I have being working in the AAA industry for tha last 3 years and the crunch is what is forcing me to find something else to do in life even if I love 3d. Some places may be more respectful with their employees but in my experience the crunch is even calculated in advance cause they know the workers will accept that. Some people is very passionate and don´t mind to do it and that is fine but a lot of people have families and they want to build a healthy environment with them or other goals outside the working ours. Not to mention non-payed overtime and other abuses I faced. Hope this industry fixs this problem.
Those tilesets are sexy. Seeing new tilesets is like getting introduced to a new lego set.
We were fortunate to talk to Eric A. Anderson – the art director of Cyan’s Obduction and the artist, who contributed to Jonathan Blow’s The Witness. Eric is an amazing developer and an incredible world builder. In this interview he discussed the way he approaches world creation in games, talked about the artistic side of things and discussed the way he worked with Unreal Engine 4.
My name is Eric A. Anderson. I live in Spokane, Washington, and I’ve been developing video games for over 15 years. I have been with Cyan (creators of Myst and Riven) on and off for most of that time, and I’m currently Art-Directing Cyan’s latest project, Obduction. Over the years, I’ve worked on numerous games – large and small – the most notable of which are probably Uru: Ages Beyond Myst (Cyan’s ill-fated Myst MMO game, released in 2003) and MystV (2005). More recently, I was hired by Jonathan Blow as one of three core artists for his ambitious project The Witness (2016), for which I constructed roughly a third of the world and helped to define the very unique look of the game.
I may not be a lot of things, but I think I might be some kind of specialist in the “walk around a complex world and solve complex puzzles” genre, because that’s what I’ve spend most of the last two decades doing. There were also some less-than-great mobile-game efforts, as well as countless commercial animation and work-for-hire jobs throughout… But sometimes an artist just needs to pay the bills, especially in the often tumultuous games industry. I think it’s better to look at every experience, including the failures (ESPECIALLY the failures) as an opportunity to learn, grow, and apply that knowledge to future projects.
World building is a vast topic, and it would be impossible to summarize all of the important elements required to construct a successful game world, but I might be able to isolate a few key points that often get overlooked… Reference is king: Especially with new artists, there is a tendency to design and build spaces completely “from imagination”. The danger here is that artwork built without real-world reference often lacks an innate “believability”. Even for fantastical environments, using real-world reference is absolutely key. Keep a repository of relevant visual resources, and refer to them constantly during production. Using reference to inform decisions about texture, scale, atmosphere, and construction details is crucial to building spaces that will resonate with players and ultimately be more immersive. Stay loose and flexible: When building in-game spaces, try to keep them as malleable as possible. This can be achieved in several ways, most notably things like modular-construction and easily-editable tools (such spline-deformation, or other procedural automation). I say keep it flexible because often times during development, it becomes necessary for environment artists to make broad, sweeping changes to a game space (often in response to gameplay needs). Nothing is worse than having to start from scratch, so anticipating the need for malleable design helps alleviate some of that pain if (when) a large change is required. Don’t be sloppy: This one may seem silly, but honestly the biggest mistakes I see made by environment artists (and sometimes even in shipped games) are simple things than can be easily avoided. Have some minor UV stretching on an asset? Missing some polys on the back side of a boring prop? Mesh smoothing acting a little funky? Fix it NOW. Be your own harshest critic. Don’t wait for that small, sloppy mistake to get propagated all over the map, only to be reported by QA weeks down the line, forcing you to go back into a source asset and waste time resolving trivial matters. Precision matters. Having a well-tuned critical eye that can automatically catch mistakes like these is what separates a competent 3D Artist from an exceptional one. Fancy new tools are great, but sometimes old-school is best: After you’ve been working as a game artist long enough, you start to acquire an array of techniques for tacking different problems. Just because a technique has been used for a long time doesn’t make it obsolete or even remotely inferior. Sure, Bitmap2Material can tile your textures for you with the press of a button, but nine times out of ten, I can do a much better job just using Photoshop, with only minimally more effort. Same goes for modeling… Zbrush is unparalleled in how it’s changed the world of 3D sculpting, but sometimes you gotta get in there in Max manually weld some vertices, just like the pilgrims did. Don’t be afraid to skip the “easy” button and do things manually once in a while… especially if it means you can save time and be more surgical about your artistic intent.
Obduction’s Artistic Direction
Obduction is a curious project from an Art Direction standpoint, because although it is in no way connected to Myst series, it is very much intended to follow in the footsteps of those games – especially as far as the player’s experience of exploration and discovery goes. As such, we didn’t want to stray too far from the overall visual style of Myst and Riven, which I guess I would describe and “surreal photo-realism”. It means I didn’t want to push the stylization too far from photo-reference (say, like The Witness did with it’s heavy simplification and visual deconstruction)… and we are instead relying heavily on photo-sourced texturing and not shying as much away from surface detail. Of course we need our sweeping vistas and dramatic reveals, but I think the single biggest recurring theme in Obduction is visual juxtaposition or otherwise disparate elements. The main world of the game is heavily designed around this concept, with elements from various different origins (and even time periods) being blended together in the same space. This theme is further reinforced by the central narrative, which is all about parts of multiple worlds being thrown together in the same physical space, and how those different elements contrast each other in interesting ways… Again, not too unlike the original Myst aesthetic.
