I have a project idea, that is very well needed in our church, but I wonder if it would be interesting for wider circles. We often need to transcribe (i.e. turn audio into text) of sermons, interviews, conferences etc for our websites our newspapers. Usually volunteers help out with it. Current workflow is:
- Convert some weird audio format into mp3
- Split mp3 into, let's say, 10 minute pieces
- Distribute those pieces among volunteers
- Volunteers have to download some audio player that supports keyboard shortcuts, slowing down speed and setting volume to more than 100%
- Track progress and then collect transcribed texts from volunteers
- Stich those texts into one file and proofread it, edit it etc
You get the point... Perfect place to introduce some automation: I want all of this to be automated in an online service.
- You upload a file and select group of people who get access to this file (like in Google Drive).
- They immediately see the list of fragments available for transcription and can choose one. As soon as they do so, the status of the fragment changes from available to taken.
- They are presented with a nice web audio player with keyboard shortcuts, speed and audio level controls, waveform to see quite places etc.
- They type the transcribed text into textarea (perhaps with bold and italic controls) that is transparently saving in the background. As audio progresses, the percent of complication on main page of that fragment may also be updated to help tracking status.
- When all fragments are done, you can see the whole text and export it to the format of choice. I wanted to implement this on some modern stack (probably React+Redux+Node/Express+Mongo). I hope I'll be able to start working on this project by the end of the year, so if any of you have similar needs, feel free to join, I'd be more than happy (edited)