- distinguish between parallel and perspective projection
- introduction to clipping
- Camera viewer:
→ the image plaine is a window to the 3D world
→ 3d objects are projected on the 2d image plane
→ we're going to be setting the camera's lens
- requires:
→ one or more objects, something to look at
→ a projection surface (imagep lane)
→ a camera that projects 3d objects on a 2d surface
- parallel ( orthographic)
→ makes everything look flat but is super precise
→ without depth cues it can be hard to tell where things are, because when we're looking at 3d models on a 2d plane, any pixel can be anywhere between our eyeball and infinity unless we have some kind of depth indicator
- perspective
→ looks more like our vision
- projectors:
→ lines that either converge at the center of projection are are all parallel (orthographic)
→ these projectors preserve liens but not necessarily angles, so if you have a right angle, and your objects right angle is not parallel to the projector, it will probably not look like a right angle
→ it's like the point is leaving a stain on the image plane as it passes through
Properties of perspective
- diminution
→ something farther away appears smaller
- foreshortenting
→ equal distances along a line are not projcted into equal distances on the image plane
- andgles are preserved..... soemtihng
- orthographic projection
→ projectors are orthogonal to projection surface
→ onyl really matters what is in front of the other
→ what we've been using so far
→ preserves both angles and distances
→ shapes are fully preserved
Perspective Clipping
- stricly an efficency measure
- we could use it if we're doing scientific visualization... we could use a clipping plane to slice into something
- maybe some kind of game..
- we are often only interested int a limted volume of space to render
- in perpsective mode, the viewing volume is like a truncated pyramid.
- you cut off the tip and the tail and only render what's in between
- we want to cut off the fat as early as possible in the rendering pipeline
→ right after we position our camera and objects
→ before we do any lighting or whatever
→ as soon as we know where things will be and what wont be visible
- in orthographic mode its basically a rectangular box
→ we say between xmin and xmax, we're gonna see stuff, between ymin and max, we're gonan see stuff, and between near and far distance we will see stuf. this is what we define when we call glOrtho2d()
→ with ortho
Summary of ortho
- projector liens are parallel
- no center of projection
- positioning
→ equivalent to positioning the imahe plane
→ positioning camera done with gluLookAt
- clipping volume is a rectangular box
- near and far always measured from the camera
summary of perspective
- projector liens go through image plane to center of projection
- positionaing the camera is like positioing the center of projection
- clipping volume is a “frustum” truncated pyramid
- near and far clipping distances measured from the camera
- other stuff i didnt manage to copy down
Index