The following is an excerpt of an interview that appeared on Art of VFX on April 27, 2021. The entire interview, conducted by Vincent Frei, can be read here.

INTERVIEW:

Earlier this month, Bryan Hirota explained in detail the work done by Scanline VFX on Zack Snyder’s Justice League. He’s back to tell us about the epic confrontation between titans!

Chris Mulcaster started his career in visual effects in 2010. He joined Scanline VFX in 2013 and has worked on a wide range of projects including Game of Thrones, Aquaman, Spider-Man: Far From Home and Zack Snyder’s Justice League.

Jim Su has over 20 years of experience in visual effects. He has worked on several shows such as The Three Musketeers, 300: Rise of an Empire, Star Trek Beyond and Ant-Man and the Wasp.

Prior to joining Scanline VFX, Kishore Singh worked at numerous studios such as Framestore, ILM and MPC. He has worked on projects like Gravity, Warcraft, Black Panther and The Meg.

Jonathan Freisler has been working in visual effects for over 10 years. He has worked on projects like Ted, After Earth, Mad Max: Fury Road and Tomb Raider.

With over 20 years of experience in visual effects, Eric Petey, has worked on many films such as Transformers, Guardians of the Galaxy Vol. 2, Rampage and Captain Marvel.

How did you feel bringing this epic confrontation between two iconic movie monsters to life?

Bryan Hirota: I was quite excited at the opportunity to work on this film. The original King Kong vs. Godzilla was one of my favorite films when I was young as it was shown on afternoon television. When it looked like I was going to get a chance to contribute to a remake of a classic kaiju film, I felt quite lucky.

How was this new collaboration with VFX Supervisor John ‘D.J.’ DesJardin?

Bryan Hirota: DJ and I often look for opportunities to collaborate together. I was finishing up our work on Aquaman when he was hired to supervise Godzilla vs. Kong. They wanted to make sure Scanline could do big creatures so Stephan Trojanksy and myself met with DJ/Tamara and picked some shots to do as a test. We took the Kong asset from Kong: Skull Island as a starting point and bulked him up and aged him to reflect the amount of time that had passed from when Skull Island was set. They liked it so much they had us continue to develop the hero asset.

How did you organize the work with your VFX Producer?

Bryan Hirota: I’ve worked with both VFX Producers Julie Orosz and Ryan Flick on a number of projects. Both of them were instrumental in keeping the show run smoothly as the facility needed to switch our worldwide operations to a work from home operation during the pandemic. They deserve a lot of credit for keeping the teams in all of our offices informed and moving forward with the needs of the show.

How did you split the work among the Scanline VFX offices?

Bryan Hirota: The company has grown to have offices in a number of locations now. Luckily everyone shares the same infrastructure so we don’t have to worry about transferring assets and so on between offices. An artist can immediately pick up work created in one office in another.

What are the sequences made at Scanline VFX?

Bryan Hirota: We developed what we called “old man Kong” asset that was shared with Weta and MPC for their sequences. The sequences we worked on were the initial attack on Pensacola by Godzilla, Kong on the transport ship including his fight with Godzilla at sea, Mecha Godzilla in the arena, Mecha Godzilla’s battle with Godzilla, and the team up with Kong and Godzilla against Mecha Godzilla in Hong Kong. We also did the shots where the two remaining titans acknowledged each other and went their own ways.

Can you explain in detail about the design and creation of Kong?

Chris Mulcaster: At the beginning of the project we took a trip to the LA Zoo to study the gorillas and capture valuable photo reference and videos that we used in our development. As a base, we started with the Kong asset from Skull Island who was an adolescent and much smaller. Our model lead, Damien Lam, went through several model concepts and paintings before re-sculpting Kong to show the age deterioration and battle scars that he had incurred in the 50 years that had elapsed. In that time Kong has also substantially bulked up, for the impressive weight and muscle mass on Kong, we studied veteran body builders and how a strong physique degrades over time.

In the ocean battle, Kong stood at around 300ft tall which is 3-4 times larger than he was in Skull Island and in our Hong Kong sequences, we had to scale him up to over 500 feet.

