
			 VOXIN 3.4 README


Table of Contents

1 Check the voxin archive
2 General Install
.. 2.1 System-wide install
.. 2.2 Local install
.. 2.3 Voxin updates
.. 2.4 Speech Dispatcher
..... 2.4.1 spd-conf
..... 2.4.2 Editing speechd.conf
.. 2.5 Orca
.. 2.6 Emacspeak
.. 2.7 Fenrir
.. 2.8 Speakup connector
.. 2.9 How to uninstall Voxin?
..... 2.9.1 System-wide as superuser
..... 2.9.2 Locally as regular user
3 Compatible GNU/Linux Distributions
.. 3.1 Voxin 3.4
4 voxin.ini configuration file
5 Dictionaries
6 voxin-say command


For support or feedback, please send a mail to contact@oralux.org with
possibly the log file (log/voxin.log).


1 Check the voxin archive

  Before any install, you may want to check that your voxin archive (the
  .tgz file) has a correct checksum.

  The checksum file can be downloaded from GitHub under the Oralux
  account:
  https://raw.githubusercontent.com/Oralux/voxin-installer/master/check/3.4/SHA512SUM_3.4

  In the following example, our voxin archive
  (voxin-update-3.4.tgz) is checked against the checksum file
  (SHA512SUM_3.4). These two files have been copied in the same
  directory.

  In a console, change/cd to this directory; then type:
  sha512sum --ignore-missing --check SHA512SUM_3.4

  The previous command must return "OK":
  voxin-update-3.4.tgz: OK


2 General Install

  Voxin can be installed system-wide as superuser or locally as regular
  user.


2.1 System-wide install

  Just run voxin-installer.sh as root and answer to the questions.  The
  packages corresponding to the distribution will be installed.


  For example, with voxin-enu-3.4 :

  cd voxin-3.4
  cd voxin-enu-3.4
  sudo --login $PWD/voxin-installer.sh

  Or without sudo:
  su --login -c $PWD/voxin-installer.sh


2.2 Local install

  By default, voxin will be installed in your $HOME/.local

  ./voxin-installer.sh

  The -d option supplies a dedicated installation directory:
  ./voxin-installer.sh -d path_to_directory

  Note that speech-dispatcher (>= 0.11) automatically calls the local
  sd_voxin module (in ~/.local/libexec/speech-dispatcher-modules/).


2.3 Voxin updates

  Your download link is usually permanent and points on the last voxin
  release.

  To update Voxin, use one of your download links to get the last
  version.  If you have Cerence TTS Embedded and IBM TTS voices, just
  use one of your Cerence TTS Embedded link to update Voxin.

  The releases are announced in the [Voxin RSS feed].

  Voxin is released:
  - to complete the list of compatible distros (usually after user
    request),
  - to add new features and fixes (libvoxin offers paths of
    improvement).

  Besides the voxin release, a lighter tarball, the voxin update is
  available and does not include the text-to-speech package.


[Voxin RSS feed] http://voxin.oralux.net/rss.xml


2.4 Speech Dispatcher

  The Speech Dispatcher voxin output module, sd_voxin, is either already
  provided by your distribution or otherwise will be installed from the
  voxin tarball.

  The voxin-installer will install its own voxin-speechd-conf and/or
  voxin-speechd packages if the voxin module is not already installed in
  your system.

  After your authorization, the installer may check each speechd.conf
  file on your system and add the voxin module if needed.  You will have
  to select voxin via your screen reader (e.g. see the Orca or Fenrir
  paragraphs below) because the installer does not modified the default
  output module in speechd.conf.

  speechd.conf can also be modified using spd-conf or simply edited.
  These two possibilities are described below.


2.4.1 spd-conf
--------------

  spd-conf can be launched as normal user (not root user) in a terminal.

  Here is an example in Ubuntu 18.04.

  spd-conf asks several questions, type voxin as default output module.
  spd-conf will accept it even if voxin is not yet in its list of known
  modules.

  Note: ENTER is written below when the ENTER key has been pressed.

  $ spd-conf

  Speech Dispatcher configuration tool

  Do you want to setup a completely new configuration? [yes] :
  > ENTER

  Do you want to create/setup a 'user' or 'system' configuration [user] :
  > ENTER

  User configuration already exists.

  Do you want to rewrite it with a new one? [no] :
  >yes
  User configuration created in /home/user/.config/speech-dispatcher
  Configuring user settings for Speech Dispatcher

  Default output module [espeak] :
  >voxin
  The value you have chosen is not among the suggested values.
  You have chosen 'voxin'.
  The suggested values are ['espeak', 'flite', 'festival', 'cicero', 'ibmtts']
  Do you want to correct your answer? [yes] :
  >no

  Default language (two-letter iso language code like "en" or "cs") [en] :
  > ENTER

  Default audio output method [pulse] :
  > ENTER

  Default speech rate (on the scale of -100..100, 0 is default, 50 is faster, -50 is slower) [0] :
  > ENTER

  Default speech pitch (on the scale of -100..100, 0 is default, 50 is higher, -50 is lower) [0] :
  > ENTER

  Do you want to have Speech Dispatcher automatically started from ~/.config/autostart ?
  This is usually not necessary, most applications will start Speech Dispatcher automatically. [no] :
  > ENTER

  Do you want to start/restart Speech Dispatcher now and run some tests? [yes] :
  > ENTER
  Starting Speech Dispatcher in user-mode
  [Sat Dec 26 10:20:47 2020 : 338076] speechd: Speech Dispatcher 0.8.8 starting
  Testing Speech Dispatcher using spd_say

  Did you hear the message about Speech Dispatcher working? [yes] :
  > ENTER
  Speech Dispatcher is installed and working!
  Speech Dispatcher works. Do you want to skip other tests? [yes] :
  > ENTER


  Diagnostics results:
  Speech Dispatcher is working
  End of diagnostics results


2.4.2 Editing speechd.conf
--------------------------

  The Speech Dispatcher configuration file can be edited manually:
  - either comment each 'AddModule' line, then speech-dispatcher should
    load each present module,
  - or add this line to select explicitly the voxin module:
  AddModule "voxin" "sd_voxin" "voxin.conf"

  Optionally, to set voxin as default module, modify the unique
  DefaultModule line: DefaultModule voxin


  Finally restart speech-dispatcher (e.g. logout or reboot).

