Jump to content


Real time procedural planets


3 replies to this topic

#1 jstier

    New Member

  • Members
  • PipPip
  • 12 posts

Posted 09 July 2009 - 02:00 PM

Posted Image


Description
This is a screenshots of the real time Geist3D graphics engine which we are using for an independent game development project called Pangea. The most distinct feature of the game is a huge environment supporting seamless travel between any two points in a solar system of detailed procedural planets. The planets are

Although the graphics engine is coming along the game mechanics and artwork are still in an early stage. You can download a trailer that allows you to use characters, ground vehicles, star gates and space ships to explore the environment. You can also have a look at videos and additional pics at www.geist3d.com


Some of the key features of the engine are:

- Rudimentary rigid body physics throughout the entire environment
- Deferred lighting with point, directional and spot lights.
- Histogram based HDR, SSAO and Bloom
- Atmospheric scattering
- Large procedural planets up 1.5 resolution.

The engine also includes an editor to customize the planets, import 3D meshes, define GUIs and develop GLSL programs and Lua scripts. Everything is still under development and on shacky grounds at times, but there are a few indie artists who are currently using the tools to develop content.

#2 TheNut

    Senior Member

  • Moderators
  • 1701 posts
  • LocationCyberspace

Posted 09 July 2009 - 03:12 PM

Nice, though it would have been nice to see a YouTube video or some such. I assume you're using shaders to handle atmospheric scattering? I based my solution off O'Neil's work, but I sometimes run into anomalies where the atmosphere doesn't always render correctly, especially when the you don't use a large number of vertices for the atmosphere. I found the distance from the planet to be a number one algorithm breaker. Have you run into any such scenarios and / or solved them? I haven't put a whole lot of time into it.
http://www.nutty.ca - Being a nut has its advantages.

#3 jstier

    New Member

  • Members
  • PipPip
  • 12 posts

Posted 09 July 2009 - 09:51 PM

Yeah, I realized that I didn't put in any links to the website. I didn't know how to go back and edit the post. So, here are some movies and additional images:

http://www.geist3d.com/gallery
Check out Geist3D at www.geist3d.com

#4 jstier

    New Member

  • Members
  • PipPip
  • 12 posts

Posted 09 July 2009 - 09:58 PM

Quote

I based my solution off O'Neil's work, but I sometimes run into anomalies where the atmosphere doesn't always render correctly, especially when the you don't use a large number of vertices for the atmosphere. I found the distance from the planet to be a number one algorithm breaker. Have you run into any such scenarios and / or solved them? I haven't put a whole lot of time into it.

I ran into all kinds of problems with the scattering. I also started with O'Neil's method but somehow his shader didn't seem complete to me. In the last iteration I worked over the extinction factors again and then things started looking a bit better. You can have a look at the shaders in the Geist3D editor.
Check out Geist3D at www.geist3d.com





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users