Texture MappingLearning how to texture map has many benefits. Lets say you wanted a missile to fly across the screen. Up until this tutorial we'd probably make the entire missile out of polygons, and fancy colors. With texture mapping, you can take a real picture of a missile and make the picture fly across the screen. Which do you think will look better? A photograph or an object made up of triangles and squares? By using texture mapping, not only will it look better, but your program will run faster. The texture mapped missile would only be one quad moving across the screen. A missile made out of polygons could be made up of hundreds or thousands of polygons. The single texture mapped quad will use alot less processing power. Important: This way of loading textures is deprecated and does not work anymore with current versions of Visual Studio. The theory of this tutorial is still valid though. An update of the code which is responsible for loading the texture can be found here: http://nehe.gamedev.net/tutorial/lesson_06_texturing_update/47002/
Lets start off by adding five new lines of code to the top of lesson one. The first new line is #include <stdio.h>. Adding this header file allows us to work with files. In order to use fopen() later in the code we need to include this line. Then we add three new floating point variables... xrot, yrot and zrot. These variables will be used to rotate the cube on the x axis, the y axis, and the z axis. The last line GLuint texture[1] sets aside storage space for one texture. If you wanted to load in more than one texture, you would change the number one to the number of textures you wish to load. #include <windows.h> // Header File For Windows #include <stdio.h> // Header File For Standard Input/Output ( NEW ) #include <gl\gl.h> // Header File For The OpenGL32 Library #include <gl\glu.h> // Header File For The GLu32 Library #include <gl\glaux.h> // Header File For The GLaux Library HDC hDC=NULL; // Private GDI Device Context HGLRC hRC=NULL; // Permanent Rendering Context HWND hWnd=NULL; // Holds Our Window Handle HINSTANCE hInstance; // Holds The Instance Of The Application bool keys[256]; // Array Used For The Keyboard Routine bool active=TRUE; // Window Active Flag bool fullscreen=TRUE; // Fullscreen Flag GLfloat xrot; // X Rotation ( NEW ) GLfloat yrot; // Y Rotation ( NEW ) GLfloat zrot; // Z Rotation ( NEW ) GLuint texture[1]; // Storage For One Texture ( NEW ) LRESULT CALLBACK WndProc(HWND, UINT, WPARAM, LPARAM); // Declaration For WndProc Now immediately after the above code, and before ReSizeGLScene(), we want to add the following section of code. The job of this code is to load in a bitmap file. If the file doesn't exist NULL is sent back meaning the texture couldn't be loaded. Before I start explaining the code there are a few VERY important things you need to know about the images you plan to use as textures. The image height and width MUST be a power of 2. The width and height must be at least 64 pixels, and for compatability reasons, shouldn't be more than 256 pixels. If the image you want to use is not 64, 128 or 256 pixels on the width or height, resize it in an art program. There are ways around this limitation, but for now we'll just stick to standard texture sizes. First thing we do is create a file handle. A handle is a value used to identify a resource so that our program can access it. We set the handle to NULL to start off. AUX_RGBImageRec *LoadBMP(char *Filename) // Loads A Bitmap Image { FILE *File=NULL; // File Handle Next we check to make sure that a filename was actually given. The person may have use LoadBMP() without specifying the file to load, so we have to check for this. We don't want to try loading nothing :) if (!Filename) // Make Sure A Filename Was Given { return NULL; // If Not Return NULL } If a filename was given, we check to see if the file exists. The line below tries to open the file. File=fopen(Filename,"r"); // Check To See If The File Exists If we were able to open the file it obviously exists. We close the file with fclose(File) then we return the image data. auxDIBImageLoad(Filename) reads in the data. if (File) // Does The File Exist? { fclose(File); // Close The Handle return auxDIBImageLoad(Filename); // Load The Bitmap And Return A Pointer } If we were unable to open the file we'll return NULL. which