Amazing art director and talented artist Keith Self-Ballard talked about his career and the function of great art in video games!
80.lv is happy to open 2017 with an interview with Keith Self-Ballard. Keith is someone a journalist like me would call a ‘veteran’. He’s been doing games longer, than the majority of our readers were able to play them. He contributed to a Myst III, worked on Lords of EverQuest, Saints Row 2 and is currently working on an upcoming Volition title called Agents of Mayhem (although he can’t show anything from the game for now). Keith also worked at Blizzard for a long time, contributing to the organization of great training courses for the company’s army of artists. This is a person, who really helped to shape the modern approach to art in games.
Wow. Candidly, I’m worried that talking about my career might be boring for most of your readers. That said, I can talk about my history as well as the difference between being an artist when I started in the industry (1998) and an art director today.
I started in the industry creating, well, shovel-ware. My first break in the industry was working on a series of outdoor sportsman games. They were wildly popular in the late 90’s and included the original Deer Hunter series. Working on those games was very much a “trial by fire” experience for someone who was fresh out of graduate school and had no experience making games at all. In fact, the first game I ever worked on was a cheap knockoff of Minesweeper called LandMine. I had approximately 2 weeks to finish all of the art, which consisted of something like 200+ 8×8 pixel images plus some front end UI graphics. I was working with one offsite programmer (who had never played Minesweeper), and I was beta-testing the game while I created the art.
The breadth of content creation I experienced, as well as the pace, are difficult to properly convey. Suffice to say, my title then was “artist” and an artist was expected to be able to do just about everything. I created art for every possible aspect of those games. This was also during a time where you were limited in color palettes and had to manually adjust your color ranges in Debabelizer. We also only had 1-bit alpha, and had to ensure that the alpha color was consistently that one color and located in the same spot for each image palette. I’m shaking my head as I think back on it.
Anyway, in a little less than 2 years, I had shipped approximately a dozen of these games. The longest development cycle I had experienced was close to 3 months. I had spent many weekends sleeping under my desk, but had also collected a long list of completed games. The problem was that none of them were going to earn any accolades for their visuals. Two years into the industry and my portfolio wasn’t appreciably better than the one I had from school. I had created a lot of art, but none of it was particularly good, and there were only so many companies interested in someone who could model and animate deer.
Luckily, I had become fairly active in the mod community during this time. Polycount had become a popular site for Quake 2 modders. I tried my hand at some Q2 characters, but really found my footing in the Half-Life modding community. I released a handful of different characters, all of which pushed me to improve both my modeling and texturing skills. Ultimately, it was all of this practice on the side that led me to my next job and what I feel was one of the most significant boosts to my career.
I had friends from school who were living and working in San Diego, and I found a job posting (I think on Gamasutra) for a texture artist position at Presto Studios. I wasn’t familiar with their projects, but I did some research and it sounded like an interesting place to both work and live. I flew out there and interviewed all day. Behind closed doors, they revealed to me that they were working on the next entry in the Myst franchise, what would eventually be released as Myst III: Exile. That was one of those career moments where you kind of see yourself from the outside. The clouds part. Beams of golden sunlight shine down. And a booming voice commands, “LEVEL UP!” I hope your readers understand that, at that time, Myst was still regarded as a seminal franchise in PC gaming, and Riven was one of the high water marks that we looked to in art school. Now I was interviewing for the opportunity to work on the sequel. I was gobsmacked. Stunned is too pedestrian a term. Also, that day was also the only time in my career where I’ve interviewed and received a job offer on the same day.
I worked on Myst III: Exile for almost 2 years. It was simultaneously one of the most exhilarating and also one of the most grueling projects I’ve worked on in my career. Presto was a small studio by modern standards, but Myst was a big franchise and no one wanted to release something undeserving of the pedigree. I originally started out as a texture artist, but my responsibilities would grow over the months to come. Over the duration of the project, I did lighting work, some environment animation, as well as render management and compositing. Our render times were long and our computers were only as powerful as those that could be purchased in early 2000s. The hours were long. I remember many 80 and 90 hour work weeks. They were rough, but I also remember looking at the clock as the sun went down and thinking, “Oh good, I’ve still got five or six more solid hours before I need to go home.”
I often think back on those times when debates about crunch come up. To be clear, I don’t refer to the hours I worked as a point of accomplishment. I’m also not referencing those times as a matter of personal pride or to deflect some of the horrible crunch periods other developers have had or continue to experience. I simply call attention to it because I remember being tired, but I also remember being passionate about the project. I was proud of my work, and I desperately wanted to craft something good. Also, I was a young married man without kids. Those were sacrifices I chose to make as opposed to being in an environment into which I found myself unwillingly thrust. I wouldn’t make those same choices as an older man. Those were brutal days. But Presto was a strong family and we fought our way through that project together.
Fast forward several years and I was hired by (THQ) Volition, Inc. in Champaign, IL as a Senior Environment Artist. My wife’s family was from the Midwest and I was now a slightly older man, now with a kid, and the move made sense. At this time, Volition saw the possibility in the exploding “open world” game market and Volition was in pre-production on their first entry in this genre. The internal name was “Bling Bling” and would eventually come to be known as Saints Row. I was the first dedicated environment artist brought onto the project. Those were interesting times. Nobody really understood what was required to make an open-world game. Through sheer brute force, we figured it out. We didn’t have a world editor, so everything was hand-crafted in 3DS MAX. Over the course of the project, I found myself taking on more of a technical leadership role than a visual one. This wasn’t so much by choice as it was necessity. Since we didn’t have a world editor, technical elements like navmesh and streaming volumes had to be hand-crafted instead of automatically generated. In addition, we also had to hand-rig the props and structures with individual LOD settings, hand-place occluder volumes and hand-place the vehicle and pedestrian pathing nodes.
The studio also grew tremendously during this time period. As such, I spent more time mentoring new hires, troubleshooting technical problems, and working with dedicated tech artists to refine our production pipeline (as much as we could then). In the end, we shipped a successful title, but not without a tremendous amount of pain endured by the development team – long hours of mandated crunch and poor tools combined to take their toll on the world-building staff.
I was a Lead Environment Artist on Saints Row 2. The tools were a little better, but we made the decision to reuse-and-revise as much of the previous game’s setting as possible. We were roughly halfway through production when the studio approached me about applying for the position of Studio Artist Manager. While I was hesitant to step away from creative development, the opportunity to have broader impact on the art culture of the studio was an intriguing opportunity.
For approximately four years, I invested my time in developing artist career paths, encouraging feedback, publishing quarterly art newsletters (to encourage art visibility, engagement, and collaboration), in addition to staffing the studio with high-caliber applicants to help build the next generation of games for the studio. It was during this time that I became more actively involved in GDC, moderating roundtables on Art Leadership and Art Culture. During my final year with Volition, I took on the responsibilities of Studio Art Director. This was during the development of Saints Row the Third and Red Faction Armageddon. While I had little direct involvement in the projects themselves, it was incredibly fulfilling to observe both projects’ development and offer guidance when and where I could.
However, I had also grown to miss being immersed in the creative churn of development. My perspective and responsibilities had grown immeasurably during my tenure as a studio manager and director, but I was worried that I had grown too disconnected from the development process itself. It was at this time that Blizzard Entertainment approached me about joining their secret “Next-Gen MMO.” Blizzard was simply too great of an opportunity to pass up, and I eagerly joined the team, despite having no idea what they were working on until I arrived on site.
For two years, almost to the day, I worked alongside a truly amazing team. Most of this team eventually went on to ship the wildly successful Overwatch. It was an honor to work with them on Titan, despite the fact that it was cancelled. During my time on Titan, I learned a great deal about working with an international team as well as some of the finer points of developing strong production plans. However, I was even more impressed by what happened when Titan was cancelled.
At most other studios, one would find themselves laid off in the wake of a project cancellation. Blizzard, however, went to great lengths to find placement for almost all of the developers displaced by the project cancellation. In my case, I shifted over to join Blizzard Academy, the studio’s internal training department for development. I was brought on as a Special Project Manager to help plan and develop a new hire apprenticeship program for Dungeon Building for the World of Warcraft team. It was a challenging project, but ultimately successful and resulted in almost every candidate being hired onto Blizzard full-time afterwards. I stayed on with Blizzard Academy for a year and a half. During that time, I also helped organize classes in puppetry, improv, as well as creative brainstorming. We played host to numerous speakers and filmmakers. I even had the opportunity to teach a handful of classes on Photoshop and ZBrush during my time.
Once again, I found myself missing direct involvement in development. My training experience had proven invaluable, but I wanted to be involved in art leadership on a team once more. I interviewed with several different studios; however, I ultimately chose to come back to Volition. They were working on a new IP, and I had a long history with the staff there. For those keeping score, I had now bounced back and forth between the Midwest and southern California four times over the course of my career.
