Combine that with a speech analyzer and synthesizer and you could theoretically even eliminate the voices of individual people (particularly that annoying person who sits near you in a cafe/plane/work).