Thanks for sharing this useful info friv 4 school
the download links gives a 404 error...
Environment artist Noah Berry, who worked on Fallout 3, Morrowind, Oblivion, Skyrim, continues his series of articles about world building for 80.lv. In this post he talks about creating the look of the world, the choice of tools, developing the lighting systems and gives the main steps for building complex, open environments. Check out the first article in the series here.
The Look of the World
In the earlier days, visuals for new worlds began their shape taking process almost purely within the mind’s eye of the artist. With input from the director, and often from pre-existing lore, the process for establishing a visual identity begins with continually imagining how this new world should look and feel, from the combination of light, color, and various surfaces to work with. This step is a further, albeit visually oriented, iteration on defining what the user will experience in the end, as they take in and explore the world.
Based on the particulars of the project at hand, real world references of nearly any kind, photos, film stills, screenshots, color palettes, and the like, would be gathered and readily used – even as guides or inspiration for alien and otherwise otherworldly environs – to hone in on tangible bits of information for the beginning stages of world creation. Atmosphere and mood play a critical role here, and these early idea sketches go a long way to continually champion the laborious, lengthy undertaking of bringing landscapes and game worlds to life.
With more recent development cycles, much of this initial visual creation work would fall into the realm of dedicated concept artists, where freshly created, traditional 2D art is referenced to help establish the predominant tone, iterate upon specific visual features and key details, and for any stylization of the overall character with the world being made. A strong goal in defining the visuals for an open world is to strive for a tangible sense of time and place for the audience, where they are usually intended to feel immersed within the environment.
Just considering artwork, this player experienced sensation can come from a medley of many different graphically related factors working together in tandem, such as – as often the case with large exterior environments – the quality of daylight and atmosphere, down to the nature of the different surfaces and physical features of the world. Something seemingly simple as the silhouette for frequently occurring tree types can have a significant effect on the player’s perception of the world.
As an example, broad and dominant deciduous foliage density might lend a quality of feeling encapsulated, or enwombed by the environment. These feelings could instill a desire in the player to slow their pacing down a bit, giving them pause to more thoroughly explore a location, in a measured, even, relaxed fashion. Using a darker, muted color palette for this same forested environment could simulate an atmosphere with an unsettling feel, directly translating to the experience of being in that forest. Or, perhaps, using a cheerfully saturated palette, a sense of wistful enchantment and playfulness could be instilled within the explorer.
In contrast, what if these same broad-leafed trees were instead tall evergreens, having a more open, jagged, vertical profile? This same location might instead lead the player’s eye upward, towards distant landmarks visible towards the horizon, subconsciously ushering them forward within the space. Any art and visual element used within a location can affect the player’s experience in such ways, and is important that these early, broad visual strokes serve to flesh out and enrich the overarching design and player experience goals for the project.
The Biggest Tasks of the Environment Designer
As a general rule of thumb, start with the largest problems. Be they large in an X-Y area sense, or with implementing a key design feature that impacts a significant portion of the game. From there, work down iteratively and progressively into the smaller levels of detail. With landscaping, the artist will usually begin resolving the larger lay of the land – somewhere between the existing world layout in its broadest sense, to what the player will experience from their particular point of view. Larger features like cliffs, valleys, passes, and roads will begin to take shape, where at later steps the artist will add in specific detail and clutter, down to individual rocks and vegetation objects.
The idea is for each change along the way, be it a visual one, or a design related alteration, to have the most positive impact on any and all affected work that happens after, and even farther down the line. Start at the largest scales, addressing the needs there, while considering what the next step will require. It can be a tremendous amount of interrelated tangents to mentally take on and care for, but this kind of creation process navigation and oversight is very much in keeping with the core needs, and interconnected components comprising an open world environment.
As with the reality of open world games, nearly every undertaking in realizing them to completion is a massive one. With the world itself a principal focus of the project, nearly every department is affected by, and tied into it, in some fashion. Every pass and iteration performed requires a bit of a juggling act with consideration for other aspects of the game, and how the overall development process might be impacted.
For example, large landscape features might reach a certain point where they no longer artistically jell with, or support more recent design additions to the world – say, a heavily altered city location. Perhaps a nearby mountain exists at a visually pleasing height, with a nice topographical flow along the horizon when the map was much emptier at early points in the project, but now needs adjustment with regard to freshly added, contextual urban detail. Not only will the mountain’s alteration affect environment artists, where they may have painstakingly and beautifully propagated a large area within the world with props and foliage, but programming engineers are now having to help resolve performance issues with regard to new geometry and object-based occlusion sightlines, dramatically impacting the game’s rendering and performance for the player. Additionally, designers are further affected by this change, as preexisting AI navigation data sets for the areas in question will need a substantial pass, if not a complete overhaul.
Given the interconnected, kitchen sink nature of an open world, practically every alteration, addition or subtraction, will lead to this kind of constant problem solving. Here, history has shown that considerate, heavy thinking and planning up front, coupled with good, proactive, and clear communication, will lay down a stronger bedrock from a project’s outset, allowing each new layer to better weave into the world taking shape, rather than exposing flaws and gaps that lead to a foundational collapse. This intention, as well as the subsequent actions done in its spirit, with others on board in a like-minded manner, helps tremendously with every area of world building.
