Hello,
I was wondering what the advantages/disadvantages would be in using mesh terrain blocks over the built in terrain system with heightmaps? I'm currently using PnP TerrainCreator and i get the best results when I use meshes with vertex alpha. However, i assume that using mesh blocks means I can't use the built in terrain system and I'm wondering what I lose by going in this direction.
Thank you for your help!
Heightmap vs Mesh Terrain Blocks?
Thank you for the reply!
One last question. Would you recommend the use of mesh blocks for terrain or is this generally considered bad practice? I'm a seasoned developer but new to 3d games development so any suggestions/advice would be helpful. I'm just trying to get a feel for some best practices in this area.
Thanks again!
One last question. Would you recommend the use of mesh blocks for terrain or is this generally considered bad practice? I'm a seasoned developer but new to 3d games development so any suggestions/advice would be helpful. I'm just trying to get a feel for some best practices in this area.
Thanks again!
This question is really application specific. It depend what are you doing and for what your terrain is made.
Some of the premodeled terrain advantages are that you are not limited by grid so you can form any terrain feature you want: vertical cliff or overhang for example. Also you have complete control over polygon distribution so you can make flat parts made of few polygons and put more polygons in to parts which require more detail.
Generic terrains on the other hand are good in applications where more general approach is required. For example if you want user to load any terrain he wants from height map.
This kind of terrains tend to give you much more flexibility.
Also since they are grid based they tend to be faster in collision detection ...if they have appropriate functions build in.
Some of the premodeled terrain advantages are that you are not limited by grid so you can form any terrain feature you want: vertical cliff or overhang for example. Also you have complete control over polygon distribution so you can make flat parts made of few polygons and put more polygons in to parts which require more detail.
Generic terrains on the other hand are good in applications where more general approach is required. For example if you want user to load any terrain he wants from height map.
This kind of terrains tend to give you much more flexibility.
Also since they are grid based they tend to be faster in collision detection ...if they have appropriate functions build in.
-
hybrid
- Admin
- Posts: 14143
- Joined: Wed Apr 19, 2006 9:20 pm
- Location: Oldenburg(Oldb), Germany
- Contact:
The problem with terrain is the usually high poly count. The terrain scene node has the dynamic LOD system for this purpose, reducing the details in far away sections. The tiled terrain node uses a sector based culling for tiles which are not visible. Using a plain mesh (or the terrain mesh from GeometryCreator) would simply render the whole mesh all the time (besides usual frustum culling).
You could try to enhance this situation with an octree node or with hw buffers (available in Irrlicht 1.5/SVN). However, still all detail is rendered at any distance. So you have to check whether the target systems can run under this load.
You could try to enhance this situation with an octree node or with hw buffers (available in Irrlicht 1.5/SVN). However, still all detail is rendered at any distance. So you have to check whether the target systems can run under this load.