I'm guessing by using only one camera you mean the minimap is a pre-rendered texture that you are placing images onto at certain locations? I'm going to assume this is the method.

• First place your level's top left corner at the world origin (makes things simpler). I will assume 1000x1000 units as a terrain size in this explanation.

• Then get the x,y coordinates of the object you want to display on the minimap and multiply that by difference between your map size and minimap texture dimensions resolution. So lets say your minimap image is 200x200. Your minimap coords should be their x,y positions in the scene multiplied by 0.2

• To get their final screen position, simply add the coordinate of the minimap to their x,y counterpart. So if your minimap is bottom right (say at 590,390), add that value to all the x,y positions.

• Use these co-ords to draw a 2DTexture image for each object on top of your minimap picture.

This should get you a basic, slightly old-skool image-based minimap for the whole level (scrolling would require a little more complexity, but not impossible). If you want the terrain 3D instead of an image, dual cameras is probably the better way to go.

I'm not sure why you can't use two cameras and place one on top of the terrain. You will use two cameras anyway, one to render the terrain and one for the mini map so you don't gain much benefit from recreating the terrain smaller. But, to make a red indicator, here is what I would do. Create a red sphere and turn lighting off. Now here's where the two possible camera methods that I talked about at the top come in.

If you attempt to do it your way, you may want to try using Renderer.bounds to get the boundaries of the terrain then set up a simple proportion.

  playerDistanceFromBoundary         indicatorDistanceFromBoundary /*unknown*/
      ------------------        =      --------------------------
     distanceAcrossTerrain                distanceAcrossMiniMap 

Then you could use that to find the x and y coordinates on the mini map. The simplified equation is playerDistanceFromBoundary / distanceAcrossTerrain * distanceAcrossMiniMap. After that, set the player's position to match, and you can do the same with the z if you need to.

var terrain : Transform; //Drag your terrain here in the inpsector
var miniMap : Transform; //""   ""   minimap here
var redDot : Transform; //The red icon;

private var terrainWidth : float;
private var terrainHeight : float;
private var miniWidth : float;
private var miniHeight : float;

private var scaleX : float;
private var scaleZ : float;

private var offset : Vector3;

//private var cachedOffset : float;
//private var cachedMapPos : Vector3;

function Start () {

     terrainWidth = terrain.collider.bounds.size.x; //not needed but adds to clarity
     terrainHeight = terrain.collider.bounds.size.z; //same


     miniWidth = miniMap.renderer.bounds.size.x; //see above
     miniHeight = miniMap.renderer.bounds.size.z;

     scaleX = miniWidth / terrainWidth; //find how much we need to scale the ball
     scaleZ = miniHeight / terrainHeight; // position

     //cachedOffset = miniMap.renderer.bounds.extents.y + redDot.renderer.bounds.extents.y;
     //cachedMapPos = Vector3(0, miniMap.renderer.bounds.extents.y, 0) + miniMap.renderer.bounds.min;
}

function Update () {
    offset = transform.position - terrain.position;
    offset.y = miniMap.renderer.bounds.extents.y + redDot.renderer.bounds.extents.y;
    //place the ball right ontop of the miniMap.  If you can cache this, delete this
    // and uncomment the line in start and the variable.
    //offset.y = cachedOffset

    offset.x *= scaleX; //scale the ball position
    offset.z *= scaleZ; //scale the ball position


    redDot.position = offset + Vector3(0, miniMap.renderer.bounds.extents.y, 0) + miniMap.renderer.bounds.min;
    //chache Vector3(0, miniMap.renderer.bounds.extents.y, 0) + miniMap.renderer.bounds.min if the map never moves.
    //redDot.position = offset + cachedMapPos;
}

This should work pretty well. It aligns the the red dot with the mini map proportionately with the player's position on the terrain. It also works with non square terrains. As you can see from the above. If you find any typos or bugs. Just post a comment, and I'll see what I'll correct them here.

The other way, the way where you have one terrain and two cameras looking at it, one from the player and one from above saves you all the scaling. You add a red sphere to the players's children directly on top of him (You could use a plane with a texture too). You turn off the lighting and such. Then put the player in a layer called Cam1Only and the ball in a layer called Cam2Only. On the player camera, turn the culling settings so that cam2Only is off, and in the mini map, turn it so cam1Only is turned off. Then, given that the ball is a child of the player, it will follow him everywhere and only it will show up on the mini map, while only the player will show up in the actual game.

This will give u a start : http://unity3d.com/support/documentation/Components/class-RenderTexture.html

The way I solved this issue was just to put another camera ontop of your level, and then creat a shape (Box, Uvvsphere, ect.) and put it directly over the players head, high enough so the player cannot see it and low enough so that it is under your second camera. Then make the cube/Sphere/whatever-shape-you-made the child of your player so that it follows the player. I know alot of you probably think this is cheating because of its simplicity but if it works and if it saves time than I say use it. Hope this helps!!

It was a very nice idea! ....can you tell me where is the red and gold colors...? I look forward your answer. thank you!