## Seaside movie clips

Share with us how are you using the powerrrr of the force

Moderator: Alain

### Seaside movie clips

That boat has too much power.

Done with Truevision3D and Newton physics. No Newton buoyancy though, only SetForce and SetTorque. Works like a charm .

http://personal.inet.fi/surf/epsilon/storm_4_5.avi
http://personal.inet.fi/surf/epsilon/storm_4_6.avi

(Movies around 3 MB each, you need DivX 5.2.1)

/ Waterman

Things should be described as simply as possible - but not simpler [A. Einstein]
Waterman

Posts: 7
Joined: Sat Mar 05, 2005 3:11 pm
Location: South coast of Finland

:O brilliant, absolutely brilliant!!
Help improving the Newton Game Dynamics WIKI

JernejL

Posts: 1521
Joined: Mon Dec 06, 2004 2:00 pm
Location: Slovenia

Nice storm & surfy bits.

Considering ditching buoyancy myself - it's rather hard to tweak & get right
Aphex

Posts: 144
Joined: Fri Jun 18, 2004 6:08 am
Location: UK

Care to share some info - how did you make the dynamic wawes collision, similar stuff ?
Or coud you just point me to some links with related info ?
RedDrake

Posts: 2
Joined: Sun Jun 13, 2004 7:28 pm

Sure. Actually, a short explanation can be found in this forum. Check:

http://newtondynamics.com/forum/viewtopic.php?t=1111

There's also something in the TrueVision3D forum:

http://www.truevision3d.com/phpBB2/viewtopic.php?t=8778&highlight=storm
http://www.truevision3d.com/phpBB2/viewtopic.php?t=8222&highlight=storm
http://www.truevision3d.com/phpBB2/viewtopic.php?t=8065&highlight=storm

These are partially outdated to the water surface part, but the physics principles are still about the same. It's essential to understand that the boat would move exactly the same way even if the water mesh wasn't ever created. The height at any point in time and any location is just a f(X) that is identical to the same f(X) in the vertex shader.
Things should be described as simply as possible - but not simpler [A. Einstein]
Waterman

Posts: 7
Joined: Sat Mar 05, 2005 3:11 pm
Location: South coast of Finland

One question...
If you're duplicating the wave calculations on the cpu and the shader, wouldn't it be faster (or free up some vertex shader power) to just do it on the cpu and draw normally?
Aphex

Posts: 144
Joined: Fri Jun 18, 2004 6:08 am
Location: UK

Perhaps one could imagine so, but if looking closer at the scenario, it is obvious that it must be done this way:

The water surface is large, even with LOD it consist of around 50000 vertices. The vertex shader has to calculate the height of each of those during each re-paint. Also not that these vertices are at fixed lcations in the XZ plane.

The boat physics need to know the surface hieght also, but *only under the boat*. This is a question of calculating Y for 16 positions of XZ only. In addition, these 16 XZ pairs are different each time, unless to boat sits still.

A vertex shader is superiorly fast. Consider calculating 50000 vertex positions 80 times per second (thats 4 million height calculations / second) in the CPU. The calculation is not simple at all, so you could say "see you later" to your computer.

As a general rule, a vertex shader is ALWAYS faster. Much faster.
Things should be described as simply as possible - but not simpler [A. Einstein]
Waterman

Posts: 7
Joined: Sat Mar 05, 2005 3:11 pm
Location: South coast of Finland

50000 vertices! I can see why the speed is a must!
Aphex

Posts: 144
Joined: Fri Jun 18, 2004 6:08 am
Location: UK

I am sorry, i gave you info from a previous version, so it vas quite exaggerated, compared to the latest achievements. The latest version of the surface has:

- 24786 triangles
- 12496 vertices

Quite a lot, even so.
Things should be described as simply as possible - but not simpler [A. Einstein]
Waterman

Posts: 7
Joined: Sat Mar 05, 2005 3:11 pm
Location: South coast of Finland

### Who is online

Users browsing this forum: No registered users and 4 guests