
Outerra
The whole world as playground!
av Cri74 | publisert: 13. mars 2011 | endret: 23. oktober 2011
This is the idea behind the new engine called Outerra. It promises seamless rendering, traveling from outer space and way down to the inches of grass anywhere on the planet, travel across the seas, and vast mountains to explore the whole world in this new and innovative engine. Outerra is a small independent game studio based in Slovakia, founded in 2010 and driven by developers with a common interest to create a complete, planetary scale virtual environment for future games and simulations. Battle.no have met the makers and developers at Outerra, and got an exclusive and informative interview. With us today is Brano Kemen.

First of all. Thank you for taking the time Brano. What is Outerra and what can you do with this engine?
Outerra was initially a the name we choose for our small company when we decided to focus on our engine. But it automatically became used for the engine itself, sometimes referred to as “Outerra Engine” or OT. The engine is meant for games, simulations and visualizations that need to render whole planets with detail ranging from space views down to ground level.
Is the engine planned to work with other companies, do you have sponsors .. etc ?
We are working with several companies at the moment, they are evaluating and testing their preliminary designs. Most of these are mainly operating in simulator business. At the moment we are self-funded.
How did Outerra start and what was the goal from the beginning?
It all started way back, as I was always interested in terrain rendering and generation, and it was almost the first thing I went to do when I first got my hands on a computer. I didn’t preserve my very first attempts, but I’ve managed to keep some screens from preceding versions.
Initially the terrains were completely randomly generated, but later I got more interested in utilizing real data to render terrains, while still keeping the part that is able to refine the detail down to the level needed in FPS games. Today OT is able to use real or artificially created elevation data for whole planet, refine it down to centimeter resolutions of terrain, render it with unlimited visibility distance from outer space seamlessly down to the surface with micro details, together with vast forests, oceans and more.

Can you explain what is different in Outerra compared to other engines?
One of the major differences is the fact that Outerra will come with a world (a whole planet) that is ready from the start, in a pristine state the planet might have looked before the civilization. From the game makers’ perspective it means they will be modifying and populating an existing location on Earth instead of getting an empty constrained level to work with from scratch.
Outerra was designed with effective level of detail management in mind, so that it can be one day used as a unified ground for many simulators and games that can take place anywhere - in space, air, sea and ground.
Are there any plans for releasing it as a game, and what kind of game?
We are currently working to produce a game that would demonstrate the engine, while at the same time helping us to mature the engine more in the process. It ought to be a combination of a sand-box game and a building/transportation/driving game initially, from a first person view, also with some usual FPS fun as well.
A “hidden” goal is to evolve integrated tools and functionality that will be needed for world building so that it can eventually become a kind of a platform where developers could implement their game ideas without having to deal with the issues of world rendering and modeling.
After the model of few other indie games, the game should be released early in alpha stage giving people the ability to support and also partially drive its development.

Outerra is about the whole planet. One question that comes to mind would be that Outerra contains a huge amount of information since we are talking about a whole planet and a world?
The whole dataset can be relatively big, using ~76m real world data spacing we are getting to 14GB of terrain data. However, the data can be streamed on demand; in that case only a few tens of MBs are required for a place of landing, with full visibility towards the horizon. With realistic aircraft speeds and a normal internet connection the data can be streamed and cached on disk smoothly as you go. In fact it would take you quite a long time to traverse Earth at low altitude, to force download of the whole dataset.
In some of the pictures on the site we are able to see a web integration, could you tell us a little about that and what it could be used for?
The engine integrates a web browser and that not only allows it to render web pages directly within the game screen, but also enables an interaction between web pages and engine. This allows for integrating existing web applications easily. For example, you can have Google Maps synchronized to the camera, with possibility to move the camera to a selected place on the maps and vice versa. Things like a web based chat can be integrated really fast. In fact, most of engine tools and UI are written as normal html pages, invoking engine functionality from their embedded scripts. The potential for this is yet to be fully exploited.

How long have Outerra been in development, what stage would you say Outerra is at this time and what is the main focus of development right now?
That is kind of hard to tell exactly, I was working on it as a hobby for a long time before we joined together with Laco Hrabcak (known also as “angrypig” on our forums), and started to work on it more seriously along with our day jobs. As the potential and our commitment gradually grew, this mode of operation became unbearable, so for over a year we are working full time on Outerra. Nonetheless, the engine is still relatively alpha, especially with regards to the tools. That’s also why we want to mature it during a development of a game that will better show what’s needed.
In Outerra forums there is a wide range of interest and suggestions on how to use the engine, does this mean you are also developing by request from fans?
There’s indeed a wide range of interests and ideas. We obviously cannot be working on everything people want, but we are listening to them and trying to answer their questions about what will be supported and what are the possibilities. With the engine we are trying to be enticing to a wide range of interests there are, be it simulators - flight, railroad, sea, driving ones, or the multitude of game types that can use the engine.
We are also using the feedback we are getting from our forum but also from other forums out there discussing Outerra, to determine what should go into our game. It’s a bit hard to combine the various interests of people, but I hope the game will be able to achieve it in the end.
We also saw a thread where forum members can request pictures by supplying longitude and latitude, global positioning?
Yes, that was meant for the time there’s no public demo available so that people can see how locations of their interest look like. Note that currently it only shows the detail of meshes, otherwise there’s still just one climate and land type everywhere in the world. The datasets used there aren’t final either, and a better data will be coming to the final product.

