2. Split the output three ways.
3a. Run one through a sub-audible-frequency low-pass.
3b. Run another through a super-audible-frequency high-pass.
3c. Run the third through an audible-range-frequency bandpass.
4a. Upshift the output of the low-pass.
4b. Downshift the output of the high-pass.
5. Merge the upshifted audio, the downshifted audio, and the bandpassed audio.
6. Feed the merged audio into headphones.
Note: For best results, fit a mic into each headphone, use separate processing for each, and feed processed audio into the apropriate side of the headphones.