# How is orthographic projection used in computer graphics technically classified as a projection?

I'm wondering why the orthographic projection we use in computer graphics is called a projection, if all it does is just scale and translate a mesh?

According to wiki orthographic projection is Orthographic projection (sometimes referred to as orthogonal projection, used to be called analemma[a]) is a means of representing three-dimensional objects in two dimensions, but we don't use orthographic projection to project our mesh to 2d space, we just use it to transform our mesh defined in the view box with arbitrarily defined dimensions (left, right, up, bottom, far and near) to a view box of (-1, 1, -1, 1, -1, 1) dimensions. So why do we call it an orthographic projection?

• It's technically just the matrix used to prepare for the orthographic projection, the actual projection happens by dropping the $Z$ coord. Commented Jul 8, 2020 at 7:01

I think it is more of a terminology shared specifically in the computer graphics community. The transformation $$P$$ results in a vector one-step-ahead from getting the actual projection. From a 3D position transformed with $$P$$, we can obtain both

1. the 2D-projected position in normalized device coordinate (NDC) (by obtaining xy component)
2. the depth (by obtaining the z component) in NDC

This so-called projection matrix is designed specifically to satisfy the needs in the computer graphics pipeline. The key difference between an ordinary N-D-to-(N-1)-D projection is that the projection matrix results in 2D positions with "depth", which is required for obtaining a depth map used for many purposes in the computer graphics pipeline.

• Thanks! I would imagine this would cause some confusion for beginners in computer graphics. It would be nice if it was called something else. Commented Jul 8, 2020 at 17:29