Voices NPM | npm.io

Contents

voices — TTS CLI for macOS

voices is a macOS CLI for changing the default TTS (text-to-speech) voice and for printing information about and/or speaking text with multiple voices.

voices complements the standard say utility by:

providing a way to change the default voice
operating on active voices - those selected for active use - rather than all installed voices
filtering voices by language
speaking text with multiple voices.

Caveats:

As of macOS 10.12 (Sierra), there is no documented programmatic way to change the default voice. Thus, this utility makes use of undocumented system internals, which, unfortunately, means that future compatibility of this feature is uncertain. Feedback welcome.
voices currently only fully supports voices provided by Apple. Support for third-party voices such as InfoVox iVox is limited to speaking with them, and the macOS Services documented below will not work with them.

See the examples below, concise usage information further below, or read the manual.

Additionally, two macOS Services are offered:

a service for switching between two or more default voices - see below.
a service for speaking selected text with a specific voice - see below

Note: If you have Alfred with its Power Pack, consider workflow speak.awf as a superior alternative.

Examples

  # List all active voices; add -a to list all installed ones.
voices -l

  # Print information about the default voice and speak its demo text.
voices -d -k

  # Print information about voice 'Alex'.
voices alex

  # Make 'Alex' the new default voice, print information about it, and
  # speak text that announces the change.
voices -k'The new default voice is Alex.' -d alex

  # List languages for which at least one voice is active.
voices -L

  # List active French voices.
voices -l fr

  # Print information about all active voices and speak
  # their respective demo text.
voices -l -k

  # Print information about all active Spanish voices and speak their
  # respective demo text.
voices -k -l es
  
  # Say "hello", first with voice Alex, then with Jill, suppressing printed
  # output.
voices -k"hello" -q alex jill

Installation

Supported platforms

macOS

Verified to work from OS X 10.8 (Mountain Lion) up to macOS 10.12 (Sierra).

The change-the-default-voice feature makes use of undocumented system internals, so its future compatiblity is uncertain. Do let me know if you find the feature broken in a future macOS version.

Installation from the npm registry

Note: Even if you don't use Node.js, its package manager, npm, works across platforms and is easy to install; try curl -L http://git.io/n-install | bash

With Node.js or io.js installed, install the package as follows:

[sudo] npm install voices -g

Note:

Whether you need sudo depends on how you installed Node.js / io.js and whether you've changed permissions later; if you get an EACCES error, try again with sudo.
The -g ensures global installation and is needed to put voices in your system's $PATH.

Manual installation

Download the CLI as voices.
Make it executable with chmod +x voices.
Move it or symlink it to a folder in your $PATH, such as /usr/local/bin.

Usage

Find concise usage information below; for complete documentation, read the manual online, or, once installed, run man voices (voices --man if installed manually).

$ voices --help


Get or set or speak with the DEFAULT VOICE:

    voices [<options>] [-d [<newDefaultVoice>]]

LIST INFORMATION about / speak with voices:

    voices [<options>] <voice>...

List / speak with ALL VOICES, optionally FILTERED BY LANGUAGES:

    voices [<options>] -l [<lang>...]

LIST LANGUAGES among voices:

    voices -L [-a]

MANAGE VOICES in System Preferences:

    voices -m

Shared options (synopsis forms 1-3):

    -a          target all installed voices (default: only active ones)
    -k          speak demo text with all targeted voices
    -k"<text>"  speak specified text
    -k-         speak text provided via stdin
    -b          output format: print voice names only
    -i          output format: print voice internals
    -q          quiet mode: no printed output

Standard options: --help, --man, --version, --home

macOS Service for switching between default voices

This service, which uses an embedded copy of voices, is helpful if you use text-to-speech in two or more languages and want to quickly switch the default voice between multiple designated voices cyclically, in combination with the built-in speak-selected-text service.

Every time the service is invoked, the next designated voice is made the default voice, and the localized name of the new voice's language is spoken to confirm the change (this is configurable).

You can invoke the service from any application's standard Services menu, category General, or assign it a keyboard shortcut via System Preferences > Keyboard > Shortcuts > Services.

Installation

Download this ZIP file.
In Finder, open the ZIP file, which creates package Switch Default Voice.workflow in the same folder.
Open Switch Default Voice.workflow and choose Install when prompted - this will place the package in ~/Library/Services/.
Choose Done when prompted and proceed with customization below.

Customization

Invoke the service for the first time to prompt creating and editing the configuration file.
- From any application, open that application's menu and select Services > Switch Default Voice.
- You will be prompted to edit the configuration file, where you can specify the voices to switch between; follow the instructions in the file.
To assign a keyboard shortcut to the service, go to System Preferences > Keyboard > Shortcuts, category Services, scroll to sub-category General in the list on the right, select Switch Default Voice, and click just inside the right edge of the list item.
To customize the service again later, open ~/.SwitchDefaultVoice-rc in your text editor.

macOS Service for speaking selected text with a specific voice

This service provides an alternative to switching the default voice: it speaks selected text in the frontmost application with a fixed alternate voice, which allows it to be used alongside the built-in speak-selected-text service, which always uses the default voice (see System Preferences > Dictation & Speech > Text to Speech).

