btw, this is in away3dlite. Rotation on the ortho lens doesn't effect the _uvtData at all. The t is always 1. Can anyone confirm that this is the way ortho cams work in general, or is it a missing feature or something? When I manually set the t value to something else, at least I can affect the depth sorting...
it's basically this line in Mesh class: Utils3D.projectVectors(_viewMatrix3D, vertices, _screenVertices, _uvtData); And whatever the resulting _viewMatrix3D is when using ortho lens, will not give proper values for t. -Ken On Jun 9, 2:19 pm, kensuguro <[email protected]> wrote: > thanks for the reply, > just doubled checked distance, but it doesn't seem to have any effect > in this case. > > On Jun 9, 1:43 pm, Bob Warfield <[email protected]> wrote: > > > > > Kensuguro, take care how close your camera is to the objects. I discovered > > a lot of weird artifacts if my camera is too close. In particular if > > objects visible in the foreground can get significantly behind the camera, > > bad things happen. This is apparently a known problem. > > > If your application allows it, try moving the camera further away while > > increasing the zoom to keep the same relative proportions. It may help. It > > did wonders for me. > > > Cheers, > > > BW > > > On Wed, Jun 9, 2010 at 9:04 AM, kensuguro <[email protected]> wrote: > > > I'm not familiar with the mathematical implications of an orthogonal > > > lens, but can it not be rotated? It seems like when I rotate it, the > > > z-order sorting gets all messed up. > > > >http://groups.google.com/group/away3d-dev/web/Picture%20307.png > > > > see how the water behind the rocks is in front of it.
