SDL2 (C++) not all .png images render correctly - c++

Note: I am using Legacy OpenGL
So I have encountered a problem where only a few .png images get rendered correctly.
Here is an example screenshot of rendered game (Left: enemyRed.png, Right: (ship.png)):
Here is the code I use to load images:
SpriteImage* SDLOpenGL::loadImage(std::string path) {
SDL_Surface * surface = IMG_Load(path.c_str());
SpriteImage * image = new SpriteImage(&*surface);
return image;
}
Here is the SpriteImage class:
SpriteImage::SpriteImage(SDL_Surface *surface) {
this->surface = surface;
this->TextureID = 0;
}
void SpriteImage::bind() {
glTexImage2D(GL_TEXTURE_2D,
0,
this->mode,
this->surface->w,
this->surface->h,
0,
this->mode,
GL_UNSIGNED_BYTE,
this->surface->pixels
);
glBindTexture(GL_TEXTURE_2D, this->TextureID);
}
int SpriteImage::getWidth() {
return this->surface->w;
}
int SpriteImage::getHeight() {
return this->surface->h;
}
Here is where I render the images:
(note, this->getCurrentImage() returns a "SpriteImage")
void Sprite::draw(float delta) {
this->getCurrentImage()->bind();
glBegin(GL_QUADS);
glTexCoord2f(0, 0);
glVertex2f(0.0f, 0.0f);
glTexCoord2f(0, 1);
glVertex2f(0.0f, this->getHeight());
glTexCoord2f(1, 1);
glVertex2f(this->getWidth(), this->getHeight());
glTexCoord2f(1, 0);
glVertex2f(this->getWidth(), 0.0f);
glEnd();
this->next();
}

The image that doesn't render has a width indivisible by four. You should use glPixelStorei with GL_UNPACK_ALIGNMENT before glTexImage2D to specify alignment of your rows in memory -- which are probably unaligned (OpenGL assumes by default that they are four-byte aligned).
You should first glBindTexture and then upload the data with glTexImage2D.
Do you ever call to glGenTextures? You should initialize the TextureID with glGenTextures.
You should not upload the texture each and every time you bind the texture. Instead upload it in the constructor, then to switch textures you only need to glBindTexture your TextureID.

Related

Draw an IOSurface to an OpenGL context

I have an OpenGL context on which I draw successfully using OpenGL.
I need to draw a specific rectangle of an IOSurface to this context.
What is the best way to do this on 10.8?
NOTE:
I know how to do this on 10.9 using CoreImage (by createing a CIImage from the IOSurface, and render it with [CIContext drawImage:inRect:fromRect]).
However, this does not work well for me on 10.8 (each raw of the image is displayed with a different offset, and the image is distorted diagonally).
Edit: Here is the code that works on 10.9 but not on 10.8:
CGColorSpaceRef colorSpace = CGColorSpaceCreateWithName(kCGColorSpaceSRGB);
CIImage* ciImage = [[CIImage alloc] initWithIOSurface:surface plane:0 format:kCVPixelFormatType_32BGRA options:#{kCIImageColorSpace : (__bridge id)colorSpace}];
NSRect flippedFromRect = fromRect;
// Flip rect before passing to CoreImage:
{
flippedFromRect.origin.y = IOSurfaceGetHeight(surface) - fromRect.origin.y - fromRect.size.height;
}
[ciContext drawImage:ciImage inRect:inRect fromRect:flippedFromRect];
CGColorSpaceRelease(colorSpace);
Here is the solution by wrapping the IOSurface with an OpenGL texture and draw the texture to the screen. This assumes a similar API to [CIContext render:toIOSurface:bounds:colorSpace:] but a vertically flipped OpenGL coordinate system.
// Draw surface on OpenGL context
{
// Enable the rectangle texture extenstion
glEnable(GL_TEXTURE_RECTANGLE_EXT);
// 1. Create a texture from the IOSurface
GLuint name;
{
CGLContextObj cgl_ctx = ...
