OpenGL in 2D
Author
I am trying to write a 2D opengl program
But when i draw a triangle and then i want to use a bigger square to cover it. It doesn't work.
The triangle displays in front of the square.
Who can the program display the 2d objects(which have the same z-coordinate) according to the sequence of drawing in 2D?
Thx
The problem you are describing has to do with the current depth buffer settings.
Right now you might have code that sets up the depth buffer like
If you perform depth testing, your drawing order won't matter since the objects are sorted by the depth. Instead, you should not enable depth testing so the objects are drawn in the order they are coded. Here is a complete example using SFML with OpenGL.
Here's the program's output as well as a second output when taking out the green rectangle.
Right now you might have code that sets up the depth buffer like
glEnable(GL_DEPTH_TEST);glDepthFunc(GL_LESS);glClearDepth(1.f);If you perform depth testing, your drawing order won't matter since the objects are sorted by the depth. Instead, you should not enable depth testing so the objects are drawn in the order they are coded. Here is a complete example using SFML with OpenGL.
////////////////////////////////////////////////////////////// Headers////////////////////////////////////////////////////////////#include <SFML/Graphics.hpp>#include <iostream>#pragma comment(lib, "sfml-graphics-s-d.lib")#pragma comment(lib, "sfml-window-s-d.lib")#pragma comment(lib, "sfml-system-s-d.lib")#pragma comment(lib, "glu32.lib")// http://www.gamedev.net/community/forums/topic.asp?topic_id=104791void glEnable2D(){ int vPort[4]; glGetIntegerv(GL_VIEWPORT, vPort); glMatrixMode(GL_PROJECTION); glPushMatrix(); glLoadIdentity(); glOrtho(0, vPort[2], 0, vPort[3], -1, 1); glMatrixMode(GL_MODELVIEW); glPushMatrix(); glLoadIdentity();}// http://www.gamedev.net/community/forums/topic.asp?topic_id=104791void glDisable2D(){ glMatrixMode(GL_PROJECTION); glPopMatrix(); glMatrixMode(GL_MODELVIEW); glPopMatrix(); }/////////////////////////////////////////////////////////////// Entry point of application////// \return Application exit code///////////////////////////////////////////////////////////////int main(){ // Create main window sf::RenderWindow App(sf::VideoMode(800, 600), "SFML OpenGL"); App.PreserveOpenGLStates(true); // Enable Z-buffer read and write //glEnable(GL_DEPTH_TEST); //glDepthFunc(GL_LESS); //glClearDepth(1.f); // Start game loop while (App.IsOpened()) { // Process events sf::Event Event; while (App.GetEvent(Event)) { // Close window : exit if (Event.Type == sf::Event::Closed) App.Close(); // Escape key : exit if ((Event.Type == sf::Event::KeyPressed) && (Event.Key.Code == sf::Key::Escape)) App.Close(); // Adjust the viewport when the window is resized if (Event.Type == sf::Event::Resized) glViewport(0, 0, Event.Size.Width, Event.Size.Height); } glClear(GL_COLOR_BUFFER_BIT| GL_DEPTH_BUFFER_BIT); glLoadIdentity(); glEnable2D(); // The red triangle is drawn first at an offset so we can see // the green rectangle draws on top of it glTranslated(0, 75, 0); glBegin(GL_TRIANGLES); glColor3ub(255, 0, 0); glVertex2d(0, 0); glColor3ub(255, 0, 0); glVertex2d(100, 0); glColor3ub(255, 0, 0); glVertex2d(50, 50); glEnd(); glTranslated(0, -75, 0); // The green rectangle will cover the red triangle glBegin(GL_QUADS); glColor3ub(0, 255, 0); glVertex2d(0, 0); glColor3ub(0, 255, 0); glVertex2d(100, 0); glColor3ub(0, 255, 0); glVertex2d(100, 100); glColor3ub(0, 255, 0); glVertex2d(0, 100); glEnd(); // We should see this one on top of the green rectangle glTranslated(50, 0, 0); glBegin(GL_TRIANGLES); glColor3ub(0, 0, 255); glVertex2d(0, 0); glColor3ub(0, 0, 255); glVertex2d(100, 0); glColor3ub(0, 0, 255); glVertex2d(50, 50); glEnd(); glDisable2D(); // Finally, display the rendered frame on screen App.Display(); } return EXIT_SUCCESS;}Here's the program's output as well as a second output when taking out the green rectangle.