We spent a lot of time developing the fur for Kong and introducing aging grey and white hairs using the V-RayHairNext material. Once we had established the base dry look, Changmin Han, who created the textures and shading for Kong, also developed a number of other looks including various stages of wetness, ‘oily Kong’ (where he is drenched in Mecha’s mechanical oil) and a dusty version. This was further supplemented by our effects team who also simulated small building debris and dust falling from Kong’s fur as he interacts with the buildings around him.

How did you develop the body muscle system for Kong?

Jim Su: There were two setups, one using Ziva VFX which is a physics-based muscle simulation, the other was using a procedural jiggle rig. The jiggle rig takes advantage of Maya’s GPU deformers and has real-time playback. Most of the shots run through a jiggle pass since the turnaround is quick and results can be rapidly iterated. The jiggle setup is anatomical, respecting muscle insertions and attachments. Any shots that required more hero muscle simulations were run through Ziva. The setup requires modeling the skeleton, muscles and a tissue with thickness consisting of the fascia and skin. There is a muscle simulation pass, and the fascia and skin combined simulation pass.

Can you tell us more about the fur and groom work?

Jim Su/Kishore Singh: The groom was done using XGen core in Maya by our groom lead Tarkan Sarim. The fur system and simulation setup are run with Maya’s nHair. As the lead vendor on the show, we were tasked with creating the master version of the groom and provided the groom data to the other studios to match our version.

Due to the complexity of the groom, we ran into technical challenges and limitations. To overcome these, we experimented with splitting the groom up into 10 smaller chunks. This resulted in quicker hair generation and previewing times within the viewport. We also utilized the viewport render features in Maya’s Viewport 2.0 which provides full shading for fur as well as lighting and shadow previews. This gave us a very close representation of how the rendered fur would look without us having to render time consuming tests, increasing our efficiency and iterative abilities greatly.

We completely overhauled our hair system to allow for interactive manipulation of the guide hairs, and created a multi-shot hair simulation tool. All fur elements were created by sculpting guide curves. At the beginning to block things out we would start with a smaller amount of guide curves and try to push the groom to about 70% completion. We were able to change the parameters of all grooming attributes on the fly without having to redo everything like is the case with a purely sculpting based grooming workflow. Although we were using a procedural approach, we were still able to utilize some of the purely sculpting based grooming tool features to add finer details when required.

Every individual strand of hair was cached out and rendered in 3ds Max, made possible with due to the creation of a procedural tool that writes out all the XGen hairs into a custom alembic file for each frame that we referred to as a hair cache. Once the hair cache is written out it can then be loaded into 3ds Max as a V-Ray proxy and used for rendering.

Kong is really emotional in some sequences. Can you elaborate on this?

Jim Su: We established a new facial mocap system at Scanline for this movie. We had our Animation Supervisor, Eric Petey, act out a lot of the hero performances at our in-house mocap stage in an Optitrack Motive suit with a Faceware Mark III head mounted camera. The matchmove department would turn the video footage into keyframes using Faceware Analyzer and Retargeter. Animation would add creative embellishment to the facial performance. A new facial rigging system was designed to handle the facial nuance such as non-linear wrinkles and interconnecting of facial shapes. Even with the amount of facial shapes and complexity of the rig, it had real-time playback by taking advantage of Maya’s Parallel Evaluation and GPU deformation. We also obtained the Chimp FACS courtesy of the University of Portsmouth to understand the differences between human and chimpanzee faces.

Can you tell us more about Kong’s eye work?

Jim Su: We rebuilt our eye model for Kong’s eye, introducing the conjunctiva on the eyeball within our rigging and lookdev workflow. Introducing this extra layer to the surface of the eye meant we were able to get proper coloration and more realism to our eye model. We also accurately replicated the shape of the cornea, how it refracts light and interacts with the iris so the iris appeared correctly. We had full control of the meniscus and therefore we were able to control the mix of oil and water that sits on the surface of the eye.

For more on how Scanline VFX brought Godzilla and Kong to life, please read the original interview on Art of VFX.