means the file couldn't be loaded. Later on in the program we'll check to see if the file was loaded. If it wasn't we'll quit the program with an error message. return NULL; // If Load Failed Return NULL } This is the section of code that loads the bitmap (calling the code above) and converts it into a texture. int LoadGLTextures() // Load Bitmaps And Convert To Textures { We'll set up a variable called Status. We'll use this variable to keep track of whether or not we were able to load the bitmap and build a texture. We set Status to FALSE (meaning nothing has been loaded or built) by default. int Status=FALSE; // Status Indicator Now we create an image record that we can store our bitmap in. The record will hold the bitmap width, height, and data. AUX_RGBImageRec *TextureImage[1]; // Create Storage Space For The Texture We clear the image record just to make sure it's empty. memset(TextureImage,0,sizeof(void *)*1); // Set The Pointer To NULL Now we load the bitmap and convert it to a texture. TextureImage[0]=LoadBMP("Data/NeHe.bmp") will jump to our LoadBMP() code. The file named NeHe.bmp in the Data directory will be loaded. If everything goes well, the image data is stored in TextureImage[0], Status is set to TRUE, and we start to build our texture. // Load The Bitmap, Check For Errors, If Bitmap's Not Found Quit if (TextureImage[0]=LoadBMP("Data/NeHe.bmp")) { Status=TRUE; // Set The Status To TRUE Now that we've loaded the image data into TextureImage[0], we will build a texture using this data. The first line glGenTextures(1, &texture[0]) tells OpenGL we want to generate one texture name (increase the number if you load more than one texture). Remember at the very beginning of this tutorial we created room for one texture with the line GLuint texture[1]. Although you'd think the first texture would be stored at &texture[1] instead of &texture[0], it's not. The first actual storage area is 0. If we wanted two textures we would use GLuint texture[2] and the second texture would be stored at texture[1]. The second line glBindTexture(GL_TEXTURE_2D, texture[0]) tells OpenGL to bind the named texture texture[0] to a texture target. 2D textures have both height (on the Y axes) and width (on the X axes). The main function of glBindTexture is to assign a texture name to texture data. In this case we're telling OpenGL there is memory available at &texture[0]. When we create the texture, it will be stored in the memory that &texture[0] references. glGenTextures(1, &texture[0]); // Create The Texture // Typical Texture Generation Using Data From The Bitmap glBindTexture(GL_TEXTURE_2D, texture[0]); Next we create the actual texture. The following line tells OpenGL the texture will be a 2D texture (GL_TEXTURE_2D). Zero represents the images level of detail, this is usually left at zero. Three is the number of data components. Because the image is made up of red data, green data and blue data, there are three components. TextureImage[0]->sizeX is the width of the texture. If you know the width, you can put it here, but it's easier to let the computer figure it out for you. TextureImage[0]->sizey is the height of the texture. zero is the border. It's usually left at zero. GL_RGB tells OpenGL the image data we are using is made up of red, green and blue data in that order. GL_UNSIGNED_BYTE means the data that makes up the image is made up of unsigned bytes, and finally... TextureImage[0]->data tells OpenGL where to get the texture data from. In this case it points to the data stored in the TextureImage[0] record. // Generate The Texture glTexImage2D(GL_TEXTURE_2D, 0, 3, TextureImage[0]->sizeX, TextureImage[0]->sizeY, 0, GL_RGB, GL_UNSIGNED_BYTE, TextureImage[0]->data); The next two lines tell OpenGL what type of filtering to use when the image is larger (GL_TEXTURE_MAG_FILTER) or stretched on the screen than the original texture, or when it's smaller (GL_TEXTURE_MIN_FILTER) on the screen than the actual texture. I usually use GL_LINEAR for both. This makes the texture look smooth way in the distance, and when it's up close to the screen. Using GL_LINEAR requires alot of work from the processor/video card, so if your system is slow, you might want to use GL_NEAREST. A texture that's filtered with GL_NEAREST will appear blocky when it's stretched. You can also try a combination of both. Make it