Thursday, August 24, 2006

Future proofing game graphics

A source of great sadness for me is the loss of so much great game artwork due to the constant erosion of obsolescence. Thousands of pieces of artwork are produced by highly talented teams, crammed into the latest title, launched into the marketplace and then, forgotten. The game eventually stops selling. The artwork was encoded in an obscure custom format that no one understands any more. Other than the occasional archived screenshot and some fuzzy memories, all those years worth of work are gone.

Permanence is difficult to achieve. Still, many artists hope that their work remains useful, pertinent, and perhaps even profitable for at least their lifetime. With games, we are lucky to get shelf life of 12 months. There is obvious room for improvement.

So here is today’s navel gazing challenge: What would graphics that last for the next 10 years look like?

Fine art tiles
I don’t really have a final answer, but here is a brainstorm about a theoretical set of graphics that might hang on for a decade. It is by no means the only answer to the longevity challenge, but it has some useful characteristics to help game graphics age more gracefully.

Imagine a set of graphics with the following characteristics:
  • 2D background tiles
  • Standard format (PNG)
  • Print resolution (300 dpi)
  • Professional quality
  • Painterly, non-photorealistic style.
  • Brand potential
The general strategy
The general survival strategy is to create graphics that can be used in as many titles as possible for as long as possible. The longer lived the titles the better. The more re-releases and sequels they spawn, the better (so long as they use the same set of graphics!)

Games are currently ephemeral. Anything highly algorithmic in nature (that requires more than the human brain to process) tends to be quite fragile. If there is a change it the technological ecosystem that supports its operation, a game has little ability to adapt. When the next console shift happens, many of the games are swept away. When the next version of the OS appears, or you install new video drivers, your games begin to die off. A single game, by itself, is not a reliable canvas for long lived art.

By giving developers a high quality, low cost option for filling a critical yet difficult part of their development process, the our graphics become an obvious choice for prototypes and smaller scale commercial projects. A hundred games released over the next decade should keep the graphics in the public eye for much longer than graphics released for a single title. Developers get cheap graphics. The graphics get a slightly longer time in the spot light.

2D vs. 3D: The importance of standards
One immediate question that popped to mind is whether 2D or 3D was a better format for longevity. 3D has some obviously attractive features. It is flexible enough to be used for a wide variety of games. Most games are going 3D now and going with 2D seems like it would cut off 95% of all future games from using the artwork.

Unfortunately, 3D models are still in a bit of a flux at the moment. There is no agreed upon file format and pixel shaders are wreaking havoc on the concept of ‘standardized’ texture mapping. A 3D model created today is highly likely to be outdated in five years. It is often easier to just remodel it from scratch to suite the newest technology.

2D has the benefits of being a mature format with strong standards. It is highly likely that bitmapped 32-bit graphics will be around in the next decade. The medium has seen little change in the last five years and there are few competitive threats on the horizon. PNG is a solid standard that is on the upswing of gaining broader support. I’d bet there is a 99% chance you’ll still be able view and read PNG files using commonly available software a decade from now.

Resolution: The importance of medium maturity
The bugaboo that plagues most technical artwork is that the bar keeps being raised. 4-bit graphics gave way to 8-bit graphics, which in turn were supplanted by 24-bit graphics. Resolution also increased over time and looks to keep increasing in the future.

For 2D graphics, there are limits. Beyond 200 to 300 DPI, there really isn’t much point in having more resolution. For softer, illustration-style graphics, the human eye tends to start blurring all those pixels together. You also don’t really need much beyond 24-bit color unless you are doing some crazy photo manipulation. Graphics created at such levels will remain useful for the foreseeable future. Provided that you are a competent artist, your graphics will look just as sexy in a decade as they do today, technology be damned.

Photorealism is an ever moving target and there will always be someone who does it better. Cartoons go in and out of style rather quickly. If the style is too unique, it is unlikely that it will be attractive to a large number of developers.

High quality illustration can hold its own for multiple decades and is generic enough to appeal to a multitude of developers. I’m making the bet that 2D tiled graphics is a style that will have pertinence in the future alongside the inevitable new styles. Ideally, designers will choose to use 2D graphics not only because they are limited by current technology or budget, but because it is the best fit for their game.

2D static graphics are relatively inexpensive to create and use.

Expenses increase as you add dimensions: The rough rule of thumb is that for every dimension that you add, the cost of production skyrockets and the number of skilled producers decreases by 10.

If you move from 2D to 3D, expect costs to rise and talent to become scarce. If you move from static 2D to animated 2D, the same thing happens. Animated 3D graphics are guaranteed to bloat your budget and leave you grasping for skilled artists.

In keeping with graphics for the masses strategy, it makes sense to aim low. That ensures that there are always hungry new teams emerging from the quantum indie vacuum, popping into existence with no money and no resources. Our inexpensive, highly cost effective graphics will be waiting.

2D tools are inexpensive and experts are easy to find. The fact that you can batch recolorize a set of graphics and have a whole new level is hard to beat. By reducing the cost of adaptation, we encourage reuse.

