A simple, no-nonsense online tool that helps you transcribe audio files. Run it on your own machine if you like. Here’s a nice audio file from NASA to test it with (download it first, then drag onto the tool).

• Multiple formats (wav, mp3, m4a, flac, possibly more).
• Play/pause with esc so you never have to take your hands off the keyboard to control the audio playback. Also, resuming after pause skips back a few seconds, so you can re-listen to short sequences easily without having to set up a loop.
• Proper loop function for tricky regions.
• Slow down the audio if it’s difficult to understand.
• Speed up the audio if the speakers are too slow.
• Transcriptions for each file are stored in-browser, on that one device. If you come back and drag in the same file again, your transcription will appear again.
• Saves your work in-browser on every keystroke.

• No signup or login.
• No file management.
• No logs.
• No tracking.
• No backend. I don’t get your audio or your transcriptions.
• No business model.
• No guarantees. Save your work.

• Long audio files could be handled better:
  • The timeline is too cramped. It’s possible to override the timeline renderer, but I haven’t had time for this yet.
  • The loop function becomes less useful because it’s not possible to make short loops without zooming in. That said, there’s an experimental branch with zooming, but the performance is pretty bad with longer files (which is where it would be most useful), with CPU and memory use rising to problematic levels.
• Sometimes, dropping a file on the drop area won’t be registered correctly, and the browser will open that file instead.

Just run a http server in the project root. If you’re on a Mac, you can just do python -m SimpleHTTPServer. Alternatively, use this super-nifty autorefreshing static server from npm: npm install -g live-server, then live-server.

There’s no package management, no build pipeline, no css compilation, no autoprefixing etc. Some stuff comes from CDNs. This will probably improve so you can use this offline in the future.