Since November 2011, we’ve been working diligently on Agents of Mayhem. This is a new open world IP for the studio, but it retains a lot of the spirit of our past games. My role on the project has been Environment Art Director. As such, I am tasked with directing and maintaining the look of the world itself, providing feedback at a high level (especially on concept art), and supporting the environment art leads. In addition, a great deal of my time is spent collaborating and coordinating efforts with my counterparts in Production, Design and Programming to ensure that expectations and goals are in alignment. I’m proud of the game we’re developing, and I’m eager to see it released to the public in 2017.
Lastly, I’ve also been collaborating with Andrew Maximov at Naughty Dog on the annual Art Direction Bootcamp at GDC. In 2017, we’ll be organizing our third annual event. Once again, we’ve collected a variety of talents on stage for a full day of talks on art direction, creativity, and leadership. I hope some of your readers will join us.
As an art director how do you manage to keep all those different artistic visions under control? How do you make a game which actually looks like it has one vision and not a hundred different visions?
I’ll let you know when I figure that out for myself.
Seriously, in both my experience and having discussed this with other art directors, there is no certain and infallible way to control the vision of the project. In fact, I feel that “control” is a potentially misleading term. The truth is that most games are, in fact, the culmination of a large number of individual visions. The art director’s primary responsibility to establish the core of said vision — candidly, an idea easier expressed than executed. An art director can’t afford to get lost directing the fine details, even though we often do get distracted by them. Rather, the art director describes the “bounding box” and they know and expect every member of their team to contribute to the vision in his or her own way. From there, the art director can focus on maintaining the big picture and culling the outliers — those elements that don’t “fit.”
As for the how, there are many ways. The problem I most often here cited from art directors is that they use the tool that is the most useful to them rather than the tools which are most useful to the team. This can be quite challenging, but it does highlight that the way in which art direction is established must flex to the needs of the project first, the team second and the art director third. Sounds counterintuitive, and it is. I still struggle to remember this day-to-day. More specific examples of how to establish the vision include vision walls, art pillars, color scripts, ref boards and/or a mountain of concept art if you can get it. These are all tools available to the art director.
For me, my opinion is that the most critical element is a dedicated group of Leads who are aligned with the Art Director. The Art Director has to evangelize the vision to the Leads and make sure they understand what and, more importantly, why. From there, it’s just about managing the constant and inevitable changes over a long development cycle.
How do you usually approach the production of high quality look for the game? What are the main important elements to get a cohesive and productive vision of the project?
I rely heavily on research. The starting point for any research is really trying to understand the narrative elements and the tone of the product. From there, I prefer to branch off in lots of different directions. I like collecting information and images. I actually use Pinterest to collect my images and inspirations. Anything visual can, and often does, help me in my search. I’ll look to movies or cinematic styles that possess a similar tone. I’ll browse various online art sites to collect images that resonate with whatever I’m currently thinking. Comic books. Television shows. Classical paintings and illustrations. Essentially, I go into consume mode and just hoard as much as I can.
That’s the “top of the hourglass” for me. I’m pouring lots of content in. The next stage is the “neck.” Now, I have to filter. That can often be the challenging point for me. Any art director has to constantly ask themselves why something works or why it doesn’t. If I were to show this image to someone else, would they get it? Despite how hard the filtering process is, the more I remove the clearer the picture can become. I should be clear on this point: it’s not a good idea to do all of this in total isolation. I show the images to other artists, designers and writers. The images I’ve selected might resonate with me, but might form a very different picture from what the Creative Director had in mind. This is a good exercise for seeking alignment.
Lastly, I hit the “bottom of the hourglass.” This is where you can start broadcasting and talking about the vision and the look. This is also the point at which you can start really broad concept development. If you have a selection of strong reference points, concept development can happen sooner, but of course, you’ll have to integrate those concepts into your filtering process.
Another key idea I wanted to get across is that the hourglass analogy repeats itself. Once you’ve done this for concepts, it’s time to take those concepts and prove them out in the game. Can the rendering engine achieve the look you’re targeting. If it can’t, how do you react? Is it time for new concepts? At a macro level, the funnel has to narrow, but in early development it’s important to try lots of things and then narrow down to find what works.
You’ve had quite a lot of years devoted to building worlds for games. I don’t think a lot of people really understand what it means. How do you build a virtual world? how do you keep it both interesting and functional at the same time?
