Possible to use Python to do text-2-speech and output speech as if it's coming from the microphone?

Question

I was looking at the Python speech libraries in this post. I'm in a quiet office environment and I have to speak to people on a handheld phone but I'd rather not disturb people around me if I don't have to.

I feel that if I typed on a keyboard on a computer and my text was converted into speech and I could feed that speech into the port of a microphone (Virtual?) then I'll be able to communicate simply by typing and they can hear a robot voice on the other end.

For clarity: I don't want to use the phone. I want to use the computer. I want to type on the keyboard that is connected to the computer and have that text converted into speech and feed it into the microphone port of my computer so it can send that computerized voice into the GoToMeeting application. This way it looks as though I'm speaking into the microphone but I'm not, I'm just typing. I want the computer to think I'm talking into the microphone in order to have other applications e.g. GoToMeeting be able to interpret it.

And the GoToMeeting chat client is useless due its default setting of hiding the chat. Many people ignore the chat unless you're speaking.

My preference would be some kind of library that would allow me to create a virtual microphone so I could write audio to it. I would ultimately want to run this in Linux with a Windows VM running GoToMeeting. This way I can continue to use Linux, type in the text, and pipe in the speech into the VM so GoToMeeting users can hear me.

Thanks

Answer 1

You need to write a driver for a virtual audio device, then you will be able to pass audio to the driver and it will be available through a virtual microphone.

Microsoft publishes the source code for a Virtual Audio Driver as part of the Windows Driver Development Kit. If you have access to Visual Studio, or if you know a programmer who does, compiling this virtual driver and installing it allows you to create a required device.

Then you need to write a tool using any TTS library that will take the text you need and pass it into the driver for loopback.