github.com/lesleyrs/clipboard-narrator

Clipboard Narrator

Turn any web page into an audiobook, works in the background on desktop!

Comparison to Microsoft Edge tts

Edge TTS	Clipboard Narrator
no external application needed if you use Edge	download required
online (even for offline files opened in browser)	offline
many languages available	English only on Windows1
delay each time you start or restart a voice	no delay on Windows2
natural sounding voices	dependant on platform

Windows built-in narrator has a different use case but does use the system voices.

This tool reads from your clipboard after copying text, with the ability to stop a voice by copying 1 word³. This allows you to start and stop a voice without requiring window focus. It auto-saves the text field to a .txt file and settings to .dat file on closing, this means you should close the application before deleting the saves to reset settings. You could also use the text field to take notes.

Inside the application press h to get a tts explanation. Copying the same text back to back does not work, press r for that. The i key can also be used for entering text. Background colour picker is the button on the top left corner.

Bugs (mostly visual) are commented in main.gd

interrupt, queue, and manual modes

This application works differently based on which mode you are in (changed by pressing tab or shift-tab). Exception to this is that recent clipboards over 1 word long³ are saved in 1 to 0 number keys allowing you to interrupt current speech with past clipboards regardless of mode.

• interrupt

Will immediately interrupt current voice with Ctrl-c, and only reads off clipboard. Copying 1 word will stop the voice.

• queue

Will put all clipboards into a queue using Ctrl-c, only reads off clipboard but allows you to add non-sequential text to it. Copying 1 word will pause the voice.

• manual

Ignores clipboard, only reads from what you put in the text editor and then hit speak, space or r.

Footnotes

1. No limit on other platforms, Microsoft has a restriction for loading voices in third party software https://github.com/godotengine/godot/issues/69788#issuecomment-1343912420 ↩

2. Linux seems to be slower at loading system voices compared to Windows https://docs.godotengine.org/en/stable/tutorials/audio/text_to_speech.html#caveats-and-other-information ↩

3. 1 word means a maximum of 1 space in the copied selection. ↩ ↩²