I’ve consistently had an innate passion for the worlds of film and games. I remember the early Sierra Online games I played, and how they transported me into a place that was (despite the graphical limitations) far more rich and complex than the arcade games of the time.
However, it took many years of building them and being guided by other art directors before I think I really understood what world building entailed. Looking back on it, I really should have been researching what background artists in traditional animation had been doing for decades. At the time I didn’t completely understand the nuances of storytelling let alone the role that value and color played in world building. I didn’t have a traditional art or illustration background, so I came into world building (before it was even a dedicated discipline of art) more unprepared than others can today.
Today, I thankfully have the wealth of experiences accumulated over time. For me, world building starts with the story. I believe that this is the key reason why changes to story or gameplay so severely impact world buildings. If the story needs to change, so too does the world. From the story, I’ve found it best to understand the culture or the history of the world (if it has either). Worlds don’t spring into existence from nothing; companies spend millions of dollars trying to make it appear as if they didn’t. Rather, worlds are shaped by the people (culture), events (history) and the environment as well (biosphere/location).
This may seem overly complex, but I’ve found the better you understand the world the more capable you are going to be able to adapt to creative changes – and all projects change in development.
I prefer to start with the function of the world. This actually becomes more crucial the further you drift from the real world. Regardless of whether you’re crafting high-fantasy or science fiction, the viewer must be able to relate on some level to the world you’re creating. It’s that kernel of relatability that should provide a sense of function. Without it, the player is left a bit “adrift” lacking the ability to bond with the world. However, if you infuse the world with elements that are recognizable, the viewer will be more readily able to accept those aspects that are fantastical.
The fantasy layer, that’s the part that makes the world interesting. That’s also the part that you really want to promote and “play up” within the space. In a really effective environment, the fantasy layer becomes the feature or the point of interest. Also, as long as that relatable/functional layer is sturdy, then it can support a wildly abstract fantasy layer — the viewer will readily accept it. All of these elements, both the relatable and the fantastical, come together to form the ingredients of your world. Furthermore, it is the ratio in which these ingredients are used that create the world’s “flavor.”
If you’re working in big companies, how do you make the artists work on the same kind of level. People come to art from different places and their skills are obviously not even everywhere. So how do you help them to develop new skills and become better at what they do?
I’m going to break this question down into two categories: How does the company help the artist grow? And also, how does the artist grown on their own. In the best of cases, the issue of growth is a shared responsibility. However, there are many reasons why both don’t happen.
In the workplace, individual artist growth is a persistent challenge; however, positive steps can be taken if there’s effective collaboration between leadership and management. I call out each group separately, because they both have a responsibility to the success of the art and the artist.
At the leadership level, the directors and leads should take steps to ensure that junior artists are partnered with senior staff artists. Ideally, there should be clear documentation that establishes visual targets. Where such documentation is lacking, the lead should step in to ensure that junior staff is receiving regular, productive critique. In short, a lot of informal mentorship should be happening week-to-week to ensure that all work meets visual standards and that junior artists get plenty of opportunities to ask questions and learn from others.
At the management level, the project or department managers should have conversations geared towards the long-term. It is far more effective when artists define their own career goals, and the managers should look to provide options (when possible) for challenges as well as cater to consistent feedback on the quality of their development skills.
How does the artist grow themselves, without the direct guidance of leadership or management? Throughout my own career, as well as observing others, the most significant growth in individual skill happens “off the clock.” Those individuals who commit to personal projects or independent study in their free time will far exceed their peers who only focus on their craft during work hours. Personal projects, if well-conceived, both challenge the artist as well draw upon their own individual motivations. As long as one can control the scope, and that is often one of the hardest things, the artist will learn more things faster than they will in the course of their job.
The reasoning for this is simple. On the job, the artist is likely to receive similar assignments over a long period of time. Work will be assigned, more often than not, based on already proven competencies — project completion is the goal. By comparison, a personal project is self-defined. They are often designed to challenge the artist in an area where they want to learn — growth is the goal.
We’ve got a lot of students on our website, so I wanted to ask, what do people usually look for in hiring the artist? Maybe you could give a little look into what the students need to learn and how should they approach their job search in the game industry?
While I don’t want to say that there is a specific formula for world building / environment art portfolios, I absolutely have recognized that there’s a process to how I review a portfolio and what I’m looking for.