glGenTextures(1, &name);
GLsizei surface_w = (GLsizei)IOSurfaceGetWidth(surface);
GLsizei surface_h = (GLsizei)IOSurfaceGetHeight(surface);
glBindTexture(GL_TEXTURE_RECTANGLE_EXT, name);
CGLError cglError =
CGLTexImageIOSurface2D(cgl_ctx, GL_TEXTURE_RECTANGLE_EXT, GL_RGBA, surface_w, surface_h, GL_BGRA, GL_UNSIGNED_INT_8_8_8_8_REV, surface, 0);
glBindTexture(GL_TEXTURE_RECTANGLE_EXT, 0);
}
// 2. Draw the texture to the current OpenGL context
{
glBindTexture(GL_TEXTURE_RECTANGLE_EXT, name);
glTexParameteri(GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
glTexParameteri(GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
glTexEnvf(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_REPLACE);
glBegin(GL_QUADS);
glColor4f(0.f, 0.f, 1.0f, 1.0f);
glTexCoord2f( (float)NSMinX(fromRect), (float)(NSMinY(fromRect)));
glVertex2f( (float)NSMinX(inRect), (float)(NSMinY(inRect)));
glTexCoord2f( (float)NSMaxX(fromRect), (float)NSMinY(fromRect));
glVertex2f( (float)NSMaxX(inRect), (float)NSMinY(inRect));
glTexCoord2f( (float)NSMaxX(fromRect), (float)NSMaxY(fromRect));
glVertex2f( (float)NSMaxX(inRect), (float)NSMaxY(inRect));
glTexCoord2f( (float)NSMinX(fromRect), (float)NSMaxY(fromRect));
glVertex2f( (float)NSMinX(inRect), (float)NSMaxY(inRect));
glEnd();
glBindTexture(GL_TEXTURE_RECTANGLE_EXT, 0);
}
glDeleteTextures(1, &name);
}
If you need to draw in the display's color profile, you can explicitly call ColorSync and pass it your source profile and destination profile. It will return to you a “recipe” to perform the color correction. That recipe actually has a linearization, a color conversion (a 3x3 conversion matrix) and a gamma.
FragmentInfo = ColorSyncTransformCopyProperty (transform, kColorSyncTransformFullConversionData, NULL);
If you like, you can combine all those operations into a 3D lookup table. That's actually what happens in the color management of many of the OS X frameworks and applications.
References:
Apple TextureUpload sample code
Draw IOSurfaces to another IOSurface
OpenGL Options for Advanced Color Management

glReadPixels returns darker image

I'm using glReadPixels to create a copy of the final frame of a state in my game (e.g. in-game, in-menu, etc...) for use in a transitional effect. I get that copy but it seems darker than the original:
here is the relevant code:
void GameStateManager::GetLastFrame()
{
if (last_frame_buffer.id != 0)
glDeleteTextures(1,&last_frame_buffer.id);
uint8* data = new uint8[Constants::SCREEN_WIDTH * Constants::SCREEN_HEIGHT*3];
glReadPixels(0,0,Constants::SCREEN_WIDTH,Constants::SCREEN_HEIGHT,GL_RGB,GL_UNSIGNED_BYTE,data);
last_frame_buffer.target = GL_TEXTURE_RECTANGLE_ARB;
glGenTextures(1,&last_frame_buffer.id);
glBindTexture(last_frame_buffer.target,last_frame_buffer.id);
glTexParameterf(last_frame_buffer.target,GL_TEXTURE_MAG_FILTER,GL_NEAREST);
glTexParameterf(last_frame_buffer.target,GL_TEXTURE_MIN_FILTER,GL_NEAREST);
glTexImage2D(last_frame_buffer.target,0,GL_RGB,Constants::SCREEN_WIDTH,Constants::SCREEN_HEIGHT,0,GL_RGB,GL_UNSIGNED_BYTE,data);
glBindTexture(last_frame_buffer.target,0);
delete[] data;
}
this is the actual render code:
glPushMatrix();
glTranslatef(0,last_frame_pos,0);
glCallList(last_frame_quad);
glPopMatrix();
and this is the display list:
glBegin(GL_QUADS);
glTexCoord2f(0 , 0); glVertex2f(0 , height);
glTexCoord2f(0 , height); glVertex2f(0 , 0);
glTexCoord2f(width, height); glVertex2f(width, 0);
glTexCoord2f(width, 0); glVertex2f(width, height);
glEnd();
Here are the things I've tried so far:
change buffer format : BYTE (made image darker), UNSIGNED_SHORT, UNSIGNED_INT, FLOAT
enable/disable : TEXTURE_2D, TEXTURE_RECTANGLE, BLEND
manually brighten pixels (failed miserably)

Why can I not load a texture into my app?

