1> Both parties use headphones, preferably in-ear or other closed headphones which do not let much noise back into the room.

2> Both parties run recording software which records what is audible through their computer microphone. Quicktime will do that out of the box.

3> The interview is conducted over Skype or a similar tool.

At the end, both parties have a recording of their half the interview in reasonably high quality, direct from the microphone. The other party’s voice isn’t audible, because you were listening to it over headphones.

You should also have Skype record the call if possible (Gizmo Project, a skype competitor, will do this without a plug in.)

To get the broadcast mix, take the two audio tracks, put them on top of each other using, say, Audacity (a free audio tool) and you have what sounds reasonably like two people in a room talking to each other, rather than one person interviewing another over a crackly VOIP line.

I dunno if this has a name, but if not, I dub it “double source podcasting.”