Apply transformations to Bounding Box - c++

I am trying to make a tank game. I have successfully loaded an OBJ model, and calculated its bounding box for the model at the origin.
I am now trying to apply the transformations done to my model in the game logic to the original coordinates for the bounding box. For this, I grab the modelview matrix right before drawing my model, then multiply this matrix for the two vectors that define the BBox.
Here is the code that draws my tank:
void drawTank()
{
bBox = calcBBox(modelo, 1);
glPushMatrix();
glBindTexture(GL_TEXTURE_2D, texTank);
glScalef(0.2, 0.2, 0.2);
glTranslatef(posTank.x,posTank.y,posTank.z);
glRotatef(anguloTanque, 0, 1, 0); // rotate around Y (horizontal)
glRotatef(90, 0, 1, 0);
glRotatef(-90, 1, 0, 0);
glGetFloatv(GL_MODELVIEW_MATRIX, matrix);
glmDraw(modelo, GLM_TEXTURE | GLM_MATERIAL);
glColor3f(1,0,0);
drawBBox(bBox);
glPopMatrix();
}
With this snippet, my bbox is properly drawn over the tank model (transformations are applied in rendering by the glTranslate & glRotate functions). As you can see I also grab here my ModelView matrix.
Then I apply this matrix as follows (this is my entire display function):
void Display(void) {
// Clear the window with current clearing color
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);
glMatrixMode(GL_MODELVIEW);
glLoadIdentity();
glPushMatrix();
camera();
glEnable(GL_TEXTURE_2D);
//glTranslatef(0,-40,150);
//PLANE
glBindTexture(GL_TEXTURE_2D, texArena);
glBegin(GL_POLYGON);
glTexCoord2f( 0.0f, 0.0f );
glVertex3f(-500, 0, -500);
glTexCoord2f( 5.0f, 5.0f );
glVertex3f(500, 0, -500);
glTexCoord2f(5.0f, 0.0f );
glVertex3f(500, 0, 500);
glTexCoord2f( 0.0f, 5.0f );
glVertex3f(-500, 0, 500);
glEnd();
drawTank();
glPopMatrix();
point3D max = bBox.max;
point3D min = bBox.min;
point3D resultMax;
point3D resultMin;
//Transformacion
multVectorByMatrix(matrix, max, resultMax);
multVectorByMatrix(matrix, min, resultMin);
bBox.max.x = resultMax.x; bBox.max.y = resultMax.y; bBox.max.z = resultMax.z;
bBox.min.x = resultMin.x; bBox.min.y = resultMin.y; bBox.min.z = resultMin.z;
glPushMatrix();
glColor3f(1,1,1);
drawBBox(bBox);
glPopMatrix();
glFlush();
glutSwapBuffers();
}
The function that multiplies a vector by a matrix:
void multVectorByMatrix(float* matrix, point3D vector, point3D &result)
{
result.x = (matrix[0] * vector.x) +
(matrix[4] * vector.y) +
(matrix[8] * vector.z) +
matrix[12];
result.y = (matrix[1] * vector.x) +
(matrix[5] * vector.y) +
(matrix[9] * vector.z) +
matrix[13];
result.z = (matrix[2] * vector.x) +
(matrix[6] * vector.y) +
(matrix[10] * vector.z) +
matrix[14];
}
If I draw the bounding box with this render loop, then my bounding box gets drawn but transformations are not applied properly. I can see the bounding box moving correctly with translations, but rotations are not done right.
What might be the problem here?
edit: some screenshots

Your problem is in this code.
point3D max = bBox.max;
point3D min = bBox.min;
point3D resultMax;
point3D resultMin;
//Transformacion
multVectorByMatrix(matrix, max, resultMax);
multVectorByMatrix(matrix, min, resultMin);
bBox.max.x = resultMax.x; bBox.max.y = resultMax.y; bBox.max.z = resultMax.z;
bBox.min.x = resultMin.x; bBox.min.y = resultMin.y; bBox.min.z = resultMin.z;
glPushMatrix();
glColor3f(1,1,1);
drawBBox(bBox);
glPopMatrix();
You take two vertices from your box and then apply transformations to them, then you use this transformed vertices to display a box, which of course will be axis aligned, because that's the only box you can get from just two opposite vertices. And you can see on your screenshot, that you bbox and the correct bbox have common vertices - these are exactly the vertices you applied your transformations to. So, in order to get a correct bbox, you need to get all vertices of the bbox and apply these transformations to all of them. Then you'll get exactly what you want.

Related

How to change the view perspective in OpenGL?

I draw many lines to form a grid. I want to see the grid rotated on its X-axis, but I never get the intended result. I tried glRotatef and gluLookAt which does not work the way I want. Please see the pictures below.
this is the grid
this is how I want to see it
Edit: geez, posting the code here is also hard, lol, anyway here it is.
Edit2: removed, only leave the code that has issues.
Please find the code below, no matter how I set the gluLookAt, the grid result won't be in the perspective I want.
#include <GL/glut.h>
void display() {
...
glClear(GL_COLOR_BUFFER_BIT);
glBegin(GL_LINES);
for (int i = 0; i < 720; i += 3)
{
glColor3f(0, 1, 1);
glVertex3f(linePoints[i], linePoints[i + 1], linePoints[i + 2]);
}
glEnd();
glFlush();
}
void init() {
glClearColor(0.0, 0.0, 0.0, 1.0);
glColor3f(1.0, 1.0, 1.0);
glMatrixMode(GL_PROJECTION);
glLoadIdentity();
gluPerspective(60.0, 4.0 / 3.0, 1, 40);
glMatrixMode(GL_MODELVIEW);
glLoadIdentity();
gluLookAt(0, -2, 1.25, 0, 0, 0, 0, 1, 0);
}
Lets assume, that you have a grid in the xy plane of the world:
glColor3f(0, 1, 1);
glBegin(GL_LINES);
for (int i = 0; i <= 10; i ++)
{
// horizontal
glVertex3f(-50.0f + i*10.0f, -50.0f, 0.0f);
glVertex3f(-50.0f + i*10.0f, 50.0f, 0.0f);
// vertical
glVertex3f(-50.0f, -50.0f + i*10.0f, 0.0f);
glVertex3f( 50.0f, -50.0f + i*10.0f, 0.0f);
}
glEnd();
Ensure that the distance of to the far plane of the projection is large enough (see gluPerspective). All the geometry which is not in between the near an far plane of the Viewing frustum is clipped.
Further more ensure that the aspect ratio (4.0 / 3.0) match the ratio of the viewport rectangle (window).
glMatrixMode(GL_PROJECTION);
glLoadIdentity();
gluPerspective(60.0, 4.0 / 3.0, 1, 200);
For the use of gluLookAt, the up vector of the view has to be perpendicular to the grid. If the grid is arranged parallel to the xy plane, then the up vector is z axis (0, 0, 1).
The target (center) is the center of the grid (0, 0, 0).
The point of view (eye position) is ought to be above and in front of the grid, for instance (0, -55, 50). Note the point of view is used for a grid with the bottom left of (-50, -50, 0) and a top right of (50, 50, 0).
glMatrixMode(GL_MODELVIEW);
glLoadIdentity();
gluLookAt(0, -55.0, 50.0, 0, 0, 0, 0, 0, 1);

