words Alex Wiltshire
In gamespace, normal rules of architecture don't apply.
Impossible structures made of any material you can think of can be placed in infinite space, with their only constraint being the direction of their occupants' thoughts, emotion and movement. "They offer an extreme version of functionalism - they're an example of form following function almost purely," says Sam Jacob, architect at FAT and keen game player.
As computers become ever more powerful, games are becoming increasingly richly detailed and immersive, bound only by the imagination of their designers. "They're part narrative, part function, a combination of architectural ideas," says Jacob. But how are these places designed, and what does the technological promise of creative perfection mean for the future of videogames?
The aim of every gamespace is to provide for and enhance gameplay with engaging and believable environments. The design must take into account how players will use, interact with and move through the space, and how they will approach and depart from it. A successful design will subtly control the player's behaviour and evoke emotion, using an array of devices, such as architecture, lighting and camera work. It will also suspend the player's disbelief by providing a consistent set of physical and behavioural rules that govern the way the game environment and its inhabitants act. These details enhance the player's experience, from the way that Mario leaves footprints in wet sand in Super Mario Sunshine (2002), to how shooting a light bulb in Splinter Cell (2002) shrouds a room in darkness and prevents the player from being detected by enemy guards, or how cutting the rope holding a bridge up in ICO (2001) allows it to fall and the player to progress. True immersion - with a totally believable environment, rich and consistent in every detail and reaction to a player's decision - is the holy grail of games design.
Technology is a vital factor in creating believable environments, as it defines the level of possible detail in a gamespace. Greater processing power means that designers are able to make more complex models for each element in a scene, and make more of them. Blocky, jerky and largely featureless environments with fog that restricts the depth of field to reduce pressure on the processor, like those found in the first Tomb Raider games, are being superseded by more rich and organic-feeling spaces. Players also come across more incidental details that can effectively characterise these spaces, like chairs in an office block or chalices in a cathedral.
More processing power also enables more visual effects. Lighting effects like bump mapping (simulating relief on a surface with shadow and highlight) and accurate shadow casting are very successful in making a space atmospheric, such as in Silent Hill 3 (2003) and Halo (2002). Environment mapping (creating reflections on surfaces) and specular lighting (creating highlights on a surface) give form and texture, and differentiate between surface materials on different elements in a scene, like the bonnet of a sports car in Gran Turismo 3 A-Spec (2001) or a shiny balloon in Wario World (2003).
But processing power is still profoundly limited, so the designers of a game have to decide on effects, the level of complexity and the number of elements to be used in each space in order to create the right effect for the game they are making. But as Viktor Antonov, art director of the upcoming first-person shooter Half-Life 2, says: "While better technology gives us the tools to make any story or experience more real and with more sophistication and nuances than in the past, it still does not reduce the burden on the designer to make the experience original and compelling." Technology is by no means the only influence in the creation of a gamespace. Designers consider two other sets of choices when creating game environments: how to make a space that provides good gameplay, and how to present the gamespace aesthetically in order to create atmosphere and a sense of place. Those choices that directly concern gameplay are probably more fundamental to a game environment. They define how the environment is constructed so gameplay is directed, and use many established architectural principles, such as balance, scale, unity, proportion and rhythm.
Before looking at these techniques, we should look at the principles on which a videogame level is composed. Philip Campbell, a level designer for Tomb Raider and now a creative director at Electronic Arts in the US working on the James Bond franchise, conceptualises levels as "golf holes in playgrounds". Designers consider where the start point, or tee, in a level is. They must think about all the things that the player can see from that point, decide on the view distance and which hazards to show and which to hide. The goal of the level should either be shown or hinted at, with the mechanics - or, to continue the analogy, the golf clubs - to achieve the goal made clear. "Players like a directional experience, with choice over how to approach the task," he says.
A basic way of creating a sense of movement is with types of walls: long, linear walls encourage movement along them; tall, thin walls suggest movement up them; concave structures invite players inside; and convex structures encourage them to move around the building. Rhythm can be achieved with the repetition of certain structures, such as bulkheads along the length of a corridor on a space ship, which move or nudge the player forward with confidence and security. Tension can then be introduced with a sudden break in the pattern, like a collapsed strut in the corridor, that makes the pattern unpredictable. The designer can thus direct the player's mood and movement.
