Drawing Other Objects in the Scene in Word

Printer DataMatrix in Word Drawing Other Objects in the Scene

Without any objects inside the cube, your scene does not look too nice. When you load other objects into the scene, you have to consider something else: the size of the skybox. How can you make your skybox large enough so it contains all your objects but still have it be inside a reasonable distance so it will not be clipped away by the Projection matrix In this case, size really doesn t matter. You will draw the skybox at any size you want, but before anything else and while making sure XNA doesn t write to the depth buffer. This way, anything else you render later will be rendered to the scene as if the skybox weren t even there. This is how it works: whether you draw a box with lowest point (-1,-1,-1) and highest point (1,1,1), or with points (-100,-100,-100) and (100,100,100), they will both look the same if you put the camera in the middle. So, you re safe to simply draw the skybox at the size it comes in the Model file. But if the box is small, it will hide all your objects behind it, right That s why you are going to disable writing to the depth buffer while rendering the skybox. This way, after drawing the skybox as the first object, the depth buffer will still be empty! So, all objects drawn later will be rendered, not knowing the skybox has already been drawn, and the pixels corresponding to the object will get the color of the objects. For more information on the z-buffer, see the last paragraph of recipe 2-1. So, you need to add this extra line before drawing the skybox: device.RenderState.DepthBufferWriteEnable = false; Make sure you turn the z-buffer back on after you ve drawn the skybox, or your whole scene will be mixed up: device.RenderState.DepthBufferWriteEnable = true;

Loading a skybox from a file doesn t give you full control over all its aspects. For example, the skybox loaded earlier in this chapter uses six different textures, which causes the model to be split up into six different meshes, all of which require a different call to DrawPrimitives just to draw two

triangles. If there s one thing that slows down an XNA application, it s a lot of DrawPrimitives calls that each render a small amount of triangles (see recipe 3-4 and recipe 3-11 in 3). Although six calls extra probably won t make a huge difference, since you will now define the vertices of the skybox yourself, you will make sure the whole skybox uses only one texture so the whole skybox can be drawn in one DrawPrimitives call. Instead of using a Texture2D to store such a texture, you will use a TextureCube. This simply is a 2D image, composed of the six images of the walls of the skybox, and it looks like Figure 2-9.

Figure 2-9. Six sides of a skybox combined into one image Add variables to hold TextureCube, Effect, and VertexBuffer to your project: VertexBuffer skyboxVertexBuffer; TextureCube skyboxTexture; Effect skyboxEffect; And fill the effect and texture in the LoadContent method: effect = content.Load<Effect>("skyboxsample"); skyboxTexture = content.Load<TextureCube>("skyboxtexture"); You will create the skyboxsample effect file at the end of this recipe.

DirectX SDK. After installing the SDK, you should find a shortcut to this utility in your Start menu. You can simply select File New and then choose Cubemap Texture. Next, you can select which of the six sides is displayed by selecting the side from the View Cube Map Face menu. Alternatively, you can define the contents of the sides during runtime. For more information on this, see recipe 3-7 in 3.

By defining each vertex manually, it is possible to draw a skybox using a single DrawPrimitives call. Of course, the vertices need to contain Position data, and because you ll want the triangles to be covered with the skybox texture, you want to add texture coordinate information to the vertices. However, you ll be using the texCUBE HLSL intrinsic function to sample the color data from TextureCube, and it doesn t need any texture coordinate data. You ll learn more details about the texCUBE intrinsic later in this recipe. The XNA Framework, however, does not come with a vertex format that can store only a position. You could of course simply use a more complex vertex format, like the VertexPositionColor struct, and leave the color data empty. This will, however, still cause this color data to be sent to your graphics card, which simply wastes some bandwidth. To keep things clean, you ll create the simplest vertex format possible, one that holds only positional data. See recipe 5-14 in 5 for how to define your own vertex formats: public struct VertexPosition { public Vector3 Position; public VertexPosition(Vector3 position) { this.Position = position; } public static readonly VertexElement[] VertexElements = { new VertexElement( 0, 0, VertexElementFormat.Vector3, VertexElementMethod.Default, VertexElementUsage.Position, 0 ) }; public static readonly int SizeInBytes = sizeof(float) * 3; } This struct defines a custom vertex format, capable of storing only one Vector3 that will hold the Position data. When defining the vertices of the skybox, it is necessary to take the winding order of the triangles into account or some faces will be culled away. To learn more about culling, see recipe 5-6 in 5. When the camera is inside the skybox, none of the faces must be culled away, so all triangles must be wound in clockwise order, as seen by the camera inside the box. To make things easier, I have given the eight corner points of the cube a name, relative to the camera in the center of the cube: