Skip to content

Self-contained multilingual TTS speech synthesizer for Node.js in pure js

License

Notifications You must be signed in to change notification settings

abbr/text2wav.node.js

Repository files navigation

Build Status

text2wav.node.js

Self-contained multilingual TTS speech synthesizer for Node.js in pure js.

Why

There are not many self-contained FOSS speech synthesizer npm modules. Most TTS npm modules are just a wrapper to an external online service or a separately installed host program. For those self-contained FOSS modules, they are either designed to run in browser to play out the speech rather than output the audio file, or lacking multilingual support.

text2wav.node.js is arguably the first, if not the only node.js module, to support all of following features

  • convert text to wav
  • support over 100 languages and accents out-of-the-box
  • allow input to be UTF-8 encoded text in native languages
  • self-contained FOSS, not even depending on other npm modules
  • 100% Javascript/Typescript

How

Install

npm i -S text2wav

to use API or

npm i -g text2wav

to use CLI.

Use

Synopsis

  • API

    • Javascript
    const text2wav = require('text2wav')
    let out = await text2wav(<text>[,<opts>])
    
    • Typescript
    import text2wav = require('text2wav')
    let out = await text2wav(<text>[,<opts>])
    

    where <text> is string and <opts> is a json object. Returns wav content of type Uint8Array.

  • CLI

    text2wav <text> [<opts>]
    

    Ouputs wav content to stdout.

<opts> in API or CLI may contain following parts

API CLI type default description
voice -v string en Use voice file of this name from /espeak-ng-data. A variant can be specified using <voice>+<variant>, such as af+m3. A list of variants can be found here.
amplitude -a integer 100 Amplitude, 0 to 200.
wordGap -g integer Word gap. Pause between words, units of 10ms at the default speed.
capital -k integer Indicate capital letters with: 1=sound, 2=the word "capitals", higher values = a pitch increase (try -k20).
lineLength -l integer Line length. If not zero (which is the default), consider lines less than this length as end-of-clause.
pitch -p integer 50 Pitch adjustment, 0 to 99.
speed -s integer 175 Speed in words per minute.
encoding -b <1|2|4> 1 Input text encoding, 1=UTF8, 2=8 bit, 4=16 bit.
hasTags -m boolean false Indicates that the text contains SSML (Speech Synthesis Markup Language) tags or other XML tags. Those SSML tags which are supported are interpreted. Other tags, including HTML, are ignored, except that some HTML tags such as <hr> <h2> and <li> ensure a break in the speech.
noFinalPause -z boolean false No final sentence pause at the end of the text.
punct --punct[=] string or boolean false Speak the names of punctuation characters during speaking. If =<string> is omitted, all punctuations are spoken.

Basic Usage Example

  • API

    (async () => {
      const text2wav = require('text2wav')
      let out = await text2wav('test')
      // out is of type Uint8Array
      const assert = require('assert')
      assert.equal(out[0], 82) //R
      assert.equal(out[1], 73) //I
      assert.equal(out[2], 70) //F
      assert.equal(out[3], 70) //F
    })()
  • CLI

    text2wav test > test.wav
    

    outputs wav to file test.wav

More examples

  • API: let out = await text2wav('测试', {voice: 'zh'})

  • CLI: text2wav '测试' -v zh > test.wav

    TTS in chinese

  • API: let out = await text2wav('test', {voice: 'en+whisper'})

  • CLI: text2wav test -v en+whisper > test.wav

    use whisper variant

  • API: let out = await text2wav('"test", I say.', {punct: '"'})

  • CLI: text2wav '"test", I say.' --punct='"' > test.wav

    speak the punctuation "

Under the hood

text2wav.node.js compiles espeak-ng into Javascript and WebAssembly libraries using Emscripten. text2wav.node.js also includes compiled voice data from espeak-ng. The text2wav.node.js API or CLI takes the text and voice parameters as input, and generates the wav ouput.

To compile

git clone https://github.com/espeak-ng/espeak-ng.git
cd espeak-ng
./autogen.sh
./configure --without-async --with-extdict-zh --with-extdict-zhy --with-extdict-ru
make
cd src/ucd-tools/
./autogen.sh
make clean
emconfigure ./configure
emmake make
cd ../..
emconfigure ./configure --without-async --without-mbrola --without-sonic --enable-shared=false
emmake make src/espeak-ng
mv src/espeak-ng src/espeak-ng.bc
emcc -s ERROR_ON_UNDEFINED_SYMBOLS=0 -s MODULARIZE=1 -s 'EXPORT_NAME="EspeakNg"' -o espeak-ng.js src/espeak-ng.bc
echo // espeak-ng git hash `git rev-parse HEAD` >> espeak-ng.js

The above commands generate

  • espeak-ng.js file
  • espeak-ng.wasm file
  • voice data files in espeak-ng-data folder

Copy espeak-ng.js and espeak-ng.wasm to text2wav.node.js /lib folder and espeak-ng-data to /. Insert the content of file /lib/post.js to /lib/espeak-ng.js below the line // {{MODULE_ADDITIONS}}.

Contribution

Contributions are welcome. All compiled code and data will be generated from espeak-ng by text2wav.node.js author. Please do not submit any PRs containing the compiled code or data, i.e. most files in /lib or /espeak-ng-data folder. To contribute to those files, submit PRs to espeak-ng instead. Once the PRs are accepted, notify text2wav.node.js author to recompile.

License

The MIT License (MIT)

Copyright (c) 2018-Present, @abbr

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

About

Self-contained multilingual TTS speech synthesizer for Node.js in pure js

Resources

License

Stars

Watchers

Forks

Sponsor this project

 

Packages

No packages published