speech-stream

Text-to-speech using streams. Stream configuration data, text. Output audio buffers.

API

The API is very minimal. You just create a transform stream that takes in voice configuration data and outputs

`SpeechStream([default_voice])``

Creates a new SpeechStream and optionally takes a default voice to use. Documentation on the voice options can be found here.

parameters

• [default_voice]: Optional Object containing voice configuration. If this isn't provided, a random is generated using randomvoice.

returns

• SpeechStream: A Duplex stream which takes in objects that contain a message property which is a string, and optionally a voice property which contains the aforementioned voice configuration data.

Example:

var streamArray = require("stream-array");
var makeProp = require("make-prop-stream");
var speechStream = require("speech-stream");
var fs = require("fs");
 
streamArray(["Hello World!"])
.pipe(makeProp("message"))
.pipe(speechStream())
.pipe(fs.createWriteStream("example.wav"));

Note: this example doesn't work well with multiple messages since it'll concatenate the buffer data together. A better use would be to pipe this into something that either saves everything or outputs to your speakers.