Hi there! Until now I created impostors in a fast/dirty way, which looked OK. But now I want to do that the RIGHT way! What I did so far was to place the camera some units away from the object, render it to texture and then put that onto an axial-bilboard. The problem here is that the original object and the pre-rendered one don''t look the same; the bilboard-image is a little diplaced/stretched etc... So I''d like to know more about the way the camera and the object have to be placed/orientated when rendering the texture. I''m also not quite sure what size (X/Y) the billboard should have later on. And: are these settings object-dependend? Or have they to be calculated for every object? Thanks for your help!!!

Huh? Nobody?

Maybe I explained the problem not clearly enough?

Lets say I have a camera (Position, LookAt-vector etc.) and an object (vertex cloud, bounding box etc.). And now I want to exchange the "real" object by a impostor (billboard).
What is the right setup for this render-to-texture-scene?

Where do I have to place the camera, where the object? How do I have to orientate the camera? And how big has the billboard-quad to be?

And can I use these settings for every object, or is there a formula to calculate these things (for every object) ???

Thanks for your help!

Allright. If nobody knows an answer, then maybe someone got a link to a paper or something? I didn''t find a good one on this topic...

I don''t think that that''s so simple! (Or is it???)

I have to place the camera near the object, so that the obj. covers the whole screen when rendering the texture. But I don''t kwo how exactly...