Avalanche Studios talks scalar fields and digital clay ahead of Rico's return
Avalanche Studios has always prided itself on creating some of the biggest world in video games, and the developer is once again upping the ante for the upcoming Just Cause 3.
In a recent developer diary, the studio revealed that it actually had to build new technology in order to sculpt the ficitional Mediterranean region of Medici – and the origins are slightly unusual.
We caught up with Avalance's co-founder and CTO Linus Blomberg (pictured) to find out more.
What limitations did your previous world-building technology present when beginning work on Just Cause 3?
As with most other open world games, the landscapes of Just Cause 2 were based on 'height maps'. Height maps are a two dimensional representations of terrain elevation. As such, it is limited in the way that it can't represent complex volumetric or overlapping features likes overhangs, tunnels, and cave systems.
How have you built on the technology used in previous Just Cause titles? How has this overcome those limitations?
For Just Cause 3, we developed an entirely new terrain system where the data is represented as 'scalar fields'. Unlike height maps, scalar fields are true three-dimensional representations of topography. Previously, it has primarily been used in medical imaging to visualize the volumetric data generated by CAT-scans – basically three-dimensional x-rays – on super-computers. But though years of research and development we have found ways to bring this to games hardware.
What else are you able to do now that you couldn’t before when it comes to building the game world?
This literally provides us with a whole new dimension of freedom when it comes to sculpt our worlds. So instead of just working with size and scale in the horizontal plane, we can now also work with depth and complexity in the terrain.
What was important to ensure remained the same, to ensure it still feels like a Just Cause world?
Our credo when it comes to our worlds is: 'if you can see it, you can go there. If you poke it, it reacts.' and I think that is one of the great charms of the Just Cause worlds, and it is something we would never compromise on.
Why was verticality important when developing the new tech?
In games like Just Cause, you spend as much time in the air as on the ground so the world must feel interesting from that perspective as well. That's why it is so important to work with the vertical dimension in the landscape.
In Mad Max on the other hand, you spend most of your time on the ground so there we needed to focus on other things, like the detail and atmosphere of the wasteland. For the future, however, we believe all of our games can benefit greatly from the added creative freedom this technology provides.
What challenges did this present and how did you overcome them?
Adding a dimension to the terrain representation also means many orders of magnitudes of more data to process. This has required us to rewrite our entire tool chain to deal with this increase. The new tools rely on the massively parallel processing power of the modern GPU's to process and compile the data, and to maintain interactive editing frame rates.
Scalar fields give us a whole new dimension of freedom when it comes to sclupting our world. So instead of just working with size and scale in the horizontal plane, we can now also work with depth and complexity in the terrain.
Why were tunnels and caves also important? What is the advantage of your new system compared to simple meshes?
Of course, tunnels and caves are nice and enriching additions to our game worlds in itself, but it is much more than that. To me, it adds a whole new level of excitement to the experience and atmosphere of our games. Now it is not just about seeing a distant mountain and wanting to see what's behind it, now it is also about what is inside it! Or even, what's beneath the ground I'm standing on? It really makes exploration and world design so much more interesting.
Simple meshes are bad from many perspectives. They are very labor-intensive to create, they consume a lot of memory, and it is difficult to make good looking level-of-detail transitions between them.
What is the advantage of being able to sculpt in three dimensions? How did this help the development of Just Cause 3?
From an editing perspective, it is much more natural and intuitive to sculpt in three dimensions than to create two-dimensional height maps. Internally, we refer to the tech as 'digital clay' because of the similarities to sculpting in clay. In Just Cause 3, it enabled us to create environments we simply wouldn't have been able to otherwise, due to their complexity.
How did you balance building new tech and features while still enabling the series’ acclaimed level of draw distance?
The draw distance and seamless streaming is key to our technology. Those things are core and center even in the first drafts of new technologies, and everything is designed around that.
What did you have to take into account regarding the in-game destruction system when building the world? How did you create a playground that lends itself to such levels of chaos?
That is primarily a challenge from a game design and level design perspective. Like, how to design a military base so that you can achieve your mission objectives even if you've blocked something with a huge pile of debris from a destroyed structure. Sure, it is not easy to build the technology for it, but we work closely with Havok on that.
Just Cause prides itself on having the biggest open worlds available in gaming. With rival titles increasing the size and scope of their worlds, how will you maintain their claim?
We've been focusing on providing the best open-world experiences possible for almost 13 years now, and are constantly improving and evolving on our technology and design principles. That is not something you easily catch up on, no matter how many people or how much money you throw at it. We are determined to stay best-in-class in our field for many years to come.