Absolute scale - Rotating scale vector

Need to get global absolute scale of object to display it with OpenGL.
When I just multiply scale vectors of parents and children, I get absolute scale for world axes space. When I rotate object, it scales along global axes, not local.
I decided that I need to rotate scale vector too. But:
When I am trying to rotate it with direction vector - it values sometimes are going to 0, and scale too.
{ scale.x * forward.x , scale.y * forward.y , scale.z * forward.z }
When I am trying to rotate it with glm::rotate, it makes unexpected results, like infinite rotating/scaling, wrench and other effects on meshes.
auto globalScale = glm::vec3(scale.x, scale.y, scale.z);
globalScale = glm::rotate(globalScale, rotation.x, {1,0,0});
globalScale = glm::rotate(globalScale, rotation.y, {0,1,0});
globalScale = glm::rotate(globalScale, rotation.z, {0,0,1});
My rendering code:
void Render(Material *mat, Transform* tr){
glEnable(GL_COLOR_MATERIAL);
glEnable (GL_LIGHTING);
glEnable (GL_LIGHT0);
// Get object transformations
Vector3 pos = tr->globalPosition();
Vector3 rot = tr->globalRotation();
Vector3 scale = (tr->globalScale());
auto globalScale = glm::vec3(scale.x, scale.y, scale.z);
// First, scaling, then rotating, then translating in world space
// ( Initially all objects rendering starts at (0,0,0) )
glScaled(globalScale.x, globalScale.y, globalScale.z);
glRotatef(rot.x, 1.0, 0.0, 0.0);
glRotatef(rot.y, 0.0, 1.0, 0.0);
glRotatef(rot.z, 0.0, 0.0, 1.0);
glTranslated(pos.x, pos.y, pos.z);
// Rendering
glEnableClientState(GL_VERTEX_ARRAY);
glEnableClientState(GL_NORMAL_ARRAY);
glEnableClientState(GL_TEXTURE_COORD_ARRAY);
glVertexPointer(3,GL_FLOAT,0,vertexArray);
glNormalPointer(GL_FLOAT,0,normalArray);
glClientActiveTexture(GL_TEXTURE0_ARB);
glTexCoordPointer(2,GL_FLOAT,0,uvArray);
glDrawArrays(GL_TRIANGLES,0,numVerts);
glDisableClientState(GL_VERTEX_ARRAY);
glDisableClientState(GL_NORMAL_ARRAY);
glDisableClientState(GL_TEXTURE_COORD_ARRAY);
// Rolling transformations back
glTranslated(-pos.x, -pos.y, -pos.z);
glRotated(-rot.z, 0.0, 0.0, 1.0);
glRotated(-rot.y, 0.0, 1.0, 0.0);
glRotated(-rot.x, 1.0, 0.0, 0.0);
glScaled(1/globalScale.x, 1/globalScale.y, 1/globalScale.z);
}
Rendering call:
void RenderObject(GameObject* go){
for(auto goc : go->children)
goc->Update<MeshRenderer>();//RenderObject(goc);
}
void RenderScene(){
auto scene = EditorInstance::GetSingleton()->currentScene;
for(auto go : scene->hierarchy){
RenderObject(go);
if(auto mr = (go->GetComponent<Camera>())){
mr->Update();
}
}
}
... render->setOnRender(RenderScene); ...
Main rendering method:
int render()
{
#ifdef EDITOR
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, fbo); // Bind our frame buffer for rendering
glPushAttrib(GL_VIEWPORT_BIT | GL_ENABLE_BIT); // Push our glEnable and glViewport states
DrawGrid(100);
#else
if(NukeOGL::getSingleton() != this){
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, fbo); // Bind our frame buffer for rendering
glPushAttrib(GL_VIEWPORT_BIT | GL_ENABLE_BIT); // Push our glEnable and glViewport states
}
#endif
//glClearColor(0, 0, 0, 1.0);
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);
glLoadIdentity ();
gluLookAt(transform->position.x,
transform->position.y,
transform->position.z,
transform->position.x + transform->direction().x,
transform->position.y + transform->direction().y,
transform->position.z + transform->direction().z,
0.0,
1.0,
0.0);
if(_onRender.size() > 0)
for(auto _rn : _onRender){
_rn();
}
#ifdef EDITOR
glPopAttrib(); // Restore our glEnable and glViewport states
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, 0); // Unbind our texture
#else
if(NukeOGL::getSingleton() != this){
glPopAttrib(); // Restore our glEnable and glViewport states
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, 0); // Unbind our texture
}
#endif
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);
if(_onGUI.size() > 0)
for(auto _rn : _onGUI){
_rn();
}
glutSwapBuffers();
//glutPostRedisplay();
return 0;
}
What I am doing wrong? What I should do to make it right scale from local to world space?
Scale and rotation are linear transformations, their effects are relative to distance to origin of coordinates system. You must apply them in a proper order and with a proper origin each.
I explain in 2D so it's easier to understand.
Say you have a rectangle of size axb, whose center is at {cx,cy} and you want to scale and rotate (in this order). Then you first translate to {0,0} then scale, then rotate, then translate it back to {cx,cy}. Since each transformation is defined by a matrix, and usually OpenGL matrices are defined in column-mayor order, the composed matrix for this object may be:
MObj_i = MObj_i_trans(cx,cy) * MObj_i_rot(cangle, caxis) * MObj_i_sca(cfactor) * MObj_i_trans(-cx,-cy)
After you do those transformations for each object (with its own center/scale/rotation each) you want a "global" scale and rotation. Again you need a center of scale/rotation:
MGlobal = MGlo_trans(gx,gy) * MGlo_rot(gangle, gaxis) * MGlo_sca(gfactor) * MGlo_trans(-gx,-gy)
Usually the world center is `{0,0}' so you can avoid translations:
MGlobal = MGlo_rot(gangle, gaxis) * MGlo_sca(gfactor)
The good news is that the transformations can be combined into an only matrix. So, for each object, you apply the matrix:
MObjec_i = MGlobal * MObj_i
If you use glm for these maths, don't forget to inializate an Identity matrix:
glm::mat4 objScale(1.0);
objScale(objScale, vec3(fx, fy, fz));
glm::mat4 objRotate(1.0);
objRotate(objRotate, angle, vec3(axis.x, axis.y, axis.z));
etc.

