azure text to speech saver

azure-text-to-speech-saver

Microsoft Azure Text to Voice Downloader

Scripts are based on aardio programming language

tts.aardio, free interface

~~The text-to-speech cloud service on Azure has a closer effect to the real-person pronunciation, and has tone and breathing sound. Azure demo web page can only be listened to online. If you want to download and use it again, there are currently two solutions on the Internet:

1. Recording method: The disadvantages are obvious. Global recording is easily disturbed and must be played before the data can be obtained in full.
2. js injection method: The existing injection method is to rewrite SpeechSDK using the oil monkey script to obtain the data synthesized by the front-end. You don’t need to listen to the audio, and you can download it after the data transmission is completed.

The solution adopted in this example is similar to the second one, which is also an injection method. However, due to the limitations of the oil monkey script, the code of the second one is a little complicated. However, using aardio's webview library to rewrite the websocket, making the code more concise and maintainable.

Solution Overview:

Cloud services are segmented and transmitted to the front end through WebSocket. We can rewrite the WebSocket library, change the listening "message" event, and pass the data to aardio for processing. ~~