LANG: cameras

Michael Pichler ([email protected])
Tue, 10 Oct 1995 17:52:24 +0100


I have two questions concerning cameras, which are somewhat
hypothetical. Most (all?) scenes I have seen up to now either define
no camera at all (where the browser should generate a "useful" view),
or define a single camera (possibly in a Switch to select one out of a
list) before any geometry.

1. What happens on multiple cameras? I mean a sequence of a camera,
some geometry, another camera, further geometry. The spec says that
cameras are affected by the current transformation.

Does this mean a camera is essentially a Rotation by the negative
angle and a Translation by the negative position and an adjustment of
the height resp. height angle (orthogonal/perspective projection)?
Further cameras will do it the same way?

2. What happens when geometry is defined before the first camera? Does
one manipulate the first (active) camera or may the browser transform
the whole world coordinate system?

The first case would yield a fixed coordinate system for those nodes,
in the latter case they would move around as well.

Michael


  • Next message: Jan Hardenbergh: "Re: ^Z character in VRML files"
  • Previous message: Chris Marrin: "Re: Behaviour Proposal + the VRML format"