Will there be a demo available for everyone?
Yes we plan to have a public technology demo, probably merged with a benchmark as well. There have been some problems with OpenGL drivers but these are being gradually resolved. We would like to have the demo out in few months, probably around the time when an alpha of the game comes out.
What kind of Hardware would you say is needed to use this engine/demo?
The engine heavily uses modern GPU features and it requires DX10-class hardware. For the alpha, which won’t contain all the optimizations yet, we recommend a card around GTX460 level of performance. Otherwise a 8800GT (= GTS250, 9800GT) card should be considered minimum, anything below will probably need a reduction in rendering quality to keep up.
The engine itself is not CPU bound, but for smoothness it needs a 2-core CPU. Of course, additional game code can elevate the demands on CPU.
Would you say this is a mod friendly engine, will users/players be able to create easy?
Engines can be probably always considered “mod friendly”, question is if a particular game opens up the modding possibilities and how complicated the process is. But since we want to iterate towards a kind of game operating system, so to say, that would allow developers to create games by writing their game code and assets but not having to deal with world rendering, modeling and management aspects, I’d say the modding capabilities should be pretty high after we finish all the tools it requires.

Battle.no/TV2 are hosting Arma2, there has been some speculations within the community and rumours in a few forums about Outerra being the new Arma3. Would you comment on this?
You know, you should not believe me even if I said it was true.
Switching the engine is quite costly, especially when the game code is already heavily tied to the old one. And it’s not like people will not buy Arma3 unless it possesses the attributes of Outerra. So from an economic standpoint it would not make any sense, unless there’s a potential competitor coming with a similar product that exhibits these advanced features. Even though the prospect is intriguing for players, it must be also viable for the developer.
There’s a similar situation with WWII Online Battleground Europe game - we’ve attracted many of it’s players and are causing periodic requests to use Outerra as the engine - but it would mean starting to code the game almost from scratch, which requires a serious investment. Nevertheless, our contacts and feedback indicate that something is bound to happen. Developers are slowly discovering capabilities of the engine and the possibilities that are getting opened here.
Thanks again Brano, we hope to hear from you when the demo is available. Outerra and the possibilities are impressive!
Outerra’s official site is found @ www.outerra.com
Developer blogs @ http://outerra.blogspot.com/

First of all. Thank you for taking the time Brano. What is Outerra and what can you do with this engine?
Outerra was initially a the name we choose for our small company when we decided to focus on our engine. But it automatically became used for the engine itself, sometimes referred to as “Outerra Engine” or OT. The engine is meant for games, simulations and visualizations that need to render whole planets with detail ranging from space views down to ground level.
Is the engine planned to work with other companies, do you have sponsors .. etc ?
We are working with several companies at the moment, they are evaluating and testing their preliminary designs. Most of these are mainly operating in simulator business. At the moment we are self-funded.
How did Outerra start and what was the goal from the beginning?
It all started way back, as I was always interested in terrain rendering and generation, and it was almost the first thing I went to do when I first got my hands on a computer. I didn’t preserve my very first attempts, but I’ve managed to keep some screens from preceding versions.
Initially the terrains were completely randomly generated, but later I got more interested in utilizing real data to render terrains, while still keeping the part that is able to refine the detail down to the level needed in FPS games. Today OT is able to use real or artificially created elevation data for whole planet, refine it down to centimeter resolutions of terrain, render it with unlimited visibility distance from outer space seamlessly down to the surface with micro details, together with vast forests, oceans and more.

Can you explain what is different in Outerra compared to other engines?
One of the major differences is the fact that Outerra will come with a world (a whole planet) that is ready from the start, in a pristine state the planet might have looked before the civilization. From the game makers’ perspective it means they will be modifying and populating an existing location on Earth instead of getting an empty constrained level to work with from scratch.
Outerra was designed with effective level of detail management in mind, so that it can be one day used as a unified ground for many simulators and games that can take place anywhere - in space, air, sea and ground.
Are there any plans for releasing it as a game, and what kind of game?
We are currently working to produce a game that would demonstrate the engine, while at the same time helping us to mature the engine more in the process. It ought to be a combination of a sand-box game and a building/transportation/driving game initially, from a first person view, also with some usual FPS fun as well.
A “hidden” goal is to evolve integrated tools and functionality that will be needed for world building so that it can eventually become a kind of a platform where developers could implement their game ideas without having to deal with the issues of world rendering and modeling.
After the model of few other indie games, the game should be released early in alpha stage giving people the ability to support and also partially drive its development.

