Proposed Pipeline for Environment Pre-production to Alpha

By Forrest K. Keyes


The GDD (Game Design Document) is the first thing that is created for a project that starts to shape the defining parameters.  For this reason we will look to this as our initial source for drawing inspiration.

The first thing to do is to draw up a world map of your game’s universe, using the defining parameters given to us through the GDD.  What are the defining parameters?  That will depend on the subject of your GDD, but possible parameters will be.

Real World/Historical vs. Fictional Locations

Futuristic vs. Old World

Primitive vs. Advanced

Desolate vs. Lush

Mountainous vs. Flat

Urban vs. Rural

Fantasy vs. Sci-Fi

Realistic vs. Cartoon

Does the game’s storyline take your character all over the Galaxy, all over the world, on a single island, building, or room?  However vast or finite the geography that your character traverses throughout the game, that is what defines the “Game Universe Map”.

The previous page shows an example of what would make a good Game Universe Map. It is J.R.R. Tolkien’s map of Middle Earth from the Lord of the Rings series.  Even though this map was not intended to be a Game Universe Map and was created long before video games existed, it fulfills everything that a Game Universe Map is intended to accomplish.

  • It encompasses everywhere and only where the player goes throughout the game.
  • It delineates the major regions of the terrain and calls out notable landmarks, which is conducive to defining a variety of compelling backdrops to inspire level designs.
  • It details out the manner in which regions tie into one another, which is useful for coming up with graceful transitions between levels. Even if different Level Designers are working on levels that follow each other, congruency can still be achieved because they are all being guided by the same source of inspiration. In turn, this gives your entire game a level of continuity that makes sense to the player allowing him to have a larger sense of where he’s been and where he is going, as appose to a series of disjointed isolated locations. Example:

At the end of a level the player’s character comes upon a precipice in front of large temple ruins along a ridge overlooking a valley. He sees a small clearing below with a flickering campfire along a riverbank. Standing upon the tip of the over hang while surveying the valley below, the ridge gives way, sending him into an uncontrollable slide and tumble into the dark jungle canopy below, which transitions to a fade to black…level load fades from black to blurry shapes against a bright light that slowly comes into focus, revealing a first person view of the character’s hands and feet tied to a pole as they bob up and down against the sun and silhouetted tree canopy. The trees give way to an unobstructed view of the sky, and your captures slide you off of the pole and dangle you by a rope.  As you twist in the wind, you see that you are at the riverside clearing with the campfire that you previously saw from the ridge, and rising above the tree line in the distance you see the temple ruins upon the ridge that you were on.

Although not every game scenario provides such a literal tie in between levels, whenever an opportunity presents itself to indicate to the player that he is seeing a place where he has been from another vantage point, it adds to the robustness and immersion of the player’s gaming experience.

When you have your Game Universe Map and you have defined the components that distinguish your various regions, its time to tighten the zoom on your satellite view of your Universe and define in greater detail the features of each region. A region may bare a single level or it may inspire many levels, so depending on the nature of your subject matter and the vastness of your region, you may find it necessary to still stick to an overhead iconic map to further define features. Or if the regions are smaller and have less variation in terrain and landmarks, you might be ready to do an environment concept that reflects:

  • The type of terrain
  • The look of the architecture
  • The kind of foliage
  • The scale of things relative to the character and the color palette, lighting, and mood.

(Example of a regional environment concept painting)

Concepts can be done by any means necessary; here we have an example of a source photo for the example of the regional concept.  As you compare the photo to the painting you can tell quiet clearly that the artist hired by Fox Movie Studios, used some elements whole cloth, while embellishing on others to get the results they wanted.


Why go though the time and trouble to paint an environment concept/mood piece, rather than just collect photo reference of what you are thinking of?


Photo reference is a great way to quickly amass a collection of pictures of what certain details you might need, such as: on a piece of machinery, the type of rock formations you were thinking of, the lighting from a particular sunset, or the type of foliage you wanted to use to populate along a cliff face, but rarely do you hit every aspect of what you are trying to depict in one singular image, and a concept piece can pull all that together for you.

Guided by this mood piece, a two-pronged effort can go underway by the artists:

  1. Develop a texture set and start a look and feel level that reflects the above list of defining parameters from the 2D mood piece, and try to develop the terrain or architecture that reflects and further elaborates on all the qualities of the mood piece.


  1. From the mood piece, start and work from a list of non-game play dependant props and other environment components that can be placed not only in the look and feel level, but also for the future game level that the Level Designer is writing up.

Here is an example of an environment mood piece embellishing on the regional environment concept.

