Geek Answers Handbook

Perspecitve divided into vertex shaders?

When using the perspective matrix in the vertex shader, should I write code to divide by w, or is this done automatically at a later stage?

The reason for my question is that I saw a lot of vertex shaders using:

gl_Position = matrix * pos;

which makes sense if there is a later stage that divides the vector with its w-component.

However, I never worked until I used the following in my vertex shader:

gl_Position = matrix * pos;
gl_Position = gl_Position / gl_Position.w;

Is the second example correct or may some other setting be missing?

Put another thing: which of the steps shown in OpenGL Vertex transform (first image) do I need to write to the vertex shader

I know for sure that the matrix ModelView and Projection belong (or merge the two). Converting a viewport is not part of the vertex shader, but what about divide by w?

My setup is simple triangles with coordinates inside [-1 1] for x / y / z.

The prospective Prospect matrix should project the coordinates from z = -1 to -10 on z = -1, x = [- 1,1], y = [- 1,1].

-1.0   0.0   0.0   0.0
 0.0  -1.0   0.0   0.0
 0.0   0.0  -1.2  -2.2
 0.0   0.0   1.0   0.0

It was generated:

x = 2.0f * zNear / (xMax - xMin);
y = 2.0f * zNear / (yMax - yMin);
a = -(xMax + xMin) / (xMax - xMin);
b = -(yMax + yMin) / (yMax - yMin);
c = (zFar + zNear) / (zNear - zFar);
d = -(2.0f * zFar * zNear) / (zNear - zFar);

To make the matrix P:

x, 0, a, 0
0, y, b, 0
0, 0, c, d
0, 0, 1, 0;

Finally, I create the final matrix with the matrix = P * T, where T is the shift (0,0, -2)

I tried to do the math on the processor and it seems to work with expected results, however there I also do split w manually.

Update : solved, but you need to understand

( -1), . , z- , .

, , .

+3
1

, .

, - ? , .

+4

More articles:


All Articles