I think texture mapping is a really easy task. Actually, I implemented it many times but failed in this time and don't know why? And I can guarantee that the route to load the texture is right. Any other reasons for my confusion?
Here is my code:
GLuint mytexture;
// two functions below come from NeHe's tut. I think it works well.
AUX_RGBImageRec *LoadBMP(CHAR *Filename)
{
FILE *File=NULL;
if (!Filename)
{
return NULL;
}
File=fopen(Filename,"r");
if (File)
{
fclose(File);
return auxDIBImageLoadA(Filename);
}
return NULL;
}
int LoadGLTextures()
{
int Status=FALSE;
AUX_RGBImageRec *TextureImage[1];
memset(TextureImage,0,sizeof(void *)*1);
if (TextureImage[0]=LoadBMP("NeHe.bmp"))
{
Status=TRUE;
glGenTextures(1, &mytexture);
glBindTexture(GL_TEXTURE_2D, mytexture);
glTexImage2D(GL_TEXTURE_2D, 0, 3, TextureImage[0]->sizeX, TextureImage[0]->sizeY, 0,
GL_RGB, GL_UNSIGNED_BYTE, TextureImage[0]->data);
glTexParameteri(GL_TEXTURE_2D,GL_TEXTURE_MIN_FILTER,GL_LINEAR);
glTexParameteri(GL_TEXTURE_2D,GL_TEXTURE_MAG_FILTER,GL_LINEAR);
}
if (TextureImage[0])
{
if (TextureImage[0]->data)
{
free(TextureImage[0]->data);
}
free(TextureImage[0]);
}
return Status;
}
//next is my Init() code:
bool DemoInit( void )
{
if (!LoadGLTextures())
{
return FALSE;
}
glEnable(GL_TEXTURE_2D);
........//other init is ok
}
bool DemoRender()
{
...///render other things
glEnable(GL_TEXTURE_2D);
glBindTexture(GL_TEXTURE_2D, mytexture);
glColor3f(0,0,1);
glBegin(GL_QUADS);
glTexCoord2f(0, 0); glVertex2f(0, 0);
glTexCoord2f(1, 0); glVertex2f(200, 0);
glTexCoord2f(1, 1); glVertex2f(200, 200);
glTexCoord2f(0, 1); glVertex2f(0, 200);
glEnd();
glDisable(GL_TEXTURE_2D);
glBindTexture(GL_TEXTURE_2D, 0);
}
Pretty clear, ha? However, the final result only has a blue rectangle without the texture. Anybody could give me a hint?
Assuming TextureImage[0]->data is correctly populated:
However, the final result only has a blue rectangle without the texture.
You're using the default GL_MODULATE texture environment. Either switch glColor3f(0,0,1) to glColor3f(1,1,1) or use GL_DECAL.
You might also try a glPixelStorei(GL_UNPACK_ALIGNMENT, 1) before your glTexImage2D() since you're using GL_RGB for format.
The problem is I set the GL_LINE mode before I load the texture and I failed to notice that. So after I set the GL_FILL mode, everything is fine!!!

OpenGL - draw pixels to screen?

I want to draw a 2D array of pixel data (RGB / grayscale values) on the screen as fast as possible, using OpenGL. The pixel data changes frequently.
I had hoped that I would find a simple function that would let me push in a pointer to an array representing the pixel data, since this is probably the fastest approach. Unfortunately, I have found no such function.
What is the best way to accomplish this task?
Maybe glDrawPixels is the function you are looking for? Though if the data is static it would be better to create a texture with it, and then draw that each frame.
I recently had a similar problem, as I am trying to render a video to screen (ie repeatedly upload pixel data to the VRAM), my approach is:
use glTexImage2D and glTexSubImage2D to upload the data to the texture (ie bind the texture (and texture unit, if applicable) before calling that)
in my case as the video frame rate (usually about 24 fps) is lower than the framerate of my application (aimed at 60 fps), in order to avoid uploading the same data again I use a framebuffer object (check out glGenFramebuffers/glBindFramebuffer/glDeleteFramebuffers) and link my texture with the framebuffer (glFramebufferTexture2D). I then upload that texture once, and draw the same frame multiple times (just normal texture access with glBindTexture)
I don't know which platform you are using, but as I am targetting Mac I use some Apple extensions to ensure the data transfer to the VRAM happens through DMA (ie make glTexSubImage2D return immediately to let the CPU do other work) - please feel free to ask me for more info if you are using Mac too
also as you are using just grayscale, you might want to consider just using a GL_LUMINANCE texture (ie 1 byte per pixel) rather than RGB based format to make the upload faster (but that depends on the size of your texture data, I was streaming HD 1920x1080 video so I needed to make sure to keep it down)
also be aware of the format your hardware is using to avoid unnecessary data conversions (ie normally it seems better to use BGRA data than for example just RGB)
finally, in my code I replaced all the fixed pipeline functionality with shaders (in particular the conversion of the data from grayscale or YUV format to RGB), but again all that depends on the size of your data, and the workload of your CPU or GPU
Hope this helps, feel free to message me if you need further info
I would think the fastest way would be to draw a screen sized quad with ortho projection and use a pixel shader and Texture Buffer Object to draw directly to the texture in the pixel shader. Due to latency transferring to/from the TBO you may want to see if double buffering would help.
