I made some changes on the original work by Tom Erbe. It is supposed to be more efficient because the FFT on the input is only done once! I tried the idea of having each partition work with FFT saved on tables, so we wouldn't need to perform FFTs in different instances of clone, but that doesn't seem to be possible. And, well, it is made as an abstraction and all you need to do is give it the window size and the IR sound file name...

cheers