Candidly, I don’t put much stock in prop-centric portfolios. Props are easily critiqued, but they don’t tell me much about the artist’s ability. The few exceptions are where the “props” are for a very specific skill set or game feature — perhaps weapons or vehicles that might be critical to a franchise. The other exception are hero props — elements that would be considered the central set piece of an environment, rather than props that would be categorized as filler material. However, I still see artists who assemble portfolios full of filler props. Unless you have a collection of truly creative and inspiring props, a prop-centric portfolio is rarely going to generate attention or interest.
Primarily, my focus during a portfolio review is on the ability to craft a scene. What did the artist try to communicate in the scene? Are they trying to craft a story? Or are they trying to demonstrate a skill? For me, I give far more credit to the story goal rather than the skill goal; the skills should be demonstrable within the scene. From there, I evaluate the quality or success of the scene based on each level of readability:
First Read: What is the setting; can I tell where I am? Is there a story here or a clear genre? What is the mood of the scene? What is the primary feature of the environment; did the artist demonstrate good compositional choices Did the artist make color and lighting choices in support of the previous questions?
Second Read: What secondary elements in the composition support the primary feature? Also, I start evaluating the verisimilitude of the world. Is there a human presence; how is it represented? How has gravity (or the lack thereof) had an effect on the world? What kind of impact has weather or nature had on the environment?
Third Read: This is the details, and can be the most overlooked. Do the materials feel believable or, at least, artistically cohesive. Has the artist taken the time to make elements feel integrated? Are objects wedged into one another, or do they feel like they are emerging/receding? Has the artist taken the time to hide or draw attention away from the worst “seams” in the environment.
Supporting materials: Not everyone agrees on this, but I personally like to see supporting materials. I want to see the artist’s research. Show me a ref board comprised of all of the images that influenced your projects. Show me thumbnails and refined sketches. I appreciate the peek into the creative process as much as I enjoy the final work.
Why do you think visuals play such an important role in game development? Seems like games are losing their photorealistic focus, becoming more stylized, but they still work as hugely important parts of the whole game feel. What do you think is going to happen with the game art in the future?
Visuals have grown to play such an important role in response to consumer expectations. Console and PC gamers in particular have come to expect more cinematic experiences. They want their games to feel more like blockbuster films in which they get to play the lead role. At the same time, many of our consumers want freedom and more opportunities to get lost in the game world. Although level-based games are still prevalent, there is an expectation that the developer sell the appearance of a grander world beyond the boundaries of the game space. It’s no longer sufficient to give the player weapons to use and enemies to defeat. Now, the consumer wants a world, a history, a culture and a motivation to draw them through the game.
The good news is that not every developer is trying to explore the same creative space. There are many indie developers, especially on mobile platforms, who are still crafting visually impressive worlds and stories. They have made good creative choices, despite the limitations of their tech or platform, to create engaging worlds. I’ve especially appreciated those games that have pursued a retro-visual style — often because they remind me of the games I played when I was young.
As for the future, I’m terrible at prognostication. I can’t reliably predict the next big visual trend, and I wouldn’t even bother trying. For me, it’s the act of remixing or combining elements in different and unexpected ways that inspires me.
I will offer one observation based on my own years working on large-scale games. We’re rapidly approaching a crisis point where consumer appetite and production limitations are going to hit a hard wall. Consumers have demonstrated an ever-growing desire to experience larger and more complex game spaces. Thus far, developers have been able to keep up with the pace of that desire as well as improvements in technology. Some developers can afford to throw hundreds of developers onto such titles in order to achieve the desired scale and complexity. This will not be infinitely sustainable. Automation must become a growing standard for world crafting. As unnerving as that may sound to some, I see it as the most likely outcome. This will become especially critical for those games aiming for photoreal or near-photoreal environments. The technology will automate much of world creation, whereas the individual artists will be tasked with hand-crafting particular spaces as they relate to landmarks, key gameplay or cinematic moments. We’ve already seen steps taken in this direction with products like World Machine. I’ve seen some companies attempt similar goals with urban environments, but we’re not there … yet.
In contrast to automation of photoreal environments, I expect some other studios will be pushing heavier on stylized environments. They will still require technologies that allow for rapid content creation and iteration. However, they will look to shift the visual targets away from complexity in order to still achieve the consumer’s expectations of scale.
I love working in this industry, and I genuinely appreciate giving back to the community. Despite having worked in game development for so long, I still learn a great deal from others. I still make mistakes, and I still keep learning and trying to get better at what we do. Game development is not an easy profession. Sharing our paths and mentoring others is the surest path to improving our creative process together.