If speed isn't much of a concern (you just need fairly interactive framerates) glDrawPixels is easy to use and works well enough for many purposes.
My solution for getting dynamically changing image data to the screen in OpenGL,
#define WIN32_LEAN_AND_MEAN
#include "wx/wx.h"
#include "wx/sizer.h"
#include "wx/glcanvas.h"
#include "BasicGLPane.h"
// include OpenGL
#ifdef __WXMAC__
#include "OpenGL/glu.h"
#include "OpenGL/gl.h"
#else
#include <GL/glu.h>
#include <GL/gl.h>
#endif
#include "ORIScanMainFrame.h"
BEGIN_EVENT_TABLE(BasicGLPane, wxGLCanvas)
EVT_MOTION(BasicGLPane::mouseMoved)
EVT_LEFT_DOWN(BasicGLPane::mouseDown)
EVT_LEFT_UP(BasicGLPane::mouseReleased)
EVT_RIGHT_DOWN(BasicGLPane::rightClick)
EVT_LEAVE_WINDOW(BasicGLPane::mouseLeftWindow)
EVT_SIZE(BasicGLPane::resized)
EVT_KEY_DOWN(BasicGLPane::keyPressed)
EVT_KEY_UP(BasicGLPane::keyReleased)
EVT_MOUSEWHEEL(BasicGLPane::mouseWheelMoved)
EVT_PAINT(BasicGLPane::render)
END_EVENT_TABLE()
// Test data for image generation. floats range 0.0 to 1.0, in RGBRGBRGB... order.
// Array is 1024 * 3 long. Note that 32 * 32 is 1024 and is the largest image we can randomly generate.
float* randomFloatRGB;
float* randomFloatRGBGrey;
BasicGLPane::BasicGLPane(wxFrame* parent, int* args) :
wxGLCanvas(parent, wxID_ANY, args, wxDefaultPosition, wxDefaultSize, wxFULL_REPAINT_ON_RESIZE)
{
m_context = new wxGLContext(this);
randomFloatRGB = new float[1024 * 3];
randomFloatRGBGrey = new float[1024 * 3];
// In GL images 0,0 is in the lower left corner so the draw routine does a vertical flip to get 'regular' images right side up.
for (int i = 0; i < 1024; i++) {
// Red
randomFloatRGB[i * 3] = static_cast <float> (rand()) / static_cast <float> (RAND_MAX);
// Green
randomFloatRGB[i * 3 + 1] = static_cast <float> (rand()) / static_cast <float> (RAND_MAX);
// Blue
randomFloatRGB[i * 3 + 2] = static_cast <float> (rand()) / static_cast <float> (RAND_MAX);
// Telltale 2 white pixels in 0,0 corner.
