Dude,

I use speech recognition software (Dragon fwiw – which I shadow the speakers with) and a hacked Emacs transcribe mode to get intereviews transcribed in around about 3x the length of the interview, maybe slightly less. Contact me if you want more details.