How to Transcribe Audio Files or YouTube Clips with Voice Notebook

Updated

To transcribe audio or video with Voice Notebook, open the Transcription panel with the Transcription button, supply the source (an HTML5 audio/video URL, a local file from your computer, or a YouTube record ID), then press Start recording. The recognized text appears in the output field. This works in the Chrome browser on Windows, Mac, and Linux; Android and iOS users need the separate mobile apps.

Before you start

  • Browser: Voice Notebook currently works only in Chrome, on Windows, Mac, or Linux.
  • Microphone: The site strongly recommends a good external microphone. That matters for live dictation; for file transcription the audio itself is the input, but the same recording control is used.
  • Source ready: Have one of the following on hand — the direct URL of an HTML5 audio or video clip, the file saved on your computer, or the YouTube record ID of the clip you want.

Step-by-step transcription

  1. Open the panel. Click the Transcription button. It shows or hides the audio recognition panel, so click it again if the panel is hidden.
  2. Choose your source.
    • For an HTML5 audio or video clip, enter its URL.
    • For a file on your machine, pick the file from your computer.
    • For YouTube, enter the YouTube record ID (not the full watch URL).
  3. Start recognition. Press Start recording. The application recognizes the speech embedded in the clip.
  4. Check the output. The transcribed text lands in the output field, where you can edit it.

Working with the transcribed text

The output field is not read-only — the on-screen buttons let you clean up results as they come in:

Button / option What it does
A/a Changes the case of the first letter of the word nearest the cursor
Undo Removes the last entered sentence from the output field
Punctuation buttons Insert punctuation marks into the text
Replace punctuation checkbox Replaces spoken punctuation words with the corresponding marks (check the button help texts for the matching words)
Execute voice commands checkbox Enables voice commands
Transfer to clipboard checkbox Sends spoken sentences straight to the system clipboard instead of the output field; available when the Voice Notebook Chrome extension is installed

One caveat on punctuation: Google now automatically replaces "period" (or "full stop"), "comma", "question mark", and "exclamation mark", so the Replace punctuation checkbox does not affect those four words.

Changing the recognition language

If the clip is not in the default language, pick the right one from the drop-down menu before you start. Alternatively, register an account and add the desired speech input language in your user account.

Common snags

  • Nothing happens on the first click of Start recording. The first time you press it, you must confirm the action in Chrome's popup toolbar. Allow it, then press again.
  • YouTube link rejected. Use the record ID rather than the full URL.
  • Panel is missing. It is toggled by the Transcription button — it may simply be hidden.
  • Wrong words throughout. Check the recognition language setting first; a mismatch is the usual cause.

Beyond single files

The same speech-to-text engine powers live dictation: press Start recording and speak into the microphone. Installing the Chrome extension lets you dictate into any text field on any website, and the Windows, Mac, or Linux integration module extends voice input to any application on those systems — useful if transcription is one part of a larger dictation workflow.