In 1992, author R.L. Stine published Goosebumps: Welcome to Dead House, a children’s book in which a houseful of zombies causes chaos in the small American town of Dark Falls. Five years later, the Goosebumps series ran to over sixty books, with individual volumes selling millions of copies every month and regularly appearing on bestseller lists.
Since then, Stine has continued to write new Goosebumps stories at a frankly phenomenal rate. The original series spawned a TV show that first aired during the late 1990s, at which time a feature film adaptation was also in development – with Tim Burton attached as producer – but the project bit the dust.
Now, R.L. Stine’s creation has finally hit the big screen With Rob Letterman directing from a script by Darren Lemke, Goosebumps chronicles the adventures of teenage protagonists Zach (Dylan Minnette), Hannah (Odeya Rush) and Champ (Ryan Lee) as they attempt to return the monstrous creations of R.L. Stine (Jack Black) to the pages of the books from which they’ve accidentally been released.
Overseeing the visual effects for Goosebumps were production VFX supervisor Erik Nordby (Elysium, The Amazing Spider-Man 2) and VFX producer Greg Baxter (X-Men Origins: Wolverine, Divergent). The majority of the film’s 900-odd visual effects shots were handled by MPC at their facilities in Vancouver, Montreal and London, with MPC Santa Monica doing extensive previs and postvis. Additional support was provided by Vitality and Instinctual.
In this VFX Q&A, Cinefex spoke to both Nordby and Baxter about the challenges involved in bringing Stine’s menacing menagerie of monsters to life.
How did you first get involved with Goosebumps?
ERIK NORDBY: We were brought on board early in development. Preproduction hadn’t officially begun, but Rob Letterman had already spent some time working through a lot of designs. This proved very valuable for us because most of our full CG creatures had a solid design concept fleshed out – and in some cases fully approved, all the way up to the studio level. As many teams know, this is often very difficult to achieve early on, but is incredibly valuable in solidifying a shooting approach and an efficient asset plan.
GREG BAXTER: I got a call from Sony and met with Rob Letterman back in September 2013. Rob had already been developing the story and designing some key monsters for several months. Meeting with Rob and hearing about his vision for the show was amazing. I left that meeting knowing I was going to join this project.
Were you familiar with the Goosebumps series of children’s books?
ERIK NORDBY: Like most of the crew, we were familiar with them and their popularity, but didn’t have much specific knowledge beyond that.
GREG BAXTER: Familiar? Yes. Had I read them all? No. I had to do a little research on the creatures we were tasked with bringing to life. But I could really see how the series had been such a hit – the stories were more fun than scary and were all very engaging.
ERIK NORDBY: One of the biggest surprises was how intensely revered the book series is by a certain generation. One of Rob’s early points of connection – which I thought was really smart – was to equate our generation’s love of early Amblin films with this generation’s love of Goosebumps. The Goonies, for instance, was a good touchstone for us in remembering that love of the adventure/horror balance.
How closely did you work with Rob Letterman?
GREG BAXTER: Rob was fantastic. We worked very closely with him throughout the almost two-year process – from concept through development, shoot, editing and final VFX look. His background in animation and VFX compositing helped our dialogue tremendously, as we could jump directly into technical notes and bypass some of the usual steps to get from director’s vision to final on-screen content.
How long were you working on the project?
GREG BAXTER: Twenty-two months from September 2013 to July 2015. Although, honestly, I’m still working on it with wrap, stereo and home video details. VFX keys on shows like this are among the first on and last off the project.
What was the biggest challenge for the VFX team?
ERIK NORDBY: This film had an extremely wide visual effects spread … without a wide budget to go with it! That was the number one VFX challenge on Goosebumps. There were many all-CG creatures, with each one requiring a different approach and a unique balance of R&D and design development.
GREG BAXTER: The monsters were our major task, for sure. We not only had to create fully detailed hero characters, but they were all so different – in size, shape, texture and movement – that each monster was a huge project on its own. We had a hero Abominable Snowman, Army of Garden Gnomes, 30-foot Praying Mantis, Werewolf, Vampire Dog, Man-Eating Plant, Ginormous Blob …!
How did you approach integrating all these outlandish monsters into the live-action setting?
GREG BAXTER: Rob wanted this to be a very photoreal film, with even the most fantastical creatures feeling as though they existed in the real world. While we were able to reference existing animals, insects and objects for the Mantis, Werewolf, Gnomes, Poodle – even the Abominable Snowman – some of the others were a bit more difficult. With the Blob, for example, there was very little reference to draw from other than the book and Rob’s concept of what it should do in the movie.
What else did you deliver apart from the monsters?
GREG BAXTER: There was a lot more VFX work, from matte paintings and set extensions to complex FX simulations. We had to run full digital doubles of our actors for the Ferris wheel ride at the end.
You spoke about the enormous differences between the various monsters. Can you give us an example of some of the extremes?