A problem with creating richly detailed environments in games is a resulting loss of legibility, which leads to players not noticing elements that are meant to prompt specific behaviour, such as a certain action that must be performed or the direction for progression. "Managing large amounts of visual data and controlling the player's attention is one of the biggest challenges in real-time interactive environments," says Viktor Antonov. "Our art department uses basic design principles close to those of photography, such as contrast, silhouette, grouping elements together in larger masses, the use of perspective to draw attention and define a focal point and the use of colour to suggest depth and to play with moods. Sometimes we want to overwhelm the player's senses with a large amount of visual noise, and at other times offer a moment of rest with a low-contrast area."
Philip Campbell feels that foreshadowing, or previewing events in a level, is an important strategy to directing gameplay. In the Return to Atlantis level of Tomb Raider: Unfinished Business (1998), he designed two important elements that created good gameplay and visual appearance. He made what lay ahead highly visible and made the upcoming sequence of architecture logical - players can see the exact structure through many levels of the building, allowing them to "feel clever" by being able to make intelligent decisions about the direction they take. He also placed a large window right at the start that semi-reveals the very end of the level and the last enemy, a centaur: "An added bonus was the unearthly sound he made - a sound that accompanies and unnerves the player throughout the scenario. Foreshadowing can help immeasurably in the storytelling aspect of the adventure, setting up aims and goals for the player and enriching the experience."
Denis Dyack, president of Canadian game developer Silicon Knights, puts forward another way of directing the player's experience. The creation of "flow" - Dyack's reading of immersion - is a vital part of his approach to making videogames. He describes flow as a point at which the player forgets reality and time and becomes totally absorbed in the game. "Flow comes from combining camerawork, gameplay, sound, story and so on - for Eternal Darkness we tried to build slowly by combining story and game to create a smooth experience. We made few camera cuts, so it was visually smooth-flowing - jump cuts are usually disorienting and break up the play." Eternal Darkness (2002) is a strongly narrative, atmospheric adventure game that's played from a third-person perspective - you see your character on screen at all times. The camera, then, is another aspect of the player's visual experience and an important tool in directing gameplay.
Most of the examples of design above are concerned with adventure games - environments that encourage exploration along their length, with little real possibility of exploration off the prescribed track. That the story in these games is acted out by the player progressing through the gamespace means that the game narrative is embodied by the environment itself. The lack of such a narrative in multiplayer first-person shooter games, like Quake 3 Arena (1999) and Unreal Tournament (1999), means that the principles of design behind these gamespaces are entirely different. Levels are designed to enable as much player-to-player interaction as possible. Each space in a level features multiple entry points so players can snipe, assault and escape, with the only limit to freedom being their skill. Important beneficial items are placed in certain areas to intensify action, and the most successful multiplayer levels are very simple in form, such as the classic Facing Worlds level for Unreal Tournament, which features two towers connected by two bridges that cross in the middle. The objective for the two teams is to capture their opponent's flag and return it to their own tower. Such simplicity means that the level is quickly understood by players, allowing them to formulate strategies easily.
Once planned, gamespaces must be given meaning and significance for the player - a sense of place and atmosphere - with a set of aesthetic choices. Antonov explains: "At Valve, we're trying to use a variety of devices to trigger responses in players, such as character acting and physically interactive environments, as well as through art techniques, such as colour and atmosphere. When building an environment, our production team tries to create a sense of place with real history and depth. In the finished game, all these elements and different disciplines merge to create the experience of a world that comes to life around the player."