filter things up close, but not things in the distance. glTexParameteri(GL_TEXTURE_2D,GL_TEXTURE_MIN_FILTER,GL_LINEAR); // Linear Filtering glTexParameteri(GL_TEXTURE_2D,GL_TEXTURE_MAG_FILTER,GL_LINEAR); // Linear Filtering } Now we free up any ram that we may have used to store the bitmap data. We check to see if the bitmap data was stored in TextureImage[0]. If it was we check to see if the data has been stored. If data was stored, we erase it. Then we free the image structure making sure any used memory is freed up. if (TextureImage[0]) // If Texture Exists { if (TextureImage[0]->data) // If Texture Image Exists { free(TextureImage[0]->data); // Free The Texture Image Memory } free(TextureImage[0]); // Free The Image Structure } Finally we return the status. If everything went OK, the variable Status will be TRUE. If anything went wrong, Status will be FALSE. return Status; // Return The Status } I've added a few lines of code to InitGL. I'll repost the entire section of code, so it's easy to see the lines that I've added, and where they go in the code. The first line if (!LoadGLTextures()) jumps to the routine above which loads the bitmap and makes a texture from it. If LoadGLTextures() fails for any reason, the next line of code will return FALSE. If everything went OK, and the texture was created, we enable 2D texture mapping. If you forget to enable texture mapping your object will usually appear solid white, which is definitely not good. int InitGL(GLvoid) // All Setup For OpenGL Goes Here { if (!LoadGLTextures()) // Jump To Texture Loading Routine ( NEW ) { return FALSE; // If Texture Didn't Load Return FALSE ( NEW ) } glEnable(GL_TEXTURE_2D); // Enable Texture Mapping ( NEW ) glShadeModel(GL_SMOOTH); // Enable Smooth Shading glClearColor(0.0f, 0.0f, 0.0f, 0.5f); // Black Background glClearDepth(1.0f); // Depth Buffer Setup glEnable(GL_DEPTH_TEST); // Enables Depth Testing glDepthFunc(GL_LEQUAL); // The Type Of Depth Testing To Do glHint(GL_PERSPECTIVE_CORRECTION_HINT, GL_NICEST); // Really Nice Perspective Calculations return TRUE; // Initialization Went OK } Now we draw the textured cube. You can replace the DrawGLScene code with the code below, or you can add the new code to the original lesson one code. This section will be heavily commented so it's easy to understand. The first two lines of code glClear() and glLoadIdentity() are in the original lesson one code. glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT) will clear the screen to the color we selected in InitGL(). In this case, the screen will be cleared to black. The depth buffer will also be cleared. The view will then be reset with glLoadIdentity(). int DrawGLScene(GLvoid) // Here's Where We Do All The Drawing { glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); // Clear Screen And Depth Buffer glLoadIdentity(); // Reset The Current Matrix glTranslatef(0.0f,0.0f,-5.0f); // Move Into The Screen 5 Units The following three lines of code will rotate the cube on the x axis, then the y axis, and finally the z axis. How much it rotates on each axis will depend on the value stored in xrot, yrot and zrot. glRotatef(xrot,1.0f,0.0f,0.0f); // Rotate On The X Axis glRotatef(yrot,0.0f,1.0f,0.0f); // Rotate On The Y Axis glRotatef(zrot,0.0f,0.0f,1.0f); // Rotate On The Z Axis The next line of code selects which texture we want to use. If there was more than one texture you wanted to use in your scene, you would select the texture using glBindTexture(GL_TEXTURE_2D, texture[number of texture to use]). If you wanted to change textures, you would bind to the new texture. One thing to note is that you can NOT bind a texture inside glBegin() and glEnd(), you have to do it before or after glBegin(). Notice how we use glBindTextures to specify which texture to create and to select a specific texture. glBindTexture(GL_TEXTURE_2D, texture[0]); // Select Our Texture To properly map a texture onto a quad, you have to make sure the top right of the texture is mapped to the top right of the quad. The top left of the texture is mapped to the top left of the quad, the bottom right of the texture is mapped to the bottom right of the quad, and finally, the bottom left of the texture is mapped to the bottom left of the quad. If the corners of the texture do not match the same corners of the quad, the image may appear upside down, sideways, or not at all. The first value of glTexCoord2f is the X coordinate. 