OpenGL glOrtho set up match the window size

I am a beginner in OpenGl and I am struggling a bit with setting up the glOrtho camera to match the window size so that I can draw a line using the window's coordinates. For example, if I want to draw a line from coordinates 0,10 (x,y) to 600,10. I managed to draw the line (which will be a "Separator" from the viewport and a toolbar with buttons) in my current set up but it was by "try end error" approach and the coordinates that I needed to put don't make any sense to me. When I tried to draw a line using the above-mentioned coordinates, the line simply did not show up. What I need to change in the glOrtho set up in order to work with these (1000x600) screen size and draw my vertices and not these:
glVertex3f(-2.0, 11.0, 0.0);
glVertex3f(20.0, 11.0, 0.0);
Note, my current window size is 1000x600 (width/height)
This is the line (on the top that crosses the whole screen):
This is my OGWindow class that handles all of the drawing:
void OGWindow::MyReSizeGLScene(int fwidth, int fheight)
{
// Store window size in class variables so it can be accessed in myDrawGLScene() if necessary
wWidth = fwidth;
wHeight = fheight;
// Calculate aspect ration of the OpenGL window
aspect_ratio = (float) fwidth / fheight;
// Set camera so it can see a square area of space running from 0 to 10
// in both X and Y directions, plus a bit of space around it.
Ymin = -1;
Ymax = 12;
Xmin = -1;
// Choose Xmax so that the aspect ration of the projection
// = the aspect ratio of the viewport
Xmax = (aspect_ratio * (Ymax -Ymin)) + Xmin;
glMatrixMode(GL_PROJECTION); // Select The Projection Stack
glLoadIdentity();
glOrtho(Xmin, Xmax, Ymin, Ymax, -1.0, 1.0);
glViewport(0, 0, wWidth, wHeight); // Viewport fills the window
}
void OGWindow::myDrawGLScene(GLvoid) // Here's Where We Do All The Drawing
{
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); // clear the drawing area
OGWindow::myDrawModel();
drawToolbar();
glutSwapBuffers(); // Needed if we're running an animation
glFlush();
}
void OGWindow::myDrawModel(GLvoid)
{
switch ( squareColour ) {
case RED:
glColor3f(1.0, 0.0, 0.0);
break;
case BLUE:
glColor3f(0.0, 0.0, 1.0);
break;
}
glBegin( GL_QUADS );
glVertex3f( squareX, squareY, 0.0 ); // Coordinates of bottom-left corner of square
glVertex3f( squareX + squareWidth, squareY, 0.0 );
glVertex3f( squareX + squareWidth, squareY + squareHeight, 0.0 );
glVertex3f( squareX, squareY + squareHeight, 0.0 );
glEnd();
}
// Convert from screen coords returned by mouse
// to world coordinates.
// Return result in worldX, worldY
void OGWindow::screen2World(int screenX, int screenY, double & worldX, double & worldY)
{
// Dimensions of rectangle viewed by camera projection
double projWidth = Xmax -Xmin;
double projHeight = Ymax - Ymin;
// Screen coords with origin at bottom left
int screenLeft = screenX;
int screenUp = wHeight - screenY;
worldX = Xmin + screenLeft * projWidth / wWidth ;
worldY = Ymin + screenUp * projHeight / wHeight ;
}
//Method to draw the toolbar separator line
void OGWindow::drawToolbar(GLvoid) {
//draw toolbar line separator
glColor3f(0.0,0.0,0.0);
glBegin(GL_LINES);
glVertex3f(-2.0, 11.0, 0.0);
glVertex3f(20.0, 11.0, 0.0);
glEnd();
//draw create button
glPushMatrix();
glTranslatef(2.0, 10.0, 0.0);
glutSolidCube(2.0);
glPopMatrix();
}
This is my main class where I am ivoking the methods from OGWindow:
int main(int argc, char **argv) {
glutInit(&argc, argv);
glutInitDisplayMode(GLUT_RGB | GLUT_DOUBLE | GLUT_DEPTH);
glutInitWindowSize( 1000, 600 );
glutInitWindowPosition(0, 0);
glutCreateWindow("OpenGL Demo");
glEnable(GL_DEPTH_TEST); // enable the depth buffer test
glutDisplayFunc(DrawGLScene);
glutReshapeFunc(ReSizeGLScene);
glutMouseFunc(mouseClick);
glutMotionFunc(mouseMotion);
glutPassiveMotionFunc(mousePassiveMotion);
glutIdleFunc(Idle);
theWindow.initGL();
glutMainLoop();
}
Check out the documentation of glOrtho function. As you see, there are 6 parameters: left, right, bottom, top, near, far. You made mistake by setting window width to top instead of bottom parameter. Here's proper use of function:
glOrtho (0, 1000, 600, 0, -1.0, 1.0)
So, first your ortho settings. If you want your camera to match the screen dimensions, glOrtho has to use the same dimensions.
// This will anchor the camera to the center of the screen
// Camera will be centered on (0,0)
glOrtho( -screenWidth/2.f, screenWidth/2.f, -screenHeight/2.f, screenHeight/2.f, -1, 1 );
// This will anchor the camera to the lower left corner of the screen
// Camera will be centered on (screenWidth/2, screenHeight/2)
glOrtho( 0, screenWidth, 0, screenHeight, -1, 1 );
Try both and see the difference. Although if you are making some sort of editor, where your camera doesn't move, you may be looking for the second ortho setup.
Second, you only ever use (apparently) the GL_PROJECTION matrix mode. You must use this mode to set the camera projection and GL_MODELVIEW to apply transforms to the camera or the objects.
So when you call resize and don't change the matrix mode back to GL_MODELVIEW, you'll be applying translations to the projection matrix.
If you did forget to initialize the modelview matrix it may contain garbage values and yield unexpected results.