All of the world designs for Obduction actually began life as top-down maps handed to us from the Design team (led by company founder Rand Miller). Once we began mocking them up in 3D, we did so directly inside Unreal 4, using a toolset we added called “Slabs”. This very simple prototyping method is actually something I became absolutely addicted to when using Jonathan Blow’s custom engine for The Witness, and it was one of the first things our programmers implemented in Unreal. It’s essentially just a way to re-size 6-sided cubes by dragging them one face at a time… rapidly authoring complex forms out of basic volumes, all directly inside the Unreal Editor. As the game worlds got more and more refined, we started replacing the slabs with actual geometry assets, but it allowed us to flesh out the entire scope of the game immeasurably faster than if we’d tried to mock it up with custom meshes.
In fact, our runtime collision geometry is mostly still constructed from a system of slabs. It is extremely lightweight in terms of polycount (making for very cheap collision calculations) and its flexible editing allows for easy adjustment and bug-fixing of collision issues.
Making Environments Manageable
When constructing large, complex worlds like these, questions arise like “how do we teach the player what they can and cannot interact with?” or “how do we convey to the player where they can and cannot walk?”. Managing a player’s experience inside a complex environment is a constant concern when building a game space, but usually the solution lies in simply being consistent in your visual language.
The navigation question is literally a matter of ensuring that if the artwork looks like a path should be navigable to the player, then let them navigate it…. and vice versa… if a path is not navigable, intentionally author the artwork to make that obvious. Try not to allow ambiguous navigational situations into the game. It sounds trivial, but often requires iteration and tuning.
Managing a player’s experience inside a complex environment is a constant concern when building a game space.
Player interaction issues are a bit trickier, and can very drastically by game. The Witness is a good example of very clear-cut player interaction, since the only things a player ever required to interact with are puzzle-panels with a very concise visual language. Obduction is trickier, since players can interact with in-game objects – so it’s more about teaching the player what kinds of things they are allowed to interact with, and then trying to author the non-interactive scenery in a way that conveys that… and (hopefully) avoids red-herring or pixel-hunt situations.
Unreal Engine 4
Unreal Engine 4 has been a wonderful development environment. It’s also changed my job responsibilities dramatically. In addition to my role as Cyan’s Art Director, I’ve essentially become the de-facto Technical Artist on the team. Throughout my career, I’ve worked alongside many talented graphics and tools programmers, and learned quite a lot about the techniques and concepts they employ. And yet, I am no programmer. But UE4 is unique in that is has allowed me to leverage my understanding of graphics techniques and write my own tools (using Blueprints and the node-based Material Editor), which I can then give to my team to use. While it may seem odd to have a guy with a mostly-art background building tools and authoring complex shader pipelines, I’ve found it to be extremely liberating and empowering. It’s given me a much deeper understanding of the true cost of game artwork, and has made me an exponentially more capable production artist. I even recently game a talk at GDC 2016 entitled “Building Obduction: Cyan’s Custom UE4 Art Tools” which outlines a few of the unique techniques we have employed to both speed up and expand Unreal’s built-in capabilities. I enjoyed the process of breaking down and explaining my techniques, and I hope to do more of that via my website (in the form of blog entries and video tutorials), once the production crunch on Obduction settles down a bit.
Lighting and Color
Lighting and color are absolutely fundamental to good environment design. I would go so far as to say, maybe the most important elements overall. In my day-to-day, one of the most common points of feedback I give is for artists to reduce the visual noise of the texturing… to mellow out the albedos of the materials so that the lighting can come in and make a scene feel cohesive. I think there is a tendency for artists to get lost in the minutiae of a scene, and put too much emphasis on textural information. I prefer to instead focus on overall textural cohesion, and let the form and lighting do most of the work. The Witness is an extreme example of this (where modeled form and good lighting is doing almost ALL of the work), but I think it’s valuable for artists working in any art style to not rely too heavily on textural detail.
I think there is a tendency for artists to get lost in the minutiae of a scene, and put too much emphasis on textural information.
Lighting is the glue that ties a scene together, and allows the viewer’s brain to process all of the elements together, as a cohesive whole. And color is a way to help guide the player; to focus their interest where the design dictates.
Telling a Story
Environmental storytelling is definitely the cornerstone of Cyan’s games. There is no one secret to developing environments that help convey narrative to the player (and a lot of mainstream non-adventure games are doing great work in that department these days). But it helps to have a solid understanding of the story as early in the worldbuilding process as possible – especially the stories of the character that lived within those spaces. By having a good knowledge of the characters, it empowers the artists to add personal touches to the space that are more than just “art that looks cool”, but instead help to fill in narrative details that would otherwise go untold.
It’s also empowering to let the player fill in non-crucial story elements themselves. By discovering “hidden story” as part of the game world, rather than through exposition (from an NPC or journal entry), it lets the player fill the role of detective and archaeologist, adding nuance to the gameplay, even if not every player will want to delve that deep.