Recently the release of the new version of the RHVoice 1.6.0 open speech synthesis system was announced, which was initially developed to provide high-quality support for the Russian language, but later adapted for other languages, including English, Portuguese, Ukrainian, Kyrgyz, Tatar, and Georgian.
For those who are unfamiliar with RHVoice, I can tell you that this project uses the developments of the HTS project (HMM / DNN-based Speech Synthesis System) and a parametric synthesis method with statistical models (Statistical Parametric Synthesis based on HMM - Hidden Markov Model).
The advantages of the statistical model are low overhead costs and low CPU power demand. All operations are performed locally on the user's system. Three levels of voice quality are supported (the lower the quality, the higher the performance, and the shorter the response time).
The disadvantage of the statistical model is the relatively low quality of the pronunciation, which does not reach the level of synthesizers that generate speech from a combination of natural speech fragments, but nevertheless the result is quite legible and resembles a broadcast from a speaker. By way of comparison, the Silero project, which provides an open engine for speech synthesis based on machine learning technologies and a set of models for the Russian language, is superior in quality to RHVoice.
There are 13 voices available for the Russian language and the voices are formed on the basis of natural voice recordings. In the settings, you can change the speed, pitch, and volume.
The Sonic Library can be used to change the tempo. It is possible to automatically detect and change the language based on the analysis of the input text (for example, for words and citations in another language, the native synthesis model of the given language can be used). Voice profiles are supported, which define voice combinations for different languages.
The code is written in C ++ and is distributed under the LGPL 2.1 license, in addition to the fact that the system is supported on GNU / Linux, Windows and Android. The program supports typical TTS (text-to-speech) interfaces for converting text to speech: SAPI5 (Windows), Speech Dispatcher (GNU / Linux), and Android Text-To-Speech API, but can also be used on screen NVDA.
RHVoice 1.6.0 Key New Features
In this new version of the system, it stands out as the main novelty that 5 new voices are added for Russian speech, In addition, support for the Albanian language has been implemented.
Another change that stands out in this new version is that the dictionary was updated for the Ukrainian language and that support has been expanded to express emoji characters.
It also highlights the work that se made in the correction of errors in the application of the Android platform, Importing custom dictionaries has been simplified and support for the Android 11 platform has been added.
On the other hand, we can also find that new configurations and functionalities have been added to the core of the engine, including g2p.case, word_break and compatibility with equalization filters.
Finally, if you are interested in knowing more about this new version, you can consult the details In the following link.
Download RHVoice
Forthose who are interested in being able to download this system for speech synthesis, you can get the installation packages from the following link
It is also mentioned in the announcement of this new version that for Android users that have RHVoice already installed on your device, this will update automatically, if automatic updates are enabled, so there is no need to do the process manually.
In the case of having the updates disabled and you want to have the new version, you can activate the function of checking for updates manually.
As soon as the updated RHVoice runs again, it will try to download the new language data. When the new data is downloaded, RHVoice will start using it.