November 02nd, 2017

The making of Journey For Elysium: A VR Experience: Part 1

How I got involved in the story

My first experience with VR must have been at a trade show in the 90’s, where I got to try on a VFX1 and play a very blurry rendition of Duke Nukem 3D for a headache-inducing minute or two. In hindsight, it’s not hard to see why VR failed to take off back then. But the experience — no matter how physically uncomfortable — did leave an impression beyond a headache, and so when Palmer Luckey launched his first Oculus kickstarter in 2012, I took a keen interest. When got to try the DK1 (Development Kit) I decided to pre-order the DK2 when it was announced, and have been firmly aboard the VR hype train ever since.

When I joined the team at Little Miss Robot, I had been playing around with Tilt Brush, Google’s Virtual Reality painting application. At LMR, we’re invited to give a short presentation every so often about things we’re working on, or particularly passionate about, and so for my first Show Off Friday-talk I decided to present some of my Tilt Brush sketches and gave a short talk about VR as a medium for artistic expression.

People asked if they could try it themselves, so at the yearly company weekend I brought along my PC and Vive and gave a VR initiation to everyone, which ended up being a big hit.

As a result I was brought onto the Journey For Elysium project late January this year, a VR Storytelling Experience for which Dave and Alex had already secured funding. I soon found myself as the resident VR-guy.

The game is set in the Greco-Roman mythological underworld, where the player has to navigate the river Styx
The game is set in the Greco-Roman mythological underworld, where the player has to navigate the river Styx

Preparing for the Journey ahead

I had some prior experience with Unity, but nothing serious really, just a few late-night experiments. A bit of a generalist, I had read a lot of articles about games-programming and lots of experience working with Flash, Processing, Javascript and probably dozens of different tools and programming languages, so I was reasonably certain I could make something playable in a modern engine like Unity.

What I was less sure about was my ability to make the graphics look anywhere as good as the concept would demand… I knew enough about the production process to realise that I didn’t know enough to make game-ready 3D assets, and the tight schedule would probably not allow me enough time to learn.

So I was about to suggest a stylised low-poly look when Alex and Dave informed me that I would have the help of a very talented intern from the Digital Arts & Entertainment school in Kortrijk, Xavier Allen, who could create all the 3D assets we’d need. I was amazed at how quickly he was able to turn out optimised meshes with detailed normal maps, and so this shifted our graphical ambitions skywards.

Still we decided to compromise a bit: we would only use normal mapping, and forgo most other texturing options available for the prototype. We preferred to have a large world with a consistent level of (low) fidelity, rather than have one tiny level at AAA-standards, or worse: a large level with huge inconsistencies in terms of detail, which would be detrimental to player immersion.

The first iteration of the opening was a lot more “fire and brimstone”-y than what we ended up going for
The first iteration of the opening was a lot more “fire and brimstone”-y than what we ended up going for

Virtual Concept Art

The next challenge to tackle: making sure the vision of the project was understood between the different team members and everyone was in sync on how their work would fit into the whole. The traditional answer to this was concept art: talented illustrators and painters make series of sketches with increasing levels of visual fidelity until the art-director’s vision is nailed down. Then these primarily 2D artworks are passed on to teams of technical 3D artists to turn into fully realised props, creatures, environments or other assets needed to make a game (or film).

Now, we had concept art already, but the amount of art needed to flesh out the level as we had envisioned would be impossible to produce in the limited time we had. We knew through our experience with Agile that it’s generally never a good idea to spend too much time on a prototype: better to make many, smaller-scoped prototypes that are quick to iterate on, than to make one single big prototype that can’t easily be changed without spending a lot of time.

So, I suggested we skip 2D and go straight into VR by using Tilt Brush as our tool (and medium) of choice for concept-art. I sketched out the levels inside of VR with all the freedom of a drawing tool. As a result I was able to quickly get accurate feedback on it in terms of scale and feel, just by letting people teleport around the sketches. Then, using Tilt Brush’s built-in exporter, I could export 3D models that could be loaded in our modelling tools (3DS Max in this case), to serve as the rough framework to build the final assets on top of.

Around this time Google also introduced the Tilt Brush Toolkit for Unity, which lets you import your exports straight into Unity, which meant an even tighter integration between concept and prototype was now possible. Whenever I needed some small UI bits, a placeholder torch prop, or even a floating skull to prototype enemies with, I just fired up Tilt Brush, doodled something and exported it, confident that the scale and placing would match.

Here’s a montage of some of my prototyping sketches:

I made liberal use of the then relatively new Media Library feature, which allows you to import your own images and 3D models into your sketches. I made my own set of 3D “brushes” in the form of various low-poly rocks, pillars & stalactites/stalagmites which could be mixed and matched to quickly create many different cave-like environments. This technique worked so well we ended up using it with our high-quality assets as well.

Now as Xavier, our 3D artisan, began the long, arduous process of meticulously sculpting, unwrapping and baking everything, I could dive into the real challenge: making it all actually work, but I’ll save that for a future article.

Author image

Gilles Vandenoostende

Interaction Designer

Digital designer & maker of things. Passionate about new technologies and loves to explore new realities: Virtual, Augmented or beyond.