In the last weeks, I made some research for a future project of mine, which will feature procedurally generated terrain.
And one of the tests has reached a state in which I could write an article about it on my blog.
So I did!
Please note that this is not about the project, but really about the terrain research itself
The next step for me is to research another approach, fully relying on 3D noise instead of heightmaps.
After that, I will actually implement both tests using the Volume Component.
In the end, I will have to make a decision for the project, it will either use PolyVox or the Volume Component. We shall see
Also, have a picture
Procedural Heightmap-based Terrain with Overhangs
-
- OGRE Retired Team Member
- Posts: 972
- Joined: Mon Jun 02, 2008 6:52 pm
- Location: Berlin
- x 65
-
- Orc Shaman
- Posts: 788
- Joined: Mon Jan 18, 2010 6:06 pm
- Location: Costa Mesa, California
- x 24
Re: Procedural Heightmap-based Terrain with Overhangs
Use Volume component! We need an example anyway. PolyVox is ok but limited...
-
- OGRE Retired Team Member
- Posts: 972
- Joined: Mon Jun 02, 2008 6:52 pm
- Location: Berlin
- x 65
Re: Procedural Heightmap-based Terrain with Overhangs
I would prefer to, but it wouldn't make sense without paging.
But the project itself won't start before June, so I hope Philip has enough time to include paging until then
But the project itself won't start before June, so I hope Philip has enough time to include paging until then
-
- Orc Shaman
- Posts: 788
- Joined: Mon Jan 18, 2010 6:06 pm
- Location: Costa Mesa, California
- x 24
Re: Procedural Heightmap-based Terrain with Overhangs
I'm sure he will
-
- Google Summer of Code Student
- Posts: 550
- Joined: Thu Jun 04, 2009 5:07 pm
- Location: Berlin
- x 108
Re: Procedural Heightmap-based Terrain with Overhangs
In the blog posting, you are talking about smoothing the terrain. You actually use a block-kernel for this (same weight for every neighbor). Maybe a Gauss-Kernel could provide better results (gauss weights).
http://en.wikipedia.org/wiki/Discrete_G ... ian_kernel
Here is a neat tool to generate the 2D-case:
http://www.embege.com/gauss/
http://en.wikipedia.org/wiki/Discrete_G ... ian_kernel
Here is a neat tool to generate the 2D-case:
http://www.embege.com/gauss/
Google Summer of Code 2012 Student
Topic: "Volume Rendering with LOD aimed at terrain"
Project links: Project thread, WIKI page, Code fork for the project
Mentor: Mattan Furst
Volume GFX, accepting donations.
Topic: "Volume Rendering with LOD aimed at terrain"
Project links: Project thread, WIKI page, Code fork for the project
Mentor: Mattan Furst
Volume GFX, accepting donations.
-
- OGRE Retired Team Member
- Posts: 972
- Joined: Mon Jun 02, 2008 6:52 pm
- Location: Berlin
- x 65
Re: Procedural Heightmap-based Terrain with Overhangs
Of course, nothing in that test is production ready. I didn't do any threading (the topic practically begs for it), the smoothing is stupid, the triplanar texturing is still called SimpleFragment.frag , etc.PhilipLB wrote:In the blog posting, you are talking about smoothing the terrain. You actually use a block-kernel for this (same weight for every neighbor). Maybe a Gauss-Kernel could provide better results (gauss weights).
http://en.wikipedia.org/wiki/Discrete_G ... ian_kernel
Here is a neat tool to generate the 2D-case:
http://www.embege.com/gauss/
It is just a proof of concept
But thanks for the links! I'll probably implement that as soon as I work on the project itself.