0.0f is the left side of the texture. 0.5f is the middle of the texture, and 1.0f is the right side of the texture. The second value of glTexCoord2f is the Y coordinate. 0.0f is the bottom of the texture. 0.5f is the middle of the texture, and 1.0f is the top of the texture. So now we know the top left coordinate of a texture is 0.0f on X and 1.0f on Y, and the top left vertex of a quad is -1.0f on X, and 1.0f on Y. Now all you have to do is match the other three texture coordinates up with the remaining three corners of the quad. Try playing around with the x and y values of glTexCoord2f. Changing 1.0f to 0.5f will only draw the left half of a texture from 0.0f (left) to 0.5f (middle of the texture). Changing 0.0f to 0.5f will only draw the right half of a texture from 0.5f (middle) to 1.0f (right). glBegin(GL_QUADS); // Front Face glTexCoord2f(0.0f, 0.0f); glVertex3f(-1.0f, -1.0f, 1.0f); // Bottom Left Of The Texture and Quad glTexCoord2f(1.0f, 0.0f); glVertex3f( 1.0f, -1.0f, 1.0f); // Bottom Right Of The Texture and Quad glTexCoord2f(1.0f, 1.0f); glVertex3f( 1.0f, 1.0f, 1.0f); // Top Right Of The Texture and Quad glTexCoord2f(0.0f, 1.0f); glVertex3f(-1.0f, 1.0f, 1.0f); // Top Left Of The Texture and Quad // Back Face glTexCoord2f(1.0f, 0.0f); glVertex3f(-1.0f, -1.0f, -1.0f); // Bottom Right Of The Texture and Quad glTexCoord2f(1.0f, 1.0f); glVertex3f(-1.0f, 1.0f, -1.0f); // Top Right Of The Texture and Quad glTexCoord2f(0.0f, 1.0f); glVertex3f( 1.0f, 1.0f, -1.0f); // Top Left Of The Texture and Quad glTexCoord2f(0.0f, 0.0f); glVertex3f( 1.0f, -1.0f, -1.0f); // Bottom Left Of The Texture and Quad // Top Face glTexCoord2f(0.0f, 1.0f); glVertex3f(-1.0f, 1.0f, -1.0f); // Top Left Of The Texture and Quad glTexCoord2f(0.0f, 0.0f); glVertex3f(-1.0f, 1.0f, 1.0f); // Bottom Left Of The Texture and Quad glTexCoord2f(1.0f, 0.0f); glVertex3f( 1.0f, 1.0f, 1.0f); // Bottom Right Of The Texture and Quad glTexCoord2f(1.0f, 1.0f); glVertex3f( 1.0f, 1.0f, -1.0f); // Top Right Of The Texture and Quad // Bottom Face glTexCoord2f(1.0f, 1.0f); glVertex3f(-1.0f, -1.0f, -1.0f); // Top Right Of The Texture and Quad glTexCoord2f(0.0f, 1.0f); glVertex3f( 1.0f, -1.0f, -1.0f); // Top Left Of The Texture and Quad glTexCoord2f(0.0f, 0.0f); glVertex3f( 1.0f, -1.0f, 1.0f); // Bottom Left Of The Texture and Quad glTexCoord2f(1.0f, 0.0f); glVertex3f(-1.0f, -1.0f, 1.0f); // Bottom Right Of The Texture and Quad // Right face glTexCoord2f(1.0f, 0.0f); glVertex3f( 1.0f, -1.0f, -1.0f); // Bottom Right Of The Texture and Quad glTexCoord2f(1.0f, 1.0f); glVertex3f( 1.0f, 1.0f, -1.0f); // Top Right Of The Texture and Quad glTexCoord2f(0.0f, 1.0f); glVertex3f( 1.0f, 1.0f, 1.0f); // Top Left Of The Texture and Quad glTexCoord2f(0.0f, 0.0f); glVertex3f( 1.0f, -1.0f, 1.0f); // Bottom Left Of The Texture and Quad // Left Face glTexCoord2f(0.0f, 0.0f); glVertex3f(-1.0f, -1.0f, -1.0f); // Bottom Left Of The Texture and Quad glTexCoord2f(1.0f, 0.0f); glVertex3f(-1.0f, -1.0f, 1.0f); // Bottom Right Of The Texture and Quad glTexCoord2f(1.0f, 1.0f); glVertex3f(-1.0f, 1.0f, 1.0f); // Top Right Of The Texture and Quad glTexCoord2f(0.0f, 1.0f); glVertex3f(-1.0f, 1.0f, -1.0f); // Top Left Of The Texture and Quad glEnd(); Now we increase the value of xrot, yrot and zrot. Try changing the number each variable increases by to make the cube spin faster or slower, or try changing a + to a - to make the cube spin the other direction. xrot+=0.3f; // X Axis Rotation yrot+=0.2f; // Y Axis Rotation zrot+=0.4f; // Z Axis Rotation return true; // Keep Going } You should now have a better understanding of texture mapping. You should be able to texture map the surface of any quad with an image of your choice. Once you feel confident with your understanding of 2D texture mapping, try adding six different textures to the cube. Texture mapping isn't to difficult to understand once you understand texture coordinates. If you're having problems understanding any part of this tutorial, let me know. Either I'll rewrite that section of the tutorial, or I'll reply back to you in email. Have fun creating texture mapped scenes of your own :) Jeff Molofee (NeHe) * DOWNLOAD Visual C++ Code For This Lesson. * DOWNLOAD Borland C++ Builder 6 Code For This Lesson. ( Conversion by Christian Kindahl )
|