if (i < 2) {
randomFloatRGB[i * 3] = randomFloatRGB[i * 3 + 1] = randomFloatRGB[i * 3 + 2] = 1.0f;
}
randomFloatRGBGrey[i * 3] = randomFloatRGB[i * 3];
randomFloatRGBGrey[i * 3 + 1] = randomFloatRGB[i * 3];
randomFloatRGBGrey[i * 3 + 2] = randomFloatRGB[i * 3];
}
// To avoid flashing on MSW
SetBackgroundStyle(wxBG_STYLE_CUSTOM);
}
BasicGLPane::~BasicGLPane()
{
delete m_context;
}
void BasicGLPane::resized(wxSizeEvent& evt)
{
// wxGLCanvas::OnSize(evt);
Refresh();
}
int BasicGLPane::getWidth()
{
return GetSize().x;
}
int BasicGLPane::getHeight()
{
return GetSize().y;
}
void BasicGLPane::render(wxPaintEvent& evt)
{
assert(GetParent());
assert(GetParent()->GetParent());
ORIScanMainFrame* mf = dynamic_cast<ORIScanMainFrame*>(GetParent()->GetParent());
assert(mf);
switch (mf->currentMainView) {
case ORIViewSelection::ViewCamera:
renderCamera(evt);
break;
case ORIViewSelection::ViewDepth:
renderDepth(evt);
break;
case ORIViewSelection::ViewPointCloud:
renderPointCloud(evt);
break;
case ORIViewSelection::View3DModel:
render3DModel(evt);
break;
default:
renderNone(evt);
}
}
void BasicGLPane::renderNone(wxPaintEvent& evt) {
if (!IsShown())
return;
SetCurrent(*(m_context));
glPushAttrib(GL_ALL_ATTRIB_BITS);
glClearColor(0.08f, 0.11f, 0.15f, 1.0f);
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);
glFlush();
SwapBuffers();
glPopAttrib();
}
GLuint makeOpenGlTextureFromDataLuninanceFloats(int width, int height, float* f) {
GLuint textureID;
glEnable(GL_TEXTURE_2D);
glGenTextures(1, &textureID);
// "Bind" the newly created texture : all future texture functions will modify this texture
glBindTexture(GL_TEXTURE_2D, textureID);
// Give the image to OpenGL
glTexImage2D(GL_TEXTURE_2D, 0, GL_FLOAT, width, height, 0, GL_FLOAT, GL_LUMINANCE, f);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
return textureID;
}
GLuint makeOpenGlTextureFromRGBInts(int width, int height, unsigned int* f) {
GLuint textureID;
glEnable(GL_TEXTURE_2D);
glGenTextures(1, &textureID);
// "Bind" the newly created texture : all future texture functions will modify this texture
glBindTexture(GL_TEXTURE_2D, textureID);
// Give the image to OpenGL
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA, width, height, 0, GL_RGBA, GL_UNSIGNED_INT, f);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
return textureID;
}
/// <summary>
/// Range of each float is 0.0f to 1.0f
/// </summary>
/// <param name="width"></param>
/// <param name="height"></param>
/// <param name="floatRGB"></param>
/// <returns></returns>
GLuint makeOpenGlTextureFromRGBFloats(int width, int height, float* floatRGB) {
GLuint textureID;
// 4.6.0 NVIDIA 457.30 (R Keene machine, 11/25/2020)
// auto sss = glGetString(GL_VERSION);
glGenTextures(1, &textureID);
// "Bind" the newly created texture : all future texture functions will modify this texture
glBindTexture(GL_TEXTURE_2D, textureID);
// Give the image to OpenGL
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGB, width, height, 0, GL_RGB, GL_FLOAT, floatRGB);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
return textureID;
}
void BasicGLPane::DrawTextureToScreenFloat(int w, int h, float* floatDataPtr, GLuint (*textureFactory)(int width, int height, float* floatRGB)) {
if (w <= 0 || h <= 0 || floatDataPtr == NULL || w > 5000 || h > 5000) {
assert(false);
return;
}
SetCurrent(*(m_context));
glPushAttrib(GL_ALL_ATTRIB_BITS);
glPushMatrix();
glPushClientAttrib(GL_CLIENT_ALL_ATTRIB_BITS);
glClearColor(0.15f, 0.11f, 0.02f, 1.0f);
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);
glEnable(GL_TEXTURE_2D);
glLoadIdentity();
glMatrixMode(GL_MODELVIEW);
glLoadIdentity();
glMatrixMode(GL_PROJECTION);
glLoadIdentity();
// 4.6.0 NVIDIA 457.30 (R Keene machine, 11/25/2020)
// auto sss = glGetString(GL_VERSION);
float onePixelW = (float)getWidth() / (float)w;
float onePixelH = (float)getHeight() / (float)h;
float orthoW = w;
float orthoH = h;
if (onePixelH > onePixelW) {
orthoH = h * onePixelH / onePixelW;
}
else {
orthoW = w * onePixelW / onePixelH;
}
// We want the image at the top of the window, not the bottom if the window is too tall.
int topOfScreen = (float)getHeight() / onePixelH;
// If the winjdow resizes after creation you need to change the viewport.
