daunting challenges, given we needed to sell shots with our human actor and a digital orangutan with his long hair occupying the rushing water together, in broad daylight, in an impactful moment where the film's beloved mentor character delivers his final words,” Winquist explains. “Our simulation framework and tight coordination between artists ensured an incredible amount of photoreal detail in the final images, and made for a very affecting farewell.”