Skip to main content
gamedev.net
🔒 Locked

Slightly confusing Z value problem

Started by Endar Sep 2, 2005 at 11:17 PM 0 replies 700+ views
Original Post
Endar
Endar
I'm working on writing a simple graphics engine, and I've got the window creation and setup for OGL working fine, but drawing seems to be another matter.

while( engine.run() ){
	driver->beginScene();	// clears the window with glClear

	glLoadIdentity();

	glPointSize(5.0f);

	glBegin(GL_TRIANGLES);

		glColor3f(1.0f, 1.0f, 1.0f);
		glVertex3f(0,   0,   1);
		glVertex3f(50,  100, 1);
		glVertex3f(100, 50,  1);

	glEnd();

	glFlush();

	driver->endScene();	// actually renders the window
}



Here's what I'm trying to draw just to make sure that drawing works. I have this drawing, but something is confusing me. Why am I able to see it? I was under the impression that when you draw something at a positive z value, then it is drawn behind the camera, and a negative z value is in front of the camera. I mean, I haven't moved the camera or anything. Here's my OGL init code, and the beginScene and endScene functions:

/**
 * Initialises the video driver.
 * \param screenSize A dimension2d object containing the height and width of the screen
 */
bool COpenGLDriver::initDriver(const video::dimension2d& screenSize)
{
	int nPixelFormat;

	static PIXELFORMATDESCRIPTOR pfd = {
		sizeof(PIXELFORMATDESCRIPTOR),		// size of the structure
		1,									// version. Always set to 1
		PFD_DRAW_TO_WINDOW |				// support window
		PFD_SUPPORT_OPENGL |				// support opengl
		PFD_DOUBLEBUFFER,					// support double buffering
		PFD_TYPE_RGBA,						// RGBA color mode
		32,									// 32-bit color mode
		0, 0, 0, 0, 0, 0,					// ignore color bits, not used
		0,									// no alpha buffer
		0,									// ignore shift bit
		0,									// no accumulation buffer
		0, 0, 0, 0,							// ignore accumulation bits
		16,									// 16-bit z-buffer size
		0,									// no stencil buffer
		0,									// no auxillary buffer
		PFD_MAIN_PLANE,						// main drawing plane
		0,									// reserved
		0, 0, 0								// layer masks ignored
	};


	// if get hdc fails
	if( !(hDC = GetDC(Window)) ){
		util::Message::print("COpenGLDriver::initDriver: Cannot create a GL device context.");
		return false;
	}

	// chose best matching pixel format, return index
	if( !(nPixelFormat=ChoosePixelFormat(hDC, &pfd)) ){
		util::Message::print("COpenGLDriver::initDriver: Cannot find a suitable pixel format.");
		return false;
	}

	// set pixel format to device context
	if( !SetPixelFormat(hDC, nPixelFormat, &pfd) ){
		util::Message::print("COpenGLDriver::initDriver: Cannot set the pixel format.");
		return false;
	}

	// create rendering context
	if( !(hRC=wglCreateContext(hDC)) ){
		util::Message::print("COpenGLDriver::initDriver: Cannot create a GL rendering context.");
		return false;
	}

	// activate rendering context
	if( !wglMakeCurrent(hDC, hRC) ){
		util::Message::print("COpenGLDriver::initDriver: Cannot activate GL rendering context.");
		return false;
	}


	glViewport(0, 0, (int)screenSize.width, (int)screenSize.height);	// reset the current opengl viewport


	glShadeModel(GL_SMOOTH);		// enable smooth shading

	glDisable(GL_LIGHTING);			// lighting disabled
	glFrontFace(GL_CW);				//
	glEnable(GL_CULL_FACE);

	glHint(GL_PERSPECTIVE_CORRECTION_HINT, GL_NICEST);		// really nice perspective calculations

	glClearDepth(1.0f);			// set gl clear depth
	glEnable(GL_DEPTH_TEST);
	glDepthFunc(GL_LEQUAL);


	return true;	// init went well
}

/**
 * Does OGL housekeeping, like clearing the screen and depth buffer
 * \param c A CColor object that represents the color to clear the screen to. Defaults to black.
 */
void COpenGLDriver::beginScene(video::CColor c = video::CColor(0.0f,0.0f,0.0f))
{
	glClearColor(	c.getRedFloat(), c.getGreenFloat(),
					c.getBlueFloat(), c.getAlphaFloat()
				);						// Set clear color

	glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);// Clear Screen And Depth Buffer
}

/**
 * Does OGL housekeeping for the end of rendering a frame.
 */
void COpenGLDriver::endScene()
{
	SwapBuffers(hDC);	// render the scene to the window




Is it something that I've done in complete ignorance? I am able to see the triangle in a range of about -1 to +1. Any ideas? Because this doens't look different from other OGL init code that I've gotten from tutorials.
[size="2"][size=2]Mort, Duke of Sto Helit: NON TIMETIS MESSOR -- Don't Fear The Reaper
DrewGreen
DrewGreen
Yes, you are correct about the way z values work.
To nitpick, opengl doesn't actually think of the view in terms of a "camera". Think of it in terms of moving the object you're drawing towards (positive) or away from (negative) the viewer. You can always perform a translation first if you prefer to use positive numbers. Do it before your calls to glBegin ... glEnd and not during.

glTranslatef(0.0f,0.0f,-10.0f) or something. What that's doing in english is moving the object 10 units into the screen along the z axis (away from the camera).

You can only see the triangle in a short range onscreen because the viewing frustum (what you can see onscreen) is set up this way by default. You can alter this:

glViewport (0, 0, (GLsizei)(windowwidth), (GLsizei)(windowheight));	// Reset The Current ViewportglMatrixMode (GL_PROJECTION);	// Select The Projection MatrixglLoadIdentity ();	// Reset The Projection MatrixgluPerspective (45.0f, (GLfloat)(windowwidth)/(GLfloat)(windowheight),1.0f, 500.0f);	// Calculate The Aspect Ratio Of The WindowglMatrixMode (GL_MODELVIEW);	// Select The Modelview MatrixglLoadIdentity ();	// Reset The Modelview Matrix


Could be placed in the initialisation routine, although it would serve much better as part of a routine called when the window is resized as well. The key part is the gluPerspective call. (you'll need to #include ) 45.0f is the fov in the y direction (in degrees), and the 1.0f and 500.0f correspond to the front and back clipping planes of the view frustum, and are always positive values. In this example, you will only see things that are at least 1.0f in front of the "camera", and at most 500.0f units away from (still in front of) the camera. Try playing with these values.

When you rotate the camera around on the screen, you're actually rotating the objects around the camera (initially, but if you do translations followed by rotations this isn't true), but in any case it means that polygons with negative z values will be visible when you rotate them into view.

Hope that helps

Also, you're using the glVertex3f command, which expects coordinates specified using floats. Although it doesn't really matter that much, get into the habit of writing 1.0f instead of 1 (as that's an integer). Makes no difference to a human but it'll get rid of some compilation warnings (it's satisfying to see your code compile with 0 warnings & 0 errors :) )

i.e. glVertex3f(0.0f, 0.0f, 1.0f);
"I must not fear. Fear is the mindkiller. Fear is the little death that brings total obliteration. I will face my fear. I will permit it to pass over me and through me. And when it has gone past me I will turn to see fear's path. Where the fear has gone there will be nothing. Only I will remain." ~Frank Herbert, DuneMy slice of the web

Topic Locked

This topic has been locked by a moderator. New replies are not allowed.

Sign in to reply to this topic.