flipCode - Tech File - Phil Carlisle [an error occurred while processing this directive]
Phil Carlisle
Click the name for some bio info

E-Mail: phil.carlisle@team17.com



   01/12/2001, Tech File Update


Hi Guys,

Little update, Xmas has taken its toll on new stuff.

Ive gotten my quadtree terrain working again (after a lot of messing with other stuff in the meantime), the main things I'm trying right now are in methods of using TnL to brute force the landscape (using terrain patches of 32x32 and patch based culling). I'm also looking more into a more flexible terrain texturing method (before I was doing a fairly intricate alpha blending thing, which was a bit too slow to do on the fly).

Terrain texturing is one area that Ive been very interested in, because its a way to get something that looks unique, rather than copying one of the 1000 ROAM based engines, besides, after trying the roam route, I dont think its useful for my current interest ( i.e. I wanted to have a 16km x 16km terrain with a 4m resolution), not that roam precludes this, but it just doesnt suit what I require.

I liked the terrain on glVelocity in terms of texturing, he basically did what I was doing in my last test (a simple alpha blend), one thing he did that I wasnt doing, was that he actually adds a random offset into the alpha value (to give the transition between one texture and the next a nice amount of noise), he also doesnt actually blend the texture, rather, he sets the texture in a band depending on the alpha value (in effect he chooses an integer texture number from a floating point value).

One thing he hasnt looked at yet is the effect of slope on terrain texture. And also on how to best use the detail texture.

Currently, my plan is to get something looking pretty similar to XIsle demo (on nvidia's site), but on a dynamically loading patch based landscape of 16km square. If you look at the XIsle demo a bit, you notice that it doesnt actually do any mesh simplification (or at least not immediately recognisable), which is one of the reasons I am looking at methods of brute forcing things using simple patch culling etc.

My main task for now is to finish the patch management, then finish the texture generation algorithms, then pop those classes into a nice windows interface where I can drag and drop terrain features (and incorporate those into my world editor).

Basically, work as usual really :))

Phil.





  • 01/09/2002 - Tech File Update
  • 04/03/2001 - Tech File Update
  • 02/23/2001 - Tech File Update
  • 02/19/2001 - Tech File Update
  • 01/12/2001 - Tech File Update
  • 11/25/2000 - Tech File Update
  • 11/02/2000 - Tech File Update
  • 09/28/2000 - Tech File Update
  • 08/21/2000 - Tech File Update
  • 07/03/2000 - Tech File Update
  • 06/05/2000 - Tech File Update
  • 05/22/2000 - Tech File Update
  • 02/29/2000 - Tech File Update
  • 01/27/2000 - Just An Update For The Tech Files
  • 12/23/1999 - New Years Tech File Update
  • 12/08/1999 - Update To Techfile
  • 11/30/1999 - Introduction

  • This document may not be reproduced in any way without explicit permission from the author and flipCode. All Rights Reserved. Best viewed at a high resolution. The views expressed in this document are the views of the author and NOT neccesarily of anyone else associated with flipCode.

    [an error occurred while processing this directive]