Skip to main content

The main screen

A screenshot of the main screenA screenshot of the main screen

The main screen is where you'll spend most of your time. Voskle Live Transcribe is designed to give immediate access to the most important functionality without cluttering the interface too much. The following user interface elements are present on the main screen:

  • Language picker dropdown,
  • Settings button,
  • Keyboard input switch,
  • Transcript text area,
  • Start/stop transcribing button,
  • Clear transcript button
  • Status bar.

Language picker dropdown

The language picker dropdown is the element that receives focus when you start the app. It is labelled "Select language" and displays the currently selected language as well.

note

The language picker button is disabled while downloading a language model.

Learn more about how to select a language.

Settings button

While using TalkBack, the settings button is the next item that receives focus. Clicking it opens the settings dialog.

You can learn more about the available settings in the documentation for the settings dialog.

Keyboard input switch

The keyboard input switch is set to "off" by default. When you toggle it on, the transcript text input becomes editable and transcribing speech gets disabled. If transcription was running when keyboard input got enabled, it resumes automatically once keyboard input got switched off again.

note

If you have an external keyboard connected to your device, only the left and right arrow keys work the way you'd expect them to. Support for up and down arrow keys as well as page up and page down didn't get implemented yet.

Transcript text area

The transcript text area is usually not editable. When transcribing, partial results of the speech recognition get inserted at the end of the transcript in real-time. This means that a word or part of a recognized sentence may change if the speech recognizer thinks it found a better result. Once recognition of a particular snippet of speech got finalized, a period gets inserted. If speaker recognition is enabled and the current speaker changed, a speaker label gets added at the start of the recognized snippet where the speaker changed. If keyboard input is enabled, the transcript text area becomes editable and you can add or delete text as you please.

note

Android currently has a known issue leading to the transcript text area not getting recognized as a text field by TalkBack. This also means that you cannot enter text using the braille keyboard. We're in touch with the relevant Android support and hoping this issue gets fixed soon.

Start/stop transcribing button

The start/stop translation button starts and stops the speech recognition and transcription process. The first time you click it, microphone permissions will get requested so that speech can get recorded and transcribed. The button will be disabled if one of the following conditions applies:

  • No speech model got loaded,
  • The app is downloading a speech model,
  • Keyboard input is enabled or
  • Microphone permissions got denied.

Clear transcript button

The clear transcript button deletes all text from the transcript text area. It is disabled if there is no text in the transcript text area.

Status bar

The status bar gives you information about the current state of the app. When the state changed, TalkBack will automatically report that change. The following states exist:

  • No model loaded: No language model got loaded; transcribing isn't possible.
  • Downloading model: A speech model is getting downloaded; transcribing isn't possible.
  • Processing model: A speech model is getting processed and prepared for usage; transcribing isn't possible.
  • Ready: A model is loaded and the app is ready to transcribe.

While a model is getting downloaded, a progress indicator will get shown next to or underneath the status bar and TalkBack will speak the progress in percent if you navigate to that progress indicator.