If you want binaural audio effects, you will either have to
1) write your own binaural sound processor. (which, yes, generally involves convolution )
2) licensing a binaural engine from a 3rd party
X3DAudio doesn't do binaural audio processing--it just transforms x,y,z into speaker volumes (if you're doing surround sound over speakers).
If you really want to jump in with both feet and do 1, then you will need to find some dummy head HRTF data sets (there are a few around the 'net). Then you'll need to write filtering algorithms.
it's not rocket science, but does require a bit of specialized knowledge to get it right.
You may also want to look into OpenALSoft, which IIRC includes an HRTF implementation.