A process that we use (found also used by BBC in recent visit) is to listen to file and speak into recognition software. This can be highly accurate if trained well on one voice (particularly male voice). Also pick up a copy (free) of Transcriber (http://www.etca.fr/CTA/gip/Projets/Transcriber/) -- a good tool for transcribing since it gives you fine control over audio and links to transcript. Final product is xml so you can spit it out to do many things. dean Dean Rehberger Associate Director of Matrix Associate Professor Michigan State University 310 Auditorium East Lansing, MI 48824-1120 rehberger@mail.matrix.msu.edu matrix.msu.edu/rehberger wk: (517) 355-9300 fax: (517) 355-8363 hm: (517) 347-7372 Dean Rehberger Associate Director of Matrix Associate Professor Michigan State University 310 Auditorium East Lansing, MI 48824-1120 rehberger@mail.matrix.msu.edu matrix.msu.edu/rehberger wk: (517) 355-9300 fax: (517) 355-8363 hm: (517) 347-7372