Author
Quote:Original post by Drew_Benton
The problem you are describing has to do with the current depth buffer settings.
Right now you might have code that sets up the depth buffer likeglEnable(GL_DEPTH_TEST);glDepthFunc(GL_LESS);glClearDepth(1.f);
If you perform depth testing, your drawing order won't matter since the objects are sorted by the depth. Instead, you should not enable depth testing so the objects are drawn in the order they are coded. Here is a complete example using SFML with OpenGL.
*** Source Snippet Removed ***
Here's the program's output as well as a second output when taking out the green rectangle.
thx~it works
But Can i work with object picking(glRenderMode(GL_SELECT);) when depth test turn off?
Quote:Original post by gohkgohk
But Can i work with object picking(glRenderMode(GL_SELECT);) when depth test turn off?
Ah, an unfortunate side affect for my simple solution [lol]
Here's the problem I see with that technique. If two objects have the same depth, which one is the closer one? The result should be undefined, because which one draws last is irrelevant to OpenGL since depth is sorted by the Z value. (OpenGL gurus correct me if I am mistaken).
When you have your picking logic, which I would assume looks like this or this, you could depth test only for the picking mode, but you are left with the problem you first saw visually, the triangle is on top of the square. When you went to click on the square, there's a good chance you are actually picking the triangle because that was the object that had the closest depth, so to speak.
Instead, I think you should be implementing a custom solution that you track the positions of your objects as well as an incrementing draw index. This way, when the user clicks on the screen, you take the mouse click coords, let's say 100, 100, then find if that point is on the inside of any of your objects. If it is, you track the object with the largest draw index, which would represent the object "on top" and that's the object you process.
Alternatively, you can do this, keep depth testing, but set the Z index with the "draw index" so you can "properly" use OpenGL. I.e.
glTranslated(0, 0, .1)... Obj 1glTranslated(0, 0, .2)... Obj 2glTranslated(0, 0, .3)... Obj 3etc...Those are my two best ideas currently, maybe someone else has an idea [smile]
Quote:Original post by Drew_BentonQuote:Original post by gohkgohk
But Can i work with object picking(glRenderMode(GL_SELECT);) when depth test turn off?
Ah, an unfortunate side affect for my simple solution [lol]
Here's the problem I see with that technique. If two objects have the same depth, which one is the closer one? The result should be undefined, because which one draws last is irrelevant to OpenGL since depth is sorted by the Z value. (OpenGL gurus correct me if I am mistaken).
The depth buffer is irrelevant to picking. The built in picking method is completely geometry based, and hits are recorded based on whether the objects intersects the clip volume or not. Therefore, hits are recorded before rasterization, and before the depth buffer has anything to say.
Relative depth order is also irrelevant for picking if it's inside the clip volume. Along with what objects that actually made a hit record, you also find the maximum and minimum depth values for that hit record. So if you have two objects at exactly the same depth, you will just have two hit record with exactly the same depth.
The hit record is also recorded in order of drawing, so if you need closest hit, or something else, it's up to you to use the depth information in the hit record to select the desired hit.
So to conclude, since picking occurs even before rasterization, depth and depth buffer settings have absolutely no influence on picking.
Author
Quote:Original post by Brother BobQuote:Original post by Drew_BentonQuote:Original post by gohkgohk
But Can i work with object picking(glRenderMode(GL_SELECT);) when depth test turn off?
Ah, an unfortunate side affect for my simple solution [lol]
Here's the problem I see with that technique. If two objects have the same depth, which one is the closer one? The result should be undefined, because which one draws last is irrelevant to OpenGL since depth is sorted by the Z value. (OpenGL gurus correct me if I am mistaken).
The depth buffer is irrelevant to picking. The built in picking method is completely geometry based, and hits are recorded based on whether the objects intersects the clip volume or not. Therefore, hits are recorded before rasterization, and before the depth buffer has anything to say.
Relative depth order is also irrelevant for picking if it's inside the clip volume. Along with what objects that actually made a hit record, you also find the maximum and minimum depth values for that hit record. So if you have two objects at exactly the same depth, you will just have two hit record with exactly the same depth.
The hit record is also recorded in order of drawing, so if you need closest hit, or something else, it's up to you to use the depth information in the hit record to select the desired hit.
So to conclude, since picking occurs even before rasterization, depth and depth buffer settings have absolutely no influence on picking.
Yes! Thx!
Now no need to sort the buffer too, because all have the same z.
that means the last one in the buffer is the latest one which i draw.
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
