Comparative differences between games design and simulation design - Part 2: Environment Design

Environment / Level Design

Introduction

The scenario or the environments that are used within games / sims are dependent on the world / situation that is needed. Below is a breakdown and example of the comparisons that present one certain way I was taught how to build both an environment in the world of a game and simulation. (Note: there are many ways to process the information and build environments and this way was the predominate one that was taught to me).

Environment Design

The environment design works on both what area you are in but also what the main purpose is of that environment. Games design 1-0-1 talks about how aesthetics and good research of the environment with concept documentation gives a strong clear example of what is needed.

From concept to map research there are similarities and separation from each other that will be looked at and broken down within this section of the post.

Games Design (Unreal Engine 4)

Games design works on the pipeline process of concept to full form through various stages. Conceptual art presents the game directors feels and ideas of the environment which is then passed onto the environment designers who then build create and make the concept art as close to the same as possible.

All the placement and data is help inside the engine with the main process done from the engine itself.

If I remember the process, it goes for terrain design, blockout (placing blocks where objects would be), asset placement, lighting and post processing.

Although you can make the terrain in the world, some designers make the data beforehand making a terrain file or height file for the engine to process and make the world using colour co-ord. The data file that can be used is called a terrain height map that uses the colour scale (normally of black to white with the variants of grey as the in-between parts of the world which is easily made with programs such as photoshop, GIMP, fireworks.

The objects are manually placed in the world by hand which means that positioning is long and slow ensuring pieces are positioned correctly to allow for correct positioning and ensuring all models have the right scale, correct U.Vs and lightmaps to ensure no lightmap or texture pixilation or quality loss as well as object it is correct to the rest of the world.

Simulation (VR-Forces)

Simulation works on the same premise for parts of the environment design but not all of it.

Simulation Data Storage

The way the data is held works differently with a network of data files that point to specific xml sheets that hold key information. The files held are called earth files which are the central hub of where the project needs to go to manually find any of the following:

• Terrain Data 
• Road 
• Pavement 
• Building 
• Trees 
• Additional prop (asset) data 
• And Rivers / Lakes

The data in the files will then have an array of data files in the form of an XML list that we then layout specifically for the engine to read and pull in the data we need, such as style of object, formatting and name definition in engine.

Below is an example of how it works

Environment Design

As described in a previous blog in regards to how simulation engines work, the processing of terrain data is done normally with some DTED or DEM map data. This instead of being an image is instead a data file that holds data in a matrix styled system but is more refined with more data and use of colour gradients than the black / white of games design.

As simulation normally works within real world locations, the used of databanks such as LIDAR databases that hold terrain information can be used for such things as detecting rivers and lakes and how high above sea level is the land. Which placed into the world can be calculated to provide a more realistic position of the terrain for simulation use.

Prop Placement Files

Although the terrain can hold the information, sometimes it can crash due to excessive information being stored. With that we can create prop lists that hold the data of necessary things we need in our world whether it is the buildings, roads or other necessary information.

We can store them into what is known as shapefiles, shapefiles (please look at other blog specifically around engines) can hold key information that we need to hold and although referenced as a file; it can be called to the engine with use of these shapefiles. One positive with the use of these files is that we can un-connect the file and replace if necessary and or replace them with other files if necessary and all that is needed a reload of the environment without the necessity of removing the connections overall.

Although impractical within the games design engines, this does allow for the key parts of a sim to be easily replaceable if it is needed.

Learning Factors (When learning these skills)

The main issue that was had when learning these skills is the provided guidelines or directions in case something goes terribly wrong or if it works and then stops working. As previously stated in another blog post, the wondrous thing about game engines is that there is a plethora of information in the world in regards to how to fix, create, start or finish products from both other users or by devs (games or engine) without having to do much other than a forum search or google search which cannot be the same for SIM engines whom have very little in the form of forums or aid when issues appear.

That said, if you have a person who has had experience, they can guide you in the process of building an environment. As an example, I built a quadrant of Bristol in a sim engine within a day (7.5) with working buildings, trees, roads and working terrain file which is average for a first time person going through the proper procedure. That said, it’s the same issue in regards to if it goes wrong, who is there to help you. The main issue I can see through learning is the idea of support which is key to ensure top range engineers / designers for future products.

Conclusion

Although the process sounds different, when you sit down and work on it and look at the process the information, there isn’t much difference in how you learn how to make environments / levels. It’s all about practise but without that key support matrix behind with multiple “go to” points, you’re stuck waiting on something that could be easily swept under a rug and you’re stuck with nowhere to go and as previously said, that is not effective for your own time (which will be limited), your companies money and the restricted time you probably will have.

Overall, I think for the months of learning simulations, that is the key thing I’ve learnt, writing down and logging information in case it breaks for future systems as no doubt they’ll be common repetitive situations that will happen and personally, the idea of waiting weeks for a response for a fix or guidance to a fix is incredibly frustrating.