1

I'm writing a ray caster in C++, which outputs its results to regular image files. Now, I want to render the internal RGBA-representation (4 integers) to an OpenGL window, provided by GLUT.

I already know that I have to use glTexImage2D to generate a texture, assign that texture to a quad, and then render this quad right in front of the camera.

http://www.opengl.org/sdk/docs/man/xhtml/glTexImage2D.xml

The problem I'm having is that I don't know how I should present the data to the glTexImage2D function: how do I construct this gluByte data chunk containing my data, coming from a simple 4-integer RGBA representation?

genpfault
  • 51,148
  • 11
  • 85
  • 139
Jeroen Baert
  • 1,273
  • 2
  • 12
  • 28
  • For the record, I ended up doing this: `unsigned char* data = (data goes here in array);` `glTexImage2D(GL_TEXTURE_2D,0,GL_RGBA,width,height,0,GL_RGBA,GL_UNSIGNED_BYTE,data);` – Jeroen Baert Apr 05 '12 at 10:13

3 Answers3

1

How you store your data is, more or less, up to you. The last three parameters to glTexImage2D (as well as additional parameters) describe how your data is formatted. There are limitations as to how you can store your data (you can't have much padding beyond 4-byte alignment per-pixel, and the components have to be adjacent), but there is a lot of variation available.

As an array of width x height x 4 size, with an unsigned int for each component: RGBARGBARGBA... ?

That's one way of doing it, but it's not the only way.

Nicol Bolas
  • 449,505
  • 63
  • 781
  • 982
0

Convert your 0-(2^32-1) range down to 0-(2^8-1).

Or use GL_UNSIGNED_INT for format.

Community
  • 1
  • 1
genpfault
  • 51,148
  • 11
  • 85
  • 139
0

The data chunk is just a pointer to your RGBA representation. You'll want to call glTexImage2D(GL_TEXTURE_2D, 0, 4, width, height, 0, GL_RGBA, GL_UNSIGNED_INT, dataBuffer);. A line in your texture has to have width * sizeof(GLuint) bytes for this to work.

jfortmann
  • 233
  • 1
  • 7
  • And how do I store my data then? As an array of width x height x 4 size, with an unsigned int for each component: RGBARGBARGBA... ? – Jeroen Baert Apr 02 '12 at 17:23
  • Exactly. Note that it's probable that the OpenGL driver you use will convert the unsigned ints to e.g. unsigned chars internally before uploading. To save memory, you may still want to convert the uints to chars manually. – jfortmann Apr 03 '12 at 16:14