Now that we have a general sense of the content going into a level, and assuming that the Level Designer has his various rule sets for the games physics, player/AI and player/environment interaction, the Level Designer can start shaping the sequence of game play events in a design document complete with 2D overhead or isometric perspective.  This further increases the granularity of detail and now we can generate a specific asset list that encompasses not just non-game play objects, but everything needed to make a complete level.

While a library of static props, background components, foliage, interactables, and particle fx can be created by environment and fx artists, the Level Designer can review his walk through of the level with an artist to storyboard out the entire sequence of level events based off of the level design doc.  These storyboards can be quick 1 minute thumbnails (3”x 2”), for the sake of quick and nimble iteration to get the proper relationship between the camera and the player, and then a more methodical polished pass at 4”x 6” can be made to get a better sense of the relationship between the player and his environment. You will only need about a quarter of the number of the 4” x 6” storyboards to depict the same content of the thumbnails, as you just need a pulled back view from a few different angles that the game play takes place in.

The storyboards will be black and white line drawings that are meant to do the following:

  • Show the best camera views for each game play event from a cinematic perspective.
  • Re-enforce the characteristics of terrain and or architecture from the original mood piece for the area.
  • Confirm what the scale of the immediate game play area should be for any given space as defined by the overhead maps from the design docs.
  • Storyboarding can also be critical in finishing one of the objectives of the game universe map, which is to achieve congruency between levels.

This is an example of a sequence of larger sketches that shows the environment space relative to the player, throughout a game play scenario.

This is an example of a sequence of thumbnail sketches on a page that shows the position of the camera relative to the player, throughout a game play scenario.

You may want to have info lines to explain the various actions and camera movement, such as the example below.  This an example of something that could scratched out as dictation in a matter of seconds during a meeting.

Once the storyboards are done, the Level Designer and or Environment Artist depending on available personnel can white box the level with primitive geometry that would be appropriate to use as first pass collision for a future arted version of the level. The storyboards are initially used to guide in creating white box spaces that exploit the above objectives of the storyboards.  While this is being done, another pass at doing a colored mood piece from the storyboards of each major area of the level can be created to help the Environment Artist maintain consistency in style and color palette throughout the level. On average a level will probably have about four of these mood pieces.  The difference between these mood pieces and the initial mood piece created, is that the first mood piece was created based on vague generalities, where these next set of mood pieces are based on specific level design content, making the information gleaned from them much less ambiguous.

Example of new mood piece specifically based on the storyboards, and or screen captures of the white box in progress. This is the screen capture of the white box that guided the mood piece above.

Art props and interactables are pulled into the white box terrain as they become available, but the level remains in white box stage until the scale and game play have been vetted and approved. Basic texture and minimal lighting can be applied as needed, for the sake of the game play function.  The reason to hold back on developing polished art on a white-boxed level design is to ensure that level design changes can be made quickly without getting bogged down by changing artist details. Artists can work parallel to the white box by compose texture sets, shader combinations, and lighting treatments onto test geometry specifically made for the white box level in question in a separate test level so that when a game level’s white box design has been finalized many of the questions about the 3D artistic treatment have been worked out and then it’s just a matter of propagating more of the same in the white box level.  This method will decrease the amount of time that artists spend on throw-away-content and put their energy toward other parts of the game that need to be developed anyway, whether its 2D concept, storyboarding or prop and background art creation.

Once a solid playable white box is completed, the Environment Artist can proceed with a detailed art pass in the following order:

        Terrain and or architectural geometry

        Texture mapping and blending

        Final prop placement adjustments

        Lighting pass

        Contact/ground shadows

        Optimize visible and invisible level components

The next page shows a timeline/schedule progression on how much time is given to complete each stage, and how the various team member positions overlap in their tasks in relation to each other’s timelines.  The timeline begins post-completion of the GDD.

The timeline is a rough approximation for the amount of time needed between the tasks and does not necessarily account for iteration, that is to say, that it assumes that the concept pieces and storyboards were nailed on the first time without redoing anything.  However, the areas indicated as “on going” is extra buffer time that can absorb concept iteration. The good thing about this process is that it is designed to get the iterations out of the way quickly and up front, so that it minimizes the amount of redoes.  It should also be noted that the timeline is representing one level, rather then multiple levels over the course of one project, and obviously a final alpha level is not going to be finished just ten weeks after the GDD is completed.  But with the exception of maybe the Universe Map and the regional maps, this timeline is meant to be viewed as template for wash, rinse, and repeat.

This proposed pipeline will:

        Reduce in redoing work

        Remove ambiguity for the environment artist

        Inspire and spur on creativity among the designers from the visual reference

        Improve quality of both the level design and the art

The result of these benefits will be that the Environment Artists will be able to art out the environments in a quarter of the time and half the time overall when accounting for dependencies across involved disciplines.  Over the life of the project, the time saved could potentially be expediential.


back to the top