glViewport(0, 0, getWidth(), getHeight());
gluOrtho2D(0.0, orthoW, (double)topOfScreen - (double)orthoH, topOfScreen);
GLuint myTextureName = textureFactory(w, h, floatDataPtr);
glBegin(GL_QUADS);
{
// This order of UV coords and verticies will do the vertical flip of the image to get the 'regular' image 0,0
// in the top left corner.
glTexCoord2f(0.0f, 1.0f); glVertex3f(0.0f, 0.0f, 0.0f);
glTexCoord2f(1.0f, 1.0f); glVertex3f(0.0f + w, 0.0f, 0.0f);
glTexCoord2f(1.0f, 0.0f); glVertex3f(0.0f + w, 0.0f + h, 0.0f);
glTexCoord2f(0.0f, 0.0f); glVertex3f(0.0f, 0.0f + h, 0.0f);
}
glEnd();
glDeleteTextures(1, &myTextureName);
glFlush();
SwapBuffers();
glPopClientAttrib();
glPopMatrix();
glPopAttrib();
}
void BasicGLPane::DrawTextureToScreenMat(wxPaintEvent& evt, cv::Mat m, float brightness) {
m.type();
if (m.empty()) {
renderNone(evt);
return;
}
if (m.type() == CV_32FC1) { // Grey scale.
DrawTextureToScreenFloat(m.cols, m.rows, (float*)m.data, makeOpenGlTextureFromDataLuninanceFloats);
}
if (m.type() == CV_32FC3) { // Color.
DrawTextureToScreenFloat(m.cols, m.rows, (float*)m.data, makeOpenGlTextureFromRGBFloats);
}
else {
renderNone(evt);
}
}
void BasicGLPane::renderCamera(wxPaintEvent& evt) {
if (!IsShown())
return;
DrawTextureToScreenMat(evt, ORITopControl::Instance->im_white);
}
void BasicGLPane::renderDepth(wxPaintEvent& evt) {
if (!IsShown())
return;
DrawTextureToScreenMat(evt, ORITopControl::Instance->depth_map);
}
void BasicGLPane::render3DModel(wxPaintEvent& evt) {
if (!IsShown())
return;
SetCurrent(*(m_context));
glPushAttrib(GL_ALL_ATTRIB_BITS);
glPushMatrix();
glClearColor(0.08f, 0.11f, 0.15f, 1.0f);
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);
glFlush();
SwapBuffers();
glPopMatrix();
glPopAttrib();
}
void BasicGLPane::renderPointCloud(wxPaintEvent& evt) {
if (!IsShown())
return;
boost::unique_lock<boost::mutex> lk(ORITopControl::Instance->pointCloudCacheMutex);
SetCurrent(*(m_context));
glPushAttrib(GL_ALL_ATTRIB_BITS);
glPushMatrix();
glLoadIdentity();
glMatrixMode(GL_MODELVIEW);
glLoadIdentity();
glMatrixMode(GL_PROJECTION);
glLoadIdentity();
glViewport(0, 0, getWidth(), getHeight());
glClearColor(0.08f, 0.11f, 0.15f, 1.0f);
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);
if (ORITopControl::Instance->pointCloudCache.size() > 0) {
glMatrixMode(GL_PROJECTION);
gluPerspective( /* field of view in degree */ 40.0,
/* aspect ratio */ 1.0,
/* Z near */ 1.0, /* Z far */ 500.0);
glMatrixMode(GL_MODELVIEW);
gluLookAt(100, 70, 200, // Eye
25, 25, 25, // Look at pt
0, 0, 1); // Up Vector
glPointSize(2.0);
glBegin(GL_POINTS);
// Use explicit for loop because pointCloudFragments can grow asynchronously.
for (int i = 0; i < ORITopControl::Instance->pointCloudCache.size(); i++) {
auto frag = ORITopControl::Instance->pointCloudCache[i];
auto current_point_cloud_ptr = frag->cloud;
glPushMatrix();
// glMultMatrixf(frag->xform.data());
for (size_t n = 0; n < current_point_cloud_ptr->size(); n++) {
glColor3ub(255, 255, 255);
glVertex3d(current_point_cloud_ptr->points[n].x, current_point_cloud_ptr->points[n].y, current_point_cloud_ptr->points[n].z);
}
glPopMatrix();
}
glEnd();
}
glFlush();
SwapBuffers();
glPopMatrix();
glPopAttrib();
}

rendering SDL_TTF text onto openGL Red Square instead of text

I've been attempting to render text onto an openGL window using SDL and the SDL_TTF library on windows XP, VS2010.
