Open GoogleCodeExporter opened 9 years ago
Eye W also affects the shadows strangely. If the horizon were exactly in the
center,
it would look better but still not right. I think that we're missing is a proper
perspective transform in which everything approaches a vanishing point in the
center
of the horizon. Or maybe even some other point on the horizon would make for an
interesting effect, but as it is, the shadows are often not believable.
Original comment by cutelyaw...@gmail.com
on 4 Nov 2009 at 5:41
I think we'll have two options to make the behavior correct:
(1) We'll have to transform the horizon along with the scaling (the horizon
vertical
position would not in general be at the same location). The auto view scaling
would
probably do some pretty wild things to the horizon here though.
(2) We can merge what is the current 2D scaling factor into the 4D view matrix
of
the rotation handler, so that the scaling is done before shadows are
calculated.
This is my preferred choice. The downside here is that a big zoom might
actually
make portions of the puzzle such that they should be below the ground, and I
doubt
we're handling drawing of that correctly.
Original comment by roi...@gmail.com
on 4 Nov 2009 at 5:34
Original issue reported on code.google.com by
roi...@gmail.com
on 30 Jul 2009 at 4:51