Welcome to the CereVoice Demo. Enter text here and press play.

  • English
  • French
  • German
  • Italian
  • Spanish
×
×

CereVoice Engine Text-to-Speech SDK

The CereVoice Engine SDK 3.1.0 is a cross-platform text-to-speech (TTS) software development kit, enabling developers to add CereProc's industry-leading TTS output to their application. A simple, yet powerful, C/C++ API, CereVoice Engine, is provided across Windows, Mac OS X, iPhone OS, Linux, and Android platforms. The CereVoice Engine API includes a Python wrapper to ease development of new applications. Industry-standard SSML markup can be used to manipulate pitch, speech rate, and pronunciations. SSML is enhanced with CereVoice XML extensions, allowing unrivalled tuning and TTS manipulation, emotional synthesis control, and the insertion of vocal gestures such as laughs and coughs.

TTS output can be added to an application with a handful of lines of code:

#include <cerevoice_eng_simp.h>
CPRCEN_engine * eng = CPRCEN_engine_load("license.lic", "speaker.voice");
CPRCEN_engine_speak_