ru_tts

ru_tts speech synthesizer

An alternative implementation of the Phonemophone-5 Russian speech synthesizer by the international laboratory of intelligent systems BelSInt (the Speech Recognition and Synthesis Lab of the Institute of Technical Cybernetics of the Academy of Sciences of the Byelorussian SSR). The source code of the original implementation has been lost. This implementation is the result of a reverse engineering of the SDRV resident speech driver for MS-DOS, and it is officially approved for publication under a free license by Boris Lobanov, who is the head of the laboratory and the author of the design solutions that formed the basis of the speech synthesizer, and Alexander Ivanov, who is an engineer of the laboratory and the developer of the speech synthesizer's the original software implementation.

Build instructions

To build the application you will need GNU Autoconf, Automake and Libtool in addition to C compiler and Make utility. On a Debian-like system all necessary tools can be installed by following command:

$ sudo apt install autoconf-archive automake libtool make

The building process itself consists of thre usual stages:

$ autoreconf -ifs
$ ./configure
$ make

The configure script produced on the first stage accepts several options allowing us to control building process in some aspects. To get a grasp run it as follows:

$ ./configure --help

Cooperation with the RuLex pronunciation dictionary

This synthesizer can take advantage of the RuLex pronunciation dictionary. By default the rulex library, that provides dictionary access means, is not linked into the result executable as its mandatory part, but is dynamically loaded by need at runtime if present.

Use configure option --with-dictionary if you want to link with the rulex library on the application build stage. On the contrary, if no RuLex cooperation is needed use configure option --without-dictionary.

Installation

After building the application can be installed as follows:

# make install

Besides that, there are ready to use packages for some releases of Ubuntu, that can be directly installed as follows:

$ sudo add-apt-repository ppa:poretsky/a11y
$ sudo apt update
$ sudo apt install ru-tts

Usage

See README file for usage details.

Acknowledgements

Thanks to the following organizations and people for their contributions to the project:

• the international laboratory of intelligent systems BelSInt for the development of the original SDRV resident speech driver;
• Dmitry Paduchikh for help in the initial disassembly of the SDRV resident speech driver;
• Boris Lobanov for the creation of the Phonemophone-5 speech synthesizer and approval of the publication of the results of a reverse engineering of the SDRV resident speech driver under a free license.
• Alexander Ivanov for the development of the original software implementation of the Phonemophone-5 speech synthesizer and approval of the publication of the results of a reverse engineering of the SDRV resident speech driver under a free license.

Scientific publications

In any papers or publications that use or mentioned the design solutions of the Phonemophone-5 speech synthesizer, please cite the following article:

• Ivanov A.N., Lobanov B.M. - PHONEMOPHONE speech synthesizer for CAD based on IBM PC // Abstracts report conf. "Theory and methods of creating intelligent CAD systems". - Minsk, 1992. - pp. 29-30 (in Russian).

The above article is a basic publication about the original implementation of the Phonemophone-5 speech synthesizer for MS-DOS.

The microwave speech synthesis method developed by Boris Lobanov and used in the Phonemophone-5 is also described in the following scientific publications:

• Lobanov B.M. - Microwave Speech Synthesis from Text // Analysis and speech synthesis. - Minsk: the Institute of Technical Cybernetics of the ASB, 1991. - pp. 57-73 (in Russian).
• Lobanov B.M. - Programming Model of Microwave Speech Synthesis from Text // Automatic Acoustic Pattern Recognition (AAPR-14). - Moscow, 1991. - pp. 82-84 (in Russian).
• Lobanov B.M. - Microwave Speech Synthesis // Automatic Acoustic Pattern Recognition (AAPR-16). - Moscow, 1991. - pp. 27-31 (in Russian).
• Lobanov B. - Microwave Speech Synthesis from Text // Proc. of the 24 Fachkolloquim Informationstechnik. - Dresden, 1991. - pp. 118-120.

(For the original Russian titles of the scientific publications, see the Russian version of ReadMe.)

Related projects

There are several projects related to ru_tts speech synthesizer:

• RuLex - Russian pronunciation dictionary for ru_tts
• ru_tts for NVDA - add-on for NVDA screen reader with speech driver for ru_tts.
• RU_TTS_EX - extension to facilitate work with high-level languages and implement additional synthesizer capabilities (deprecated after ru_tts package release 6.0.4 and not supported)