Festival is a complely free, open source, easy-to-use, portable and cross-platform command-line software project implemented in C++ and designed from the offset to provide a framework for creating multilingual speech synthesis systems that contain examples of various modules.
Features at a glance
Key features include support for various languages, including American English, British English, as well as Spanish, a complete text-to-speech system, as well as a complete set of tools and documentation for building new voices.
Among other features, we can mention backwards compatibility with Festival 1.4.3, support for externally configurable language independent modules, waveform synthesizers, SABLE markup, server/client, scripting, and Emacs interfaces.
What’s in the package?
In the package you will find a complete English (British and American) text-to-speech system, a lexicon based in OALD (Oxford Advanced Learner's Dictionary) and CMULEX, the complete C++ source of the SIOD interpreter, the Scheme library and the modules, the Edinburgh speech tools, as well as the the complete documentation in various file formats.
In addition, users will also find the rab_diphon (British English Male residual LPC, diphone), cmu_us_rms_cg (American Male using clustergen), kal_diphone (American English Male residual LPC diphone), cmu_us_awb_cg (Scottish English Male (with US frontend) clustergen) and cmu_us_slt_arctic_hts (American Female, HTS) modules.
Under the hood and supported OSes
Taking a look under the hood of the Festival project, we can notice that it has been written entirely in the C++ programming language and that it is compatible with any UNIX-like operating system, including all GNU/Linux distributions, as well as FreeBSD, Mac OS X and Microsoft Windows.
Currently, the program has been successfully tested with computers supporting either of the 64-bit (x86_64) or 32-bit (x86) CPU architectures. For installation and usage instructions please read the included documentation or visit the project’s official website (click the Homepage link below).
What is new in this release:
- HTS hidden Markov model based synthesis engine API integration
- Support for gcc 4.3, 4.4 and 4.5
- Native Apple OS X audio support (OS X 10.6)
- Bug fixes galore.
What is new in version 2.1:
- HTS hidden Markov model based synthesis engine API integration
- Support for gcc 4.3, 4.4 and 4.5
- Native Apple OS X audio support (OS X 10.6)
- Bug fixes galore.
Comments not found