OpenGL via Win32
Author
I'm hoping someone knows something about displaying OpenGL in Win32 as I am fairly new.
Here's a summary of the problem:
I have been working on a project for a class where I use OpenGL only to draw lines and points. Everything else is done through custom code. Basically I have implemented the model->world->view->NDC transformations so now I have coordinates in the range of [-1,1] in both x and y axis, as they should be.
I basically need to do 2 things at this point.
1). Map my normalized device coordinates to pixels which is simple enough to do. (e.g. (0,0) would map to (512,384) at a resolution of (1024x768)
2). Display them in Win32 using glVertex3f(#,#,#);
The problem (yes I finally got to it) is that it seems like if I call glVertex(#,#,#) with any values greater than the absolute value of 1 they are clipped.
My guess is this is a Win32 related issue because if I set up the viewport and projection/modelview matrices normally I still can't pass glVertex3f() with any numbers abs() > 1.
My interim solution is to convert the vertices to screen coordinates, perform Z-Buffering, and then convert them back to NDC [-1,1] but that seems like a horrible way to do it.
Are there any Win32/Opengl experts out there?
Thanks
Hornsj2
I'm not exactly sure I got what you're trying to do, but shouldn't just setting an orthographic projection with your window's width and height should be enough?
I don't understand why you would want to do your model->view->NDC transformations in software if that can be done with OpenGL with a single matrix multiplication.
You also don't have to map your NDC coordinates to pixels. This is what OpenGL can do. With a standard orthographic projection you can directly pass your NDC coordinates to OpenGL provided that the viewport has been setup correctly.
If you insist in wanting to pass vertex values with abs() > 1 you simply have to setup the ortho projetion matrix accordingly.
Also, OpenGL does the z-buffering for you, why would you want to do it yourself? Doing it yourself defeats the purpose of using OpenGL. Then you could aswell write your own software rasterizer.
You also don't have to map your NDC coordinates to pixels. This is what OpenGL can do. With a standard orthographic projection you can directly pass your NDC coordinates to OpenGL provided that the viewport has been setup correctly.
If you insist in wanting to pass vertex values with abs() > 1 you simply have to setup the ortho projetion matrix accordingly.
Also, OpenGL does the z-buffering for you, why would you want to do it yourself? Doing it yourself defeats the purpose of using OpenGL. Then you could aswell write your own software rasterizer.
Author
Basically I created my own version of a software renderer that uses OpenGL to talk to the hardware. There is no practical reason for this, I am in a class.
Basically I have set up a projection matrix, model matrix, and view matrix. Then I load geometry from flat files and process them through that pipeline to get coordinates in screen space.
At the end if this pipeline I have geometry in normalized device coordinates (NDC), which are [-1,1] in both axis directions.
I can send these coordinates to OpenGL using glVertex3f(#,#,#); and they render to the screen just fine.
However, I now want to implement a z-buffer algorithm and I need integer values to represent scan lines. I am faced with converting my coordinates (floats) to another coordinate system (integers) in order to implement the algorithm and then transform them back to the NDC.
It is possible that I mixed up. I am new to win32/opengl.
Basically I have set up a projection matrix, model matrix, and view matrix. Then I load geometry from flat files and process them through that pipeline to get coordinates in screen space.
At the end if this pipeline I have geometry in normalized device coordinates (NDC), which are [-1,1] in both axis directions.
I can send these coordinates to OpenGL using glVertex3f(#,#,#); and they render to the screen just fine.
However, I now want to implement a z-buffer algorithm and I need integer values to represent scan lines. I am faced with converting my coordinates (floats) to another coordinate system (integers) in order to implement the algorithm and then transform them back to the NDC.
It is possible that I mixed up. I am new to win32/opengl.
There is no reason to use OpenGL at all if you are writing a software renderer. You are just making things (much) more complicated that way. So either use OpenGL the way it was intended to be used (ie. as a hardware accelerated rasterizer), or drop it alltogether. Just use your software renderer to draw to a memory buffer and blit that into a window.
Author
Yes, of course you are right. Sorry for the confusion. Unfortunately this is the assignment, OpenGl is required, so I will just do a couple of transformations between coordinates and be sad about the inefficiency.
Thanks for the tips.
Hornsj2
Thanks for the tips.
Hornsj2
So you were told to draw an OpenGL point for each of your software rasterized pixels ?! No offense, but what kind of ridiculous assignment is that supposed to be ? If so, then I would seriously question the competence of whomever gave you this assignment.
If you really have to use OpenGL to write a software renderer, then at least have your software rasterizer render into a memory buffer, upload it as a texture to OpenGL, and render it as a screen aligned quad.
If you really have to use OpenGL to write a software renderer, then at least have your software rasterizer render into a memory buffer, upload it as a texture to OpenGL, and render it as a screen aligned quad.
Quote:Original post by Yann L
No offense, but what kind of ridiculous assignment is that supposed to be ? If so, then I would seriously question the competence of whomever gave you this assignment.
LOL! My thoughts!
Author
Outstanding, so my instructor gave me a bad assignment. If anyone else has a LOL, PST because I find that kind of input extremely valuable.
Hornsj2
Hornsj2
I think what they're trying to teach is the transformation pipeline. I'd done this a while back when I was in school as well. When I did that assignment there's a drawpixel function that take care of putting the pixel on the screen for a given screen coordinate. IIRC the function doesn't use glVertex*, I have to look at the source later. If you still need it, I can look it up later today.
I found out what it was when I had the class. All they did was create a canvas (a buffer of screen width x screen height in size of pixel data (rgb) ), so when you plot a pixel it just set the correct color at the correct location, when it's time to render then you call glDrawPixels and pass in this canvas. Hope that helps.
[Edited by - nhatkthanh on July 15, 2008 12:11:51 AM]
I found out what it was when I had the class. All they did was create a canvas (a buffer of screen width x screen height in size of pixel data (rgb) ), so when you plot a pixel it just set the correct color at the correct location, when it's time to render then you call glDrawPixels and pass in this canvas. Hope that helps.
[Edited by - nhatkthanh on July 15, 2008 12:11:51 AM]
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