The HP Lovecraft-like Eternal Darkness is set in several eras in "real-world" locations around the world, like Cambodia (in 1150 and 1983) and Rhode Island (in 1760, 1952 and 2000), unified by an atmosphere of otherworldly threat and conspiracy. These locations were chosen for their historical significance, such as Amiens Cathedral, which the player visits three times in different periods - in 814 with the death of Charlemagne, in 1485 during the Inquisition, and 1916 during the First World War. The cathedral was ideal because it changed over time, allowing the designers to change the environment for each visit. On the first visit, the main cathedral hadn't yet been built, so action takes place in the original chapel. The second visit establishes the cathedral, and the third depicts it as a field hospital with a distant battle audible. "The locations we chose symbolise how time can erode or change and adapt things," says Denis Dyack. "They subconsciously suggest that things aren't going to be the same, by making the player visit the same place over time."
For the aesthetic design of the environments in Eternal Darkness, Silicon Knights had to choose motifs that players would instantly comprehend. Because the game was to be distributed worldwide (it was published by the Japanese company Nintendo), motifs had to be readily understood by an extremely wide audience. "Nintendo doesn't have the Renaissance ingrained in its culture," says Dyack.
Videogames have correspondingly developed a universally understood iconography: "The videogame iconography of the moment is that of the pop clichés created by Hollywood, as well as those derived from comic books, Manga and fantasy," says Viktor Antonov. Many motifs found in games are instantly recognisable - the space ship, gothic horror mansion, ancient Egyptian tomb, dwarven blacksmith's shop and Blade Runner dystopian future city - and are becoming increasingly tired as technology resolves more complex environments. One of the reasons why Egyptian tombs and warehouses have been such a common location for games is that the squared-off corridors, flat-sided pyramids and cubic crates are easy for low-powered computers to draw.
But the search to set games in interesting and original locations will be perennial: "The gaming industry is still young and lacks confidence, but even if it were a mature platform, we would still have to work to develop a new approach to iconography that expresses something original," says Antonov. "The creation of great fiction, a good story and fresh concepts is a universal challenge, common to all forms of narrative art."
Another choice the designers have to make is over the general look of a game. An insistence on a photo-realistic style is in danger of resulting in environments that remind players of just how far from reality they are. More stylised approaches, such as that taken by Metal Gear Solid 2: Sons of Liberty (2002), allow the processing deficiencies of the hardware in resolving "perfect" forms not to be so apparent, resulting in a smooth, compelling and believable environment. ICO, using washed-out colour and a simplicity of form, has environments that perfectly create the atmospheric solitude of a deserted castle. The occasional appearance of grassy courtyards and gently swaying trees in vivid sunny greens provide restful contrasts to the barren stone of the rest of the spaces.Lighting is a powerful way of creating atmosphere. ICO's use of a colour palette that emphasises the effect of full sunlight, shade and indirect lighting makes otherwise identically decorated areas feel very different. Silent Hill 2 and 3's use of darkness and dramatic shadows creates a sense of threat from the unseen, and Splinter Cell uses shadow as an intrinsic part of the gameplay because it offers cover from detection by enemy guards. But it is still a developing technology that places heavy demands on available processing power - while the original Doom (1994) used lighting as a simple tool to create mood, the yet-to-be-released Doom 3 contains some of the most accomplished lighting effects ever seen in a videogame.
The rise in available processing power is allowing fuller and richer gamespaces and demanding ever more from gamespace designers to make sure that they are original, compelling and believable. But while a richer gamespace can enhance gameplay by making it more resonant and immersive, the games of tomorrow will not necessarily be intrinsically better to play than those of the previous generation. It's easy to dismiss the crude visual efforts of the past but it is interesting to note that basic gameplay in the most modern of games and those released 15 years ago remains very similar. For instance, the gameplay in Metroid Prime (2002) feels perfectly fresh, yet is virtually identical to gameplay in the original Metroid game released in 1985.
So in real terms what has the development of more complex and rich game environments done for videogames? Making them less abstract and more intuitively understood and believable, videogames are becoming more and more legible - and attractive - to people who aren't versed in videogame conventions. Games like The Getaway (2002), which puts the player in an accurately modelled Central London, offer an environment for play that is instantly comprehensible and broadly compelling. The ability of these environments to conjure emotion and feeling in players, other than the traditional competitive spirit to "beat the game", means that the experience is able to appeal to a wider audience. The development of these spaces is vital to the rise of prominence of games as an entertainment - and perhaps even artistic - medium.