Understanding the Lighting
Light and color continue to be wonderful ways to soak in and exist within 3D, open world environments. From as far back as I can recall, the possibility to experience a virtual environment from all manner of weather conditions, as well as times of day, has been an immensely powerful draw and attraction, also being a strong pull for why I sought out a career in environment art to begin with. I believe sunlight and atmosphere are as much of the character of a world, as the ground beneath the player’s feet, along with all the rich detail that propagates its surface.
Even subtle intangibles such as the origination of light in relation to other landscape features – like trees and vegetation, or building structures – where it spills over and splays across physical contours, and through transmissive surfaces, adding expressive depth and oceans of mood to a scene. Something as simple as altering a key light’s direction source can impact how the environment feels, even on a subconscious level, to the player, where they can then pick up on a sense of implied time, and orientation within the larger spaces of an open world.
While beautiful and arguably important as day-to-night lighting cycles are in helping to realizing an open world environment upon a screen, they too can come with an appreciable cost with regard to the nature of their impact upon the interconnectedness of development. There are a number of good reasons many 3D games, especially level-based ones, have decided to eschew real time daylight progression, as to do so opens a large bundle of problems to resolve in order to do it well.
A beautiful example of what a game world (GTA V) can look during nightime.
With outdoor, open world spaces, the sun touches down upon nearly every surface in a given scene. As it moves and changes with strength and color, over time, in addition to any and all changing weather patterns, a massive combination of adjustable, cascading variables comes to light with which the artist must then take into consideration, across many areas of development, often just as much as with pure environment art concerns.
For example, with the player able to be, possibly, in any given spot at any time, looking in any direction, great care must be taken to ensure that the widest net is cast over the veritable sea of combinations, conditions, and contexts that any location can be seen, experienced, and played within. Day versus night, notably in outdoor spaces, is a near universal range of lighting extremes that needs to carefully balance aesthetics with user comfort, and ease of play. Here, the proverbial hair between believable realism, immersion, and player accessibility can seem to split infinitely, and where the most painstaking, belabored efforts, often with iteration upon iteration, would sometimes only serve to just get closer to a more completely satisfying result, rather than fully arriving at it.
In many scenarios, what looked and felt like nighttime in a scene, would hinder navigation, or even frustrate enjoyment for those less sensitive to subtleties of visible light ranges, compressed down onto an electronic display device. Yet, other players might view that very same compensation and feel frustrated for how their attention was called to a visually abstracted nighttime look and feel that went against the desire for a deeper cloak of night.
In bittersweet contrast to the pure joy of working with light and color, and with emulating such a wonderful aspect of real life – The Sun, sky, and Earth’s atmosphere as they are more humbly represented in gaming – these were often tough hills to climb, while still striving for the stars. In the end, time and experience proves that entrusting in your senses, where instinctual feelings and direction arise during artistic creation and refinement, will lead you more truly to your intended goal.
Tools of an Environment Artist
When creating intricate, sprawling environments, such as those found in open world games, great tools are invaluable resources to have at your disposal. When given good thought behind their purpose and design, they can become something close to a seamless extension of the user’s will, allowing what they envision within their mind’s eye to travel through one’s extremities, and onto the screen as purely and plainly as possible. The projects I’ve have the privilege to work on were benefited by excellent tools and interfaces, that facilitated nearly all that was envisioned and desired for the worlds and landscapes created. In fact, more often, it was time and the multitude of areas to address, or details to give attention to that were the biggest constraints to work within, or the dominant hurdles to overcome.
In many cases there were brand new features to undertake, where no discernable approaches and solutions were available locally, sometimes out even in the larger, collective world of software development. In response, we would forge ahead, solving from A to C, where B was unknown, with in-house utilities. Though that process was usually more efficient and direct, sometimes a particular need – especially with landscape and exterior environments – called to attention a deficiency that had me searching outside of our engineers and tool base.
An example of World Machine usage.
One such program proved to be a boon time and again, the procedural noise and terrain generation software, World Machine. One of my favorite aspects of WM involved its node-based, logic flow controls, where the reconfigurable graph structure allows various devices to be interconnected in multitudes of ways. Using the program, I was always reminded of guitar effect pedals, involving manipulation of cascaded signal flow, send and return loops, and the like, resulting in vibrant and diverse sound palettes, experimentation, and play. World Machine is well suited for extensive, large scaled generation and manipulation for all manner of procedural noise functions, which I used not only to generate and augment in-game terrain, but for a variety of texture-based effects, ranging from rendering FX, sky/cloud art, to 2D, in-game UI art for interface maps.
Often it was a joy to experiment with and refine the working toolsets, where the goals were always to facilitate the user as directly as possible in envisioning and conceiving the worlds created. The best results were tools that seemed to minimize any delay, especially with regard to input, and then subsequent feedback for the user, as the software slowly and steadily progressed over the years, striving for the world building tools to approach WYSIWYG, 1:1 working results matching the final visuals and experience that the player would have.
In this way, it was also beneficial to let the computers do what they do best, where classically rooted interfaces and control schemes allowed for the cleanest, most precise, and immediate ways to interact with the worlds being built. Great software performs its purpose purely and singularly, without interface bloat (with feature and style) or functional obfuscation, and will otherwise deftly step out of the way as to not hinder the user–also again, back to the idea that a strong foundation allows for subsequent features and functionality to layer in and operate from a healthily stable base, on up through to the latest upgrades and additions.