Versions:
SDL version 1.2.14
SDL TTF devel 1.2.10
openGL (version is at least 2-3 years old).
I have successfully created an openGL window using SDL / SDL_image and can render lines / polygons onto it with no problems.
However, moving onto text it appears that there is some flaw in my current program, I am getting the following result when trying this code here
for those not willing to pastebin here are only the crutial code segments:
void drawText(char * text) {
glLoadIdentity();
SDL_Color clrFg = {0,0,255,0}; // set colour to blue (or 'red' for BGRA)
SDL_Surface *sText = TTF_RenderUTF8_Blended( fntCourier, text, clrFg );
GLuint * texture = create_texture(sText);
glBindTexture(GL_TEXTURE_2D, *texture);
// draw a polygon and map the texture to it, may be the source of error
glBegin(GL_QUADS); {
glTexCoord2i(0, 0); glVertex3f(0, 0, 0);
glTexCoord2i(1, 0); glVertex3f(0 + sText->w, 0, 0);
glTexCoord2i(1, 1); glVertex3f(0 + sText->w, 0 + sText->h, 0);
glTexCoord2i(0, 1); glVertex3f(0, 0 + sText->h, 0);
} glEnd();
// free the surface and texture, removing this code has no effect
SDL_FreeSurface( sText );
glDeleteTextures( 1, texture );
}
segment 2:
// create GLTexture out of SDL_Surface
GLuint * create_texture(SDL_Surface *surface) {
GLuint texture = 0;
glGenTextures(1, &texture);
glBindTexture(GL_TEXTURE_2D, texture);
// The SDL_Surface appears to have BGR_A formatting, however this ends up with a
// white rectangle no matter which colour i set in the previous code.
int Mode = GL_RGB;
if(surface->format->BytesPerPixel == 4) {
Mode = GL_RGBA;
}
glTexImage2D(GL_TEXTURE_2D, 0, Mode, surface->w, surface->h, 0, Mode,
GL_UNSIGNED_BYTE, surface->pixels);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
return &texture;
}
Is there an obvious bit of code I am missing?
Thank you for any help on this subject.
I've been trying to learn openGL and SDL for 3 days now, so please forgive any misinformation on my part.
EDIT:
I notice that using
TTF_RenderUTF8_Shaded
TTF_RenderUTF8_Solid
Throw a null pointer exception, meaning that there is an error within the actual text rendering function (I suspect), I do not know how this means TTF_RenderUTF8_Blended returns a red square but I suspect all troubles hinge on this.
I think the problem is in the glEnable(GL_TEXTURE_2D) and glDisable(GL_TEXTURE_2D) functions which must be called every time the text is painted on the screen.And maybe also the color conversion between the SDL and GL surface is not right.
I have combined create_texture and drawText into a single function that displays the text properly. That's the code:
void drawText(char * text, TTF_Font* tmpfont) {
SDL_Rect area;
SDL_Color clrFg = {0,0,255,0};
SDL_Surface *sText = SDL_DisplayFormatAlpha(TTF_RenderUTF8_Blended( tmpfont, text, clrFg ));
area.x = 0;area.y = 0;area.w = sText->w;area.h = sText->h;
SDL_Surface* temp = SDL_CreateRGBSurface(SDL_HWSURFACE|SDL_SRCALPHA,sText->w,sText->h,32,0x000000ff,0x0000ff00,0x00ff0000,0x000000ff);
SDL_BlitSurface(sText, &area, temp, NULL);
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA, sText->w, sText->h, 0, GL_RGBA, GL_UNSIGNED_BYTE, temp->pixels);
glTexParameteri(GL_TEXTURE_2D,GL_TEXTURE_MAG_FILTER,GL_LINEAR);
glTexParameteri(GL_TEXTURE_2D,GL_TEXTURE_MIN_FILTER,GL_LINEAR);
glEnable(GL_TEXTURE_2D);
glBegin(GL_QUADS); {
glTexCoord2d(0, 0); glVertex3f(0, 0, 0);
glTexCoord2d(1, 0); glVertex3f(0 + sText->w, 0, 0);
glTexCoord2d(1, 1); glVertex3f(0 + sText->w, 0 + sText->h, 0);
glTexCoord2d(0, 1); glVertex3f(0, 0 + sText->h, 0);
} glEnd();
glDisable(GL_TEXTURE_2D);
SDL_FreeSurface( sText );
SDL_FreeSurface( temp );
}
screenshot
I'm initializing OpenGL as follows:
int Init(){
glClearColor( 0.1, 0.2, 0.2, 1);
glMatrixMode(GL_PROJECTION);
glLoadIdentity();
glOrtho( 0, 600, 300, 0, -1, 1 );
glMatrixMode(GL_MODELVIEW);
glLoadIdentity();
if( glGetError() != GL_NO_ERROR ){
return false;
}
glEnable(GL_BLEND);
glBlendFunc(GL_SRC_COLOR, GL_ONE_MINUS_SRC_ALPHA);
}
I think you should just add glEnable(GL_BLEND), because the code for the text surface says TTF_RenderUTF8_Blended( fntCourier, text, clrFg ) and you have to enable the blending abilities of opengl.