  - global conf: /etc/speech-dispatcher/speechd.conf
  - local conf (created by spd-conf):
    $HOME/.config/speech-dispatcher/speechd.conf


2.5 Orca

  For Orca/Speech Dispatcher users: the newly installed TTS will be
  taken in account at the next start of Orca/Speech-dispatcher (if you
  authorized voxin-installer to update your configuration).  So, you may
  want to logout or reboot your system.  Then, if needed, configure Orca
  as indicated below.

  In the Orca Screen Reader Preferences window, Voice tab:
  - set "Speech system" to "Speech Dispatcher"
  - set "Speech synthesizer" to voxin

  then press the Apply button.

  The Orca preferences window is displayed using this command line:

  orca -s

  or typing the Orca modifier key + SPACE key


2.6 Emacspeak

  The [emacspeak installer] helps to install Emacspeak with eSpeak and
  Voxin.

  If the sound is choppy under Alsa, please copy the asoundrc file from
  emacspeak to your home directory as ~/.asoundrc For example: cp
  emacspeak-45.0/share/emacs/site-lisp/emacspeak/servers/linux-outloud/asoundrc
  $HOME/.asoundrc


[emacspeak installer]
https://github.com/Oralux/emacspeak_voxin_install/tags


2.7 Fenrir

  Select the speech-dispatcher driver and the voxin module in
  settings.conf.  For example in
  /etc/fenrirscreenreader/settings/settings.conf

  driver=speechdDriver module=voxin


2.8 Speakup connector

  You may want to install speechd-up (interface between Speakup and
  Speech Dispatcher).


2.9 How to uninstall Voxin?

2.9.1 System-wide as superuser
------------------------------

  Voxin 3.4 can be uninstalled with this command line:

  sudo --login $PWD/voxin-installer.sh -u

  Or without sudo:
  su --login -c "$PWD/voxin-installer.sh -u"

  Please note that the text-to-speech files will be removed too (under
  /opt/IBM/ibmtts or /opt/oralux/nve).


2.9.2 Locally as regular user
-----------------------------

  ./voxin-installer.sh -u


3 Compatible GNU/Linux Distributions

3.1 Voxin 3.4

  The last checked distributions are listed below.  Much more should be
  compatible (particularly with a local install).  Note that the Cerence
  TTS Embedded are compatible with x86_64 and ARM architectures whereas
  IBM TTS is only compatible with x86_64.

  For support or feedback, please send a mail to contact@oralux.org with
  possibly the log file (log/voxin.log).

  --------------------------------------------
   Distributions              Architectures   
  --------------------------------------------
   Arch Linux (October 2023)  x86_64          
   Debian 13 (testing)        x86_64          
   Fedora 39 (beta)           x86_64          
   Raspberry Pi OS            armv7l, aarch64 
   Slint64-15.0               x86_64          
   Ubuntu 23.10               x86_64          
  --------------------------------------------


4 voxin.ini configuration file

  The user may now define the default value of several parameters.
  - capitalization
  - punctuation
  - voiceName

  and only for IBM TTS:
  - dictionaryDir
  - useAbbreviation

  These values will be overwritten by the calling software
  (e.g. speech-dispatcher, emacspeak,...).  The voxin.ini file is
  consequently interesting when a parameter is not yet managed by the
  calling software.

  voxin.ini is expected to be present locally at:
  $HOME/.config/voxin/voxin.ini or if not found, available under the
  installation directory at: var/opt/oralux/voxin/voxin.ini

  The installer creates a writable default file in
  var/opt/oralux/voxin/voxin.ini if none exists.

  Brief Example:

  [general] capitalization=icon #punctuation=none

  [viavoice] dictionaryDir=/var/opt/IBM/ibmtts/dict useAbbreviation=no


5 Dictionaries

  User dictionaries are currently only available for IBM TTS.  They are
  enabled by default by speech-dispatcher: in voxin.conf (or
  ibmtts.conf), the IbmttsDictionaryFolder variable is set to
  /var/opt/IBM/ibmtts/dict in case of global installation.

  Examples of dictionaries are included in this archive under:
  dictionary/var/opt/IBM/ibmtts/dict


6 voxin-say command

  voxin-say is a command which converts text to speech.

  EXAMPLES :

  # Say 'hello world' and redirect output to an external audio player:
  voxin-say "hello world" | aplay

  # Read file.txt and save speech to an audio file:
  voxin-say -f file.txt -w file.wav
  voxin-say -f file.txt > file.wav

  # The following command is incorrect because no output is supplied:
  voxin-say "Hello all"

  # Correct command to read a file in French at 500 words per minute, use 4 jobs to speed up conversion
  voxin-say -f file.txt -l fr -s 500 -j 4 -w audio.wav