Outerra is about the whole planet. One question that comes to mind would be that Outerra contains a huge amount of information since we are talking about a whole planet and a world?
The whole dataset can be relatively big, using ~76m real world data spacing we are getting to 14GB of terrain data. However, the data can be streamed on demand; in that case only a few tens of MBs are required for a place of landing, with full visibility towards the horizon. With realistic aircraft speeds and a normal internet connection the data can be streamed and cached on disk smoothly as you go. In fact it would take you quite a long time to traverse Earth at low altitude, to force download of the whole dataset.
In some of the pictures on the site we are able to see a web integration, could you tell us a little about that and what it could be used for?
The engine integrates a web browser and that not only allows it to render web pages directly within the game screen, but also enables an interaction between web pages and engine. This allows for integrating existing web applications easily. For example, you can have Google Maps synchronized to the camera, with possibility to move the camera to a selected place on the maps and vice versa. Things like a web based chat can be integrated really fast. In fact, most of engine tools and UI are written as normal html pages, invoking engine functionality from their embedded scripts. The potential for this is yet to be fully exploited.

How long have Outerra been in development, what stage would you say Outerra is at this time and what is the main focus of development right now?
That is kind of hard to tell exactly, I was working on it as a hobby for a long time before we joined together with Laco Hrabcak (known also as “angrypig” on our forums), and started to work on it more seriously along with our day jobs. As the potential and our commitment gradually grew, this mode of operation became unbearable, so for over a year we are working full time on Outerra. Nonetheless, the engine is still relatively alpha, especially with regards to the tools. That’s also why we want to mature it during a development of a game that will better show what’s needed.
In Outerra forums there is a wide range of interest and suggestions on how to use the engine, does this mean you are also developing by request from fans?
There’s indeed a wide range of interests and ideas. We obviously cannot be working on everything people want, but we are listening to them and trying to answer their questions about what will be supported and what are the possibilities. With the engine we are trying to be enticing to a wide range of interests there are, be it simulators - flight, railroad, sea, driving ones, or the multitude of game types that can use the engine.
We are also using the feedback we are getting from our forum but also from other forums out there discussing Outerra, to determine what should go into our game. It’s a bit hard to combine the various interests of people, but I hope the game will be able to achieve it in the end.
We also saw a thread where forum members can request pictures by supplying longitude and latitude, global positioning?
Yes, that was meant for the time there’s no public demo available so that people can see how locations of their interest look like. Note that currently it only shows the detail of meshes, otherwise there’s still just one climate and land type everywhere in the world. The datasets used there aren’t final either, and a better data will be coming to the final product.

Will there be a demo available for everyone?
Yes we plan to have a public technology demo, probably merged with a benchmark as well. There have been some problems with OpenGL drivers but these are being gradually resolved. We would like to have the demo out in few months, probably around the time when an alpha of the game comes out.
What kind of Hardware would you say is needed to use this engine/demo?
The engine heavily uses modern GPU features and it requires DX10-class hardware. For the alpha, which won’t contain all the optimizations yet, we recommend a card around GTX460 level of performance. Otherwise a 8800GT (= GTS250, 9800GT) card should be considered minimum, anything below will probably need a reduction in rendering quality to keep up.
The engine itself is not CPU bound, but for smoothness it needs a 2-core CPU. Of course, additional game code can elevate the demands on CPU.
Would you say this is a mod friendly engine, will users/players be able to create easy?
Engines can be probably always considered “mod friendly”, question is if a particular game opens up the modding possibilities and how complicated the process is. But since we want to iterate towards a kind of game operating system, so to say, that would allow developers to create games by writing their game code and assets but not having to deal with world rendering, modeling and management aspects, I’d say the modding capabilities should be pretty high after we finish all the tools it requires.

Battle.no/TV2 are hosting Arma2, there has been some speculations within the community and rumours in a few forums about Outerra being the new Arma3. Would you comment on this?
You know, you should not believe me even if I said it was true.
Switching the engine is quite costly, especially when the game code is already heavily tied to the old one. And it’s not like people will not buy Arma3 unless it possesses the attributes of Outerra. So from an economic standpoint it would not make any sense, unless there’s a potential competitor coming with a similar product that exhibits these advanced features. Even though the prospect is intriguing for players, it must be also viable for the developer.
There’s a similar situation with WWII Online Battleground Europe game - we’ve attracted many of it’s players and are causing periodic requests to use Outerra as the engine - but it would mean starting to code the game almost from scratch, which requires a serious investment. Nevertheless, our contacts and feedback indicate that something is bound to happen. Developers are slowly discovering capabilities of the engine and the possibilities that are getting opened here.
Thanks again Brano, we hope to hear from you when the demo is available. Outerra and the possibilities are impressive!
Outerra’s official site is found @ www.outerra.com
Developer blogs @ http://outerra.blogspot.com/
Here are 3 videos to show Outerra - PUSH HD
blog comments powered by Disqus
