Hi Alessandro, On 30/12/09 11:32 AM, alessandro terenzi wrote: > By the way, what is the general approach to RTT? I had a look at some > examples some time ago, but it could be useful to have a guide line > here...as far as I remember, RTT requires a camera that sees my scene > and that will be used to generate a texture, then I'll need another > different camera that actually will be used to render that texture, correct? > > what the texture will be, in my case that camera will be the one that > I'm currently using to render my scene, but then I'll need another > different camera that actually will be used to render the previous > texture, correct?
Yes, that's correct. You need to have two cameras: the scene camera, that is setup like your original camera (view and projection matrices) but renders to a texture instead. Best case you can just take the original camera and change its render target (and detach it from the viewer). The second camera renders (with ortho projection) a simple scene with a quad that displays the generated texture. > How about HUDs, if I'm not wrong they also use a different camera so > performing RTT on the scene will keep HUDs out, am I wrong? HUDs are usually child nodes of the scene camera so should be rendered without problems. One issue that you might run into are manipulators since they could attempt to modify the second camera instead of the scene camera. HTH, /ulrich _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
