In OpenGL, Why we are located at (0,0,0) by default in the world space?

user_14574973 0
Viewed 231

Quoting "https://learnopengl.com/Getting-started/Coordinate-Systems"

Next we need to create a view matrix. We want to move slightly backwards in the scene so the object becomes visible (when in world space we're located at the origin (0,0,0)). To move around the scene, think about the following:

This confused me, because before I know any concepts about these space, I can simply draw a triangle which all vertices has a z-coordinates equals to 0, and I still can see it(If I was located at (0,0,0), I suppose not be able to see it)

opengl
1 Answers

One of the aspects that seems to confuse a lot of beginners of OpenGL (or computer graphics in general) is, that there not really actually is such a thing as a camera.

The effects of a camera are implemented by simple geometrical transformation. The effect of perspective, i.e. lines vanishing in the distances is implemented through what's called the "homogenous division": The vectors used in 3D graphics contain actually 4 values, the 4th usually being 1. This allows to coalesce translations and rotations into a single 4×4 transformation matrix.

What it also enables is, to have the transformed Z value also be mapped into the 4th, the W coordinate.

As a last step, that's hardwired in fact, all the coordinates of the transformed vector are divided by the W element. Hence if you have Z values other than 1, this creates a scaling on the X and Y components, that looks like vanishing lines.

To enable this, the transformation matrix must have nonzero elements in the W-row, Z-column, in order to get values other than 1 in the W element.

The default identity transformation will keep the transformed W at 1, hence no homogenous divide happens.

Last but not least, what's visible is everything inside the clip space volume. And with identity transformation, that's the volume [-1;1]³

user_5243680
Related