Brand potential
Over time, these future proofed graphics will go one of two ways. If their quality is low and the players have poor experiences with them, subsequent titles will likely be seen as shoddy. Reviewers will mark down the projects for taking the cheap route and not investing in original art resources. How often have you heard the refrain “They just reused the same graphics from before! Score Deduction!”

On the other hand, if the quality of the graphics is high and players have great experiences with titles that use them, then the graphics have the possibility of creating a brand of their own. This has happened in the past with the Wilhelm scream and Space invader graphics.

Creating graphics with brand potential is a tricky feat to pull off. A history of positive player experience with the graphics is the critical ingredient. An interesting story about the graphics that appeals to educated gamers is also helpful. All of these naturally occurring factors can be augmented by a steady and effective awareness campaign.

Adaptability to new technology
Naturally, there are some fundamental technology advances that prevent our graphics from being suited for every project. We can still stack the deck in our favor, however.
  • Interactivity
  • 3D

Interactivity: The interactive elements in games are evolving at a rapid pace and their art resources bear the burden of also being interactive. A fighting game, for example, would be nothing without graphics specifically tailored to demonstrate the interactive aspects of hitting, moving and blocking. In fact there are wide swaths of the graphics spectrum that must be custom tailored to fit the interactive system of the game title.

This is why our graphics don’t really deal with characters, special effects or other areas that demand high interactivity. Instead, they focus more on background props and landscape tiles. These more static elements are less likely to demand custom created graphics that are highly tailored towards a game’s specific interactive requirements.

The rise of 3D: Over time 3D will only become more attractive. Standards will start to emerge which will make 3D assets easier to repurpose and many productions will demand that you use 3D to cut the cost of character animation. We can’t stop this trend, but we can make graphics that can still be useful within a 3D engine.

To this end, I’m focusing on faux 3D environment tiles. You can use them along side 3D characters, particle effects and such without too many difficulties. If someone insists on using a 3D engine and their title doesn’t need to move the camera, our proposed graphics remain at least a viable option for inexpensive backgrounds.

Broadening the scope of the discussion
So you’ve just read through an elaborate thought experiment. It sounds quite silly on the surface…“Making game graphics that last a decade.” Pshaw!

But ultimately, I’m asking some simple questions:
  • How can we create game artwork that remains valuable to people for a long period of time?
  • And looking at the larger picture, how do we avoid creating disposable content?
I began exploring future proofing art because it is an area where I have experience and can affect change. We can extend the question to other aspects of game development. What would a future proofed sound track look like? What would future proofed level design look like? What would an entire future proofed game look like?

Some lessons from future proofing graphics are likely applicable to games:
  • Broad usage: If a game concept is used across a broad number of titles across a wide population of users, it is likely to last. Think of this as a portfolio management matter. Where a single high risk project might easily go under, the chance of all projects going under is much slimmer.
  • Standardization: Standards in a mature medium help ensure the persistence of a game concept by facilitating reuse. A standard is simply a method of crystallizing value in a broadly accepted and reusable format.
  • Reduced Cost of Adaptation: As costs of updating and adapting a game decrease there is a greater chance the content will be brought forward as the technology ecosystem evolves. When people are looking for entertainment, they have lots of choices. If you can provide equivalent utility for less money, people have an economic incentive to reuse your work.
  • Adaptability to upcoming technologies: We can look down the road a few years and make a guess about what is coming next. By focusing on more stable areas and having upgrade paths in mind, our content can help make the transition when technological shifts occur.
  • Brand potential: A game concept with a great brand remains meainingful in the face of advancing technology and player burnout. Long lived art forms a deeper emotional connection with the audience that keeps them coming back long after the utilitarian value has faded. Brand turns a throw away experience into an evergreen experience.
There are many classic examples of games that have withstood the test of time admirably. Chess, Solitaire, Tetris are the easy examples. Many Nintendo titles successfully reinvent themselves for new audiences, generation after generation. Lastly, successful game designs such as Dune 2, or Doom manage to live on by founding entirely new genres of game play. Each of these has some or all of the elements mentioned above.

Future proofing art is not an exercise in preservation. Future proofing is an exercise in building in easily accessible value that can be reused and repurposed.

We think of a painting as a static smear of paint on a physical canvas. But if you look at its use and value throughout time, you'll find that it evolves quite radically. A great painting goes from being on a canvas, to gracing a living room, to being a work of art in a museum. At each stage the value of the painting to its owner is distinctly different. The painter sees it as a creative act that will bring money. The original buyer sees it as something to brighten up the living room. The museum sees it as a work of cultural expression that will enlighten the masses that view it. At each stage in its lifecycle, the original form is reused, reinterpreted and reapplied to a new environment that can be value to others.

Creating long term works of game art is thus about creating content that lends itself to this constant process of adaptation.
  • First, we must acknowledge that many aspects of modern game development are fragile. A delicate ecosystem of art and technology is the only thing that allows us to share our works of creativity. This momentary eddy in the cultural current slips away when time inevitably flows forward.
  • Next, we must consider how our art or design can take on new life beyond this moment and this release. What choices can we make that facilitate future audiences assuming control of our content and adapting it to their unique needs and environment? .
Wouldn’t it be neat if the game art you made today could be enjoyed by your children or grandchildren? As a creative fellow, this is a rather delightful dream.

Take care