EDIT
Okay, I finally took the time to put your code through a compiler. Most importantly, compiler with -Werror so that warning turn into errors
GLuint * create_texture(SDL_Surface *surface) {
GLuint texture = 0;
/*...*/
return &texture;
}
I didn't see it first, because that's something like C coder's 101 and is quite unexpected: You must not return pointers to local variables!. Once the functions goes out of scope the pointer returned will point to nonsense only. Why do you return a pointer at all? Just return a integer:
GLuint create_texture(SDL_Surface *surface) {
GLuint texture = 0;
/*...*/
return texture;
}
Because of this you're also not going to delete the texture afterward. You upload it to OpenGL, but then loose the reference to it.
Your code misses a glEnable(GL_TEXTURE_2D) that's why you can't see any effects of texture. However your use of textures is suboptimal. They way you did it, you recreate a whole new texture each time you're about to draw that text. If that happens in a animation loop, you'll
run out of texture memory rather soon
slow it down significantly
(1) can be addressed by not generating a new texture name each redraw
(2) can be addresses by uploading new texture data only when the text changes and by not using glTexImage2D, but glTexSubImage2D (of course, if the dimensions of the texture change, it must be glTexImage2D).
EDIT, found another possible issue, but first fix your pointer issue.
You should make sure, that you're using GL_REPLACE or GL_MODULATE texture environment mode. If using GL_DECAL or GL_BLEND you end up with red text on a red quad.
There was leaking memory of of the function in my previous post and the program was crashing after some time...
I improved this by separating the texture loading and displaying:
The first function must be called before the SDL loop.It loads text string into memory:
Every string loaded must have different txtNum parameter
GLuint texture[100];
SDL_Rect area[100];
void Load_string(char * text, SDL_Color clr, int txtNum, const char* file, int ptsize){
TTF_Font* tmpfont;
tmpfont = TTF_OpenFont(file, ptsize);
SDL_Surface *sText = SDL_DisplayFormatAlpha(TTF_RenderUTF8_Solid( tmpfont, text, clr ));
area[txtNum].x = 0;area[txtNum].y = 0;area[txtNum].w = sText->w;area[txtNum].h = sText->h;
glGenTextures(1, &texture[txtNum]);
glBindTexture(GL_TEXTURE_2D, texture[txtNum]);
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA8, sText->w, sText->h, 0, GL_BGRA, GL_UNSIGNED_BYTE, sText->pixels);
glTexParameteri(GL_TEXTURE_2D,GL_TEXTURE_MIN_FILTER,GL_LINEAR);
glTexParameteri(GL_TEXTURE_2D,GL_TEXTURE_MAG_FILTER,GL_LINEAR);
SDL_FreeSurface( sText );
TTF_CloseFont(tmpfont);
}
The second one displays the string, must be called in the SDL loop:
void drawText(float coords[3], int txtNum) {
glBindTexture(GL_TEXTURE_2D, texture[txtNum]);
glEnable(GL_TEXTURE_2D);
glBegin(GL_QUADS); {
glTexCoord2f(0, 0); glVertex3f(coords[0], coords[1], coords[2]);
glTexCoord2f(1, 0); glVertex3f(coords[0] + area[txtNum].w, coords[1], coords[2]);
glTexCoord2f(1, 1); glVertex3f(coords[0] + area[txtNum].w, coords[1] + area[txtNum].h, coords[2]);
glTexCoord2f(0, 1); glVertex3f(coords[0], coords[1] + area[txtNum].h, coords[2]);
} glEnd();
glDisable(GL_TEXTURE_2D);
}