ERIK NORDBY: The Blob required a huge amount of fluid-sim R&D – many months’ worth – and almost as much work in terms of look-dev. It was amongst our most intangible creatures on-set, since its scale was so mutable. The Blob had to fit into our blocking instead of driving it.
At the other end of the spectrum there’s the Werewolf, which required a very structured on-set approach. Since the Werewolf was a human-sized biped, it needed to be directable and shootable, and the cast needed to be able to act around it. This meant a human proxy stand-in, with a high degree of athleticism and wolf-arm extensions. We added tracking markers which allowed quick turnaround for first pass blocking, and kept the edit fluid before we had to lock things down.
Where did some of the other monsters sit on the spectrum?
ERIK NORDBY: Abby, the Abominable Snowman, was a hybrid: too large to have a human actor perform, but small enough that we could provide a blocking eyeline reference. Our PAs ran through the scene with Abby’s head on an eight-foot-tall stick. The gnomes were shot with hand-made stand-ins which the actors could interact with and smash. Shooting empty plates allowed us to fully replace and fill the frame with their CG equivalents.
How about lighting reference – did you use regular grey/mirror spheres?
ERIK NORDBY: We really tried to leverage creature-specific reference spheres. During preproduction, we tracked down many fur samples, and Rob helped us pick a good, solid, real-world equivalent for our furry creatures. We wrapped the samples around spheres, and combined them with skin samples – leather or a prosthetic piece. For every setup, we would shoot these spheres in their respective lighting. We would also shake the fur around, giving our FX artists great dynamics reference.
Let’s look at one of the creatures in more detail – the Werewolf, for instance?
GREG BAXTER: The initial 2D concept design for the Werewolf was provided by Carlos Huantes. Carlos then built a miniature maquette which we cyberscanned via 3DS in Burbank. This was our initial starting point for MPC to model and augment from.
Who performed the Werewolf on set?
GREG BAXTER: John Bernecker was our stunt proxy for the Werewolf. He performed on-camera for just about every Werewolf shot, including running on all-fours atop a grocery store freezer. Erik and MPC Supervisor Pete Dionne built a custom facial capture rig with Go-Pros and a bicycle helmet, capturing John’s snarling for additional animation reference.
Tell us about the development of the digital Werewolf character.
GREG BAXTER: MPC went round after round trying to solve the intricacies between human and canine facial expressions for the Werewolf close-ups. We often had to dive back into the model to solve some subtleties in facial performance. A canine snout doesn’t allow for much range, so we had to cheat that a bit in the model.
The run and walk cycles were also problematic, due to our design having the hind feet act as bipedal feet. The way a canine’s ankles bend results in awkward movement, and we didn’t want to come across comedic and so detract from the scarier moments.
Then there was the drool! What “real” werewolf drool might look like and how it performs had to be tested and re-tested to read the right way in the comedic beat when R.L. Stine (Jack Black) is hiding under the produce shelf in the store.
Towards the end of the film, a Ferris wheel detaches from its mountings and rolls through a forest. How did you approach this spectacular sequence?
ERIK NORDBY: The Ferris wheel sequence was meant to be visceral, entertaining and, above all, believable. We needed to create a shooting situation where we could achieve all the photography of the kids practically and insert them into a wild but well-planned ride.
Tell us about the practical shoot.
ERIK NORDBY: The previs existed early on, and it was a great guidepost. From that, we created very elaborate techvis which focused on one central idea: the Ferris wheel cab that the kids were in would be fixed in space and the camera would inherit all the wheel’s motion.
John Frasier, and his special effects team built a cab that could be safely and securely hung in the middle of our greenscreen stage. It could swing, but would not travel in any other axis. We then inverted the motion of the cab into the camera in the techvis, to illustrate to the camera and grips what they needed to accomplish with the Technocrane. There was a little give and take because the moves were sweeping and the wheel was so large. We would shoot each take, roughly key it, and comp in the previs background to check if we had accomplished our task.
Since you were shooting on a greenscreen stage, presumably the backgrounds were all created digitally?
ERIK NORDBY: Yes, everything besides the kids and their cab was CG. This required a fully-rigged CG Ferris wheel with all the other cabs, a kilometre-long CG forest for the wheel to run through, and distant 2½D matte paintings for the high angles. Once we became embroiled in the sequence in post, it was clear we needed a lot more tree interaction than previously thought. This pushed the complexity – and therefore the render time – through the roof.
How do you feel about your work on Goosebumps, looking back?
GREG BAXTER: We had an awesome crew on this show. Everybody really enjoyed the experience of making the movie … except for the late-night shoots in the hot, humid, bug-infested Georgia woods!
One memorable moment came when Rob had to direct an army of both practical and virtual creatures attacking the high school. Blocking out that performance required a lot of comedic work. In the middle of it all there was a crazy big thunderstorm, so we had to seek cover while the greenscreens took the brunt of the storm. But we survived, rebuilt the greenscreens and resumed shooting.
Special thanks to Stacey Leinson, Rebecca Rehak and Mick Mayhew. Goosebumps photographs copyright © 2015 and courtesy of Columbia Pictures/MPC Film.