|
|
Actually, the morphing stuff is some of the simplest code in my entire engine. It has to do with how I'm storing the meshes, which I'm working on a paper on, and so I don't want to give away too much, but this mesh storage mechanism also allows for very fast (and near-optimal) shadowvolumes being put out (takes O(n) time for n polygons, too), as well as all the fun morphing, resampling, and other stuff, such as being able to have wavelet-based compression (which can progressively optimize a mesh as it's downloaded, and some meshes compress down to as little as 4.5 bytes per triangle).
The storage mechanism is also technically a HUGE step backwards in terms of thinking and it's so incredibly obvious I'm surprised nobody else has done it. Which is the other reason I'm being secretive about it for now. :)
Of course, what I *COULD* do is just get around to finishing my damned paper (I've actually been sitting on it for about 6 months now) and submitting it to SIGGRAPH or something and then I could speak freely about it without fear of someone else usurping it from me. But I'm lazy, and the problem with writing academic papers of this nature is finding references to cite. (What might seem obvious to one person could be a completely foreign concept to a reader, which is why papers always have to be so anally-cited - it's not that you're proving that you're not plagiarizing as English teachers would have you think). Unfortunately, I don't know of any papers which go down to the simplistic, fundamental levels to which these techniques go. :) I also don't want to have a tiny number of citations either, though, since then people won't take me seriously as a researcher. Which would suck. (This renderer is hopefully my PhD thesis project, though I still need to clear it with my advisor... I'm sure he'll agree though.)
|
