also, AFAIKT, unless you can cleanly map the frequencies in soundA to the frequencies in soundB, you aren't going to get anything much different than a simple crossfade of the sounds. usually, mapping those frequencies requires that you do it by hand.
pretty sure you can do it with fourier transformations...there is lots of stuff in the pd documentation about that.