Thanks, I did it by hand. Emacs+Amarok. While I worked on this, I Googled for transcription software, but couldn't find any packages or readily available POSIX pipelines that I could use from Ubuntu, let alone web services that don't require a credit card. Maybe I didn't look very carefully. Took two days to complete, though. This is something that should be easily doable now, so here's my personal RFS: fix transcription software. A user should be able to either upload an audio file directly, use a Dropbox folder, or provide a link to an s/OGG/MP3/whatever, and get a text file in return. They can then listen to the audio while they make whatever edits that they need.