Typically, you would use this service to speak selected text with a voice that speaks a different language.

You can invoke it from the standard Services menu, category Text, whenever text is selected in the frontmost application, or assign it a keyboard shortcut via System Preferences > Keyboard > Shortcuts > Services; e.g., ⌥` (Opt-`) to parallel the default shortcut for the built-in service, ⌥⎋ (Opt-Esc).

Invoking the service again while text from a previous invocation is still being spoken aborts speaking.
Caveat: This only works if text - any text - is selected in the activate applciation at the time the service is invoked again.

If desired, you can duplicate the service so as to be able to speak with one of multiple alternate voices:
Once installed, duplicate ~/Library/Services/Speak With Specific Voice.workflow in Finder, give it a meaningful name, and customize the duplicate as described below.

Installation

Download this ZIP file.
In Finder, open the ZIP file, which creates package Speak With Specific Voice.workflow in the same folder.
Open Speak With Specific Voice.workflow and choose Install when prompted - this will place the package in ~/Library/Services/.
Choose Open in Automator when prompted and proceed with customization below.

Customization

In Automator, follow the instructions at the top of the document, which currently only require you to specify the name of the desired voice.
- Apply your customizations between the lines # ------- BEGIN: CUSTOMIZE and # ------- END: CUSTOMIZE.
To assign a keyboard shortcut, go to System Preferences > Keyboard > Shortcuts, category Services, scroll to sub-category General in the list on the right, select Speak With Specific Voice.workflow, and click just inside the right edge of the list item.
To customize the service again later, open ~/Library/Services/Speak With Specific Voice.workflow in Automator.
- If you have trouble navigating to ~/Library, activate Finder, hold down the Option key while selecting the Go menu, and select Library; from there, navigate to subfolder Services and open package Speak With Specific Voice.workflow.

License

Acknowledgements

This project gratefully depends on the following open-source components, according to the terms of their respective licenses.

npm dependencies below have optional suffixes denoting the type of dependency; the absence of a suffix denotes a required run-time dependency: (D) denotes a development-time-only dependency, (O) an optional dependency, and (P) a peer dependency.

npm dependencies

Changelog

Versioning complies with semantic versioning (semver).

v0.3.4 (2018-03-21):
- doc Links to workflow ZIPs fixed.
v0.3.3 (2018-03-08):
- fix It is now ensured that the system versions of standard utilities such as awk are called, to prevent unexpected behavior stemming from user-installed versions in /usr/local/bin getting called.
v0.3.2 (2017-01-03):
- doc Limitations of support for third-party voices noted.
- fix voices -m now works on macOS Sierra.
v0.3.1 (2015-11-03):
- doc Added link to Alfred 2 workflow speak.waf as a superior alternative to the OSX services.
v0.3.0 (2015-10-27):
- Potentially breaking change -i for reporting voice internals now reports an extra variable BundleID as the last item, i.e., the voice's bundle ID.
v0.2.3 (2015-09-20):
- doc voices now has a man page (if manually installed, use voices --man); voices -h now just prints concise usage information.
v0.2.2 (2015-09-15):
- dev Makefile improvements; various other behind-the-scenes tweaks.
v0.2.1 (2015-07-30):
- doc Read-me corrections.
v0.2.0 (2015-07-29):
- enhancement voices now honors custom speaking rates when requested to speak with the -k option
- enhancement OSX Service Switch Default Voice.workflow is now configuration file-based and supports more than 2 voices for cyclical switching; default confirmation text spoken on switching is now the localized name of the new voice's language.
- new OSX Service Speak With Specific Voice.workflow allows speaking selected text with a fixed alternate voice.
v0.1.9 (2015-07-28):
- doc Corrected the mistaken claim that changing the default voice also changes the VoiceOver default voice: the VoiceOver feature has its own default voice, separate from the TTS feature; this utility only changes the TTS default voice, not also the VoiceOver one.
v0.1.8 (2015-07-28):
- fix Regression: default customization data for the OSX service reset to original values (two US English-voices).
v0.1.7 (2015-07-28):
- fix When switching default voices, any custom speaking rate (words per minute) configured for a given voice via System Preferences is now honored. Note, however, that speaking text with voices' own -k option does not honor custom speaking rates due to a limitation in the underlying say utility.
v0.1.6 (2015-07-28):
- dev Pre-commit hook fixed to ensure that the modified workflow and ZIP file are added to the index before committing.
v0.1.5 (2015-07-27):
- doc Read-me and CLI help amended with respect to supported OSX versions.
v0.1.4 (2015-07-27):
- enhancement Added Automator-based OSX service for switching between two default voices.
- fix Inability to determine the default voice on a pristine system is now handled more gracefully.
v0.1.3 (2015-07-06):
- doc CLI-help copyediting; wording of --version streamlined.
v0.1.2 (2015-07-01):
- doc CLI help improved.
v0.1.1 (2015-06-30):
- doc Read-me improvements.
v0.1.0 (2015-06-29):
- Initial release.