opengl: rotating an ellipse

I am having trouble rotating an ellipse in OpenGL. So, I have some code to draw an ellipse as follows:
glPushAttrib(GL_CURRENT_BIT);
glColor3f(1.0f, 0.0f, 0.0f);
glLineWidth(2.0);
glPushMatrix();
glTranslatef(0, 0, 0); // ellipse centre
glBegin(GL_LINE_LOOP);
float inc = (float) M_PI / 500.0;
for (GLfloat i = 0; i < M_PI * 2; i+=inc)
{
float x = cos(i) * 0.4;
float y = sin(i) * 0.4;
glVertex2f(x, y);
}
glEnd();
glPopMatrix();
glPopAttrib();
This produces a picture as so:
Now what I want to do is rotate this ellipse clockwise. So I added a glrotate in between but the result was not what I had expected.
So, I did something like:
glPushAttrib(GL_CURRENT_BIT);
glColor3f(1.0f, 0.0f, 0.0f);
glLineWidth(2.0);
glPushMatrix();
glTranslatef(0, 0, 0);
glRotatef(-90, 1, 1, 0);
glBegin(GL_LINE_LOOP);
float inc = (float) M_PI / 500.0;
for (GLfloat i = 0; i < M_PI * 2; i+=inc)
{
float x = cos(i) * 0.4;
float y = sin(i) * 0.4;
glVertex2f(x, y);
}
glEnd();
glPopMatrix();
glPopAttrib();
This produced an image which was simply collapsed. What I wanted to do was rotate the ellipse along its center by the specified degrees. Also, I tried playing around with the various parameters of glRotatef but could not get it do as I expected. The resulting image looks like:
You're working in the XY plane, so you can't really rotate around a vector in XY. Instead, you want to rotate along the unit Z axis (glRotate (angle, 0, 0, 1);). Imagine your screen being the XY coordinate system and the Z axis pointing inwards. What you want is to rotate around the Z axis, so your XY plane remains in the XY plane.

