# OpenGL tilemap in 3d not 2d - 3d newb

This topic is 4422 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

## Recommended Posts

##### Share on other sites
It's quite simple to create a map:
Keep it 2D!

At least the structure of it.

How to load models/heightmaps are available on the net.

Basically a Heightmap is a grid (much like your gamemap, but with the numbers representing height instead of a tilenr) which is looped through with:
glBegin(GL_TRIANGLES);for (x < width-1) for (y < height-1)   glVertex3f(x+1, height(x+1,y+1), y+1);  //corner A   glVertex3f(x, height(x,y+1), y+1);      //corner B   glVertex3f(x, height(x,y), y);          //corner C   glVertex3f(x+1, height(x+1,y), y);      //corner DglEnd();float height(int x,int y) {// reads height from heightmap data (i.e standard grid) }

Code is not tested, and is pseudo.

Using the various techniques has their own pro/con's:

Heightmap:
Pro: Easy collision detection, quite fast.
Con: No extravagant details, no cliffs etc.

Models:
Pro: Can be any shape
Con: Harder to collision detect.

Hope all goes well!

Cheers!

/Robert

EDIT:

nehe.gamedev.net <-- LOTS of excellent tutorials for basic/immediate and advance usage!

##### Share on other sites
You must know how a point given in 3D world is mapped to be displayed on the 2D screen. Verbosely, a vertex is a point where faces (usually triangles) of a mesh have a corner. Those vertices of a mesh are specified not in the worlds global co-ordinate system, but in a local co-ordinate system. This is senseful to be able to use the same mesh (i.e. vertices at the respective same vertex positions when seen in local co-ordinates) regardless where in the world the mesh as a whole is located. However, when rendering, the mesh actually appears at a global position, and hence the vertices _are_ to be transformed from the local to the global space. This is done on-the-fly by what OpenGL calls the Model transformation.

Next, you as the observer of the scene be located also somewhere in the global space. If you move you see another cut of the world. So, what you have to display ("what you see") doesn't depend only on the placement of meshes but also of you (most often expressed as a camera). This is what OpenGL calls the View transformation.

Still being in 3D, now its time to map the cut scene onto the screen. This can be done (you already know that) in different modes, mainly in perspective or orthogonal projection.

So coarsly you have to transform a vertex like this:
local (3D) -> global (3D) -> view/camera (3D) -> screen (2D)

In OpenGL you assemble the part local (3D) -> global (3D) -> view (3D) in the so-called MODELVIEW matrix (in D3D these both parts are more distinct), and the part view (3D) -> screen (2D) in the so-called PROJECTION matrix.

Both of these matrices are seperated by the matrix mode. The PROJECTION matrix is normally set up once and used throughout the whole lifetime of the application. The MODELVIEW matrix is normally set up per frame, because the camera stuff is changed all the way.

OpenGL is normally in MODELVIEW matrix mode. Switching is to done by yourself. So somewhere at the (logical) beginnging of the app you switch to the PROJECTION matrix, set it up, switch back to the MODELVIEW matrix, and begin with your game loop. One of the parts of the game loop is setting up the MODELVIEW matrix. First the VIEW portion as a very first part of rendering, then the MODEL portion appropriately for each mesh.

Notice that the order in which the transformations are assembled by their particular matrices (for position, orientation, and perhaps scaling) is a very important thing!

##### Share on other sites
Thank you very much haegarr... I think the explaination of the matricies and their order has really helped me out a bunch. A few things I didn't understand are starting to click together.

One thing I am still missing as this relates to a tile map. I have read about "stretching a texture accross a mesh". I would think this involves calulating texture coordinates to place with all of the verticies. Not that it is really all that difficult to figure out programaticly I am just wondering if there is something I am missing that makes this part a bit easier?

Also, setting up glFrustum... I can get it to work using gluPerspective (ala NeHe) but I would like to know how to get this to work... Any ideas or suggestions that explain in depth? I have searched google results extensively but as I said in my original post, many tutorials suggest that I have some deeper understanding of 3d concepts.

I can follow along in the NeHe tutorials but I feel like I am copy, pasting and not really understanding why some things work the way they do.

1. 1
2. 2
Rutin
21
3. 3
4. 4
frob
15
5. 5

• 9
• 13
• 9
• 33
• 13
• ### Forum Statistics

• Total Topics
632593
• Total Posts
3007282

×