Rendering visually perfect squares in OpenGL?

In OpenGL's fixed pipeline, by default, specifying vertex coordinates using glVertex3f is equivalent to specifying a location between -1.0 and +1.0 in screen space. Therefore, given a set of 4 perfectly adjacent screen-space vertices using GL_TRIANGLE_STRIP (or even GL_QUADS), and unless your window is already perfectly square, you will always render a rectangle instead of a perfect square...
Knowing the width, height and aspect ratio of a window, is there some way to correct this?
I have tried multiplying the vertex coordinates by the aspect ratio, which unfortunately seemed to achieve the same visual effect.
Here's the full source code I'm currently using:
#include "main.h"
#pragma comment(lib, "glut32.lib")
int g_width = 800;
int g_height = 600;
int g_aspectRatio = double(g_width) / double(g_height);
bool g_bInitialized = false;
int main(int argc, char **argv)
{
glutInit(&argc, argv);
glutInitDisplayMode(GLUT_DEPTH | GLUT_DOUBLE | GLUT_RGBA);
glutInitWindowPosition(0, 0);
glutInitWindowSize(g_width, g_height);
glutCreateWindow("OpenGL Test App");
glutDisplayFunc(onRender);
glutReshapeFunc(onSize);
glutIdleFunc(onRender);
glutMainLoop();
return 0;
}
void onInit()
{
glFrontFace(GL_CW);
}
void onRender()
{
if(!g_bInitialized)
onInit();
static float angle = 0.0f;
const float p = 0.5f * g_aspectRatio;
glLoadIdentity();
gluLookAt(
0.0f, 0.0f, 10.0f,
0.0f, 0.0f, 0.0f,
0.0f, 1.0f, 0.0f
);
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);
glScalef(1, -1, 1); // Flip the Y-axis
glRotatef(angle, 0.0f, 1.0f, 0.0f);
glBegin(GL_TRIANGLE_STRIP);
{
glColor4f(1.0, 0.0, 0.0, 1.0); // Red
glVertex3f(-p, -p, 0.0); // Top-Left
glColor4f(0.0, 1.0, 0.0, 1.0); // Green
glVertex3f(p, -p, 0.0); // Top-Right
glColor4f(0.0, 0.0, 1.0, 1.0); // Blue
glVertex3f(-p, p, 0.0); // Bottom-Left
glColor4f(1.0, 1.0, 0.0, 1.0); // Yellow
glVertex3f(p, p, 0.0); // Bottom-Left
}
glEnd();
angle += 0.6f;
glutSwapBuffers();
}
void onSize(int w, int h)
{
g_width = max(w, 1);
g_height = max(h, 1);
g_aspectRatio = double(g_width) / double(g_height);
glMatrixMode(GL_PROJECTION);
glLoadIdentity();
glViewport(0, 0, w, h);
gluPerspective(45, g_aspectRatio, 1, 1000);
glMatrixMode(GL_MODELVIEW);
}
EDIT:
This has been solved... In the above code, I had defined g_aspectRatio as an int instead of a floating-point value. Therefore, it's value was always 1...
In my (old) experience, that's just why you have an aspect ratio argument to gluPerspective().
The manual page says:
In general, the aspect ratio in gluPerspective should match
the aspect ratio of the associated viewport. For example, aspect = 2.0
means the viewer's angle of view is twice as wide in x as it is in y.
If the viewport is twice as wide as it is tall, it displays the image
without distortion.
Check your g_aspectRatio value.
by default, specifying vertex coordinates using glVertex3f is equivalent to specifying a location between -1.0 and +1.0 in screen space
Wrong. Coordinates passed to OpenGL through glVertex or a glVertexPointer vertex array are in model space. The transformation to screen space happens by transforming into view space by the modelview matrix and from view space to clip space by the projection matrix. Then clipping is applied and the perspective divide applied to reach normalized coordinate space.
Hence the value range for glVertex can be whatever you like it to be. By applying the right projection matrix you get your view space to be in [-aspect; aspect]×[-1, 1] if you like that.
glMatrixMode(GL_PROJECTION);
glLoadIdentity();
glOrtho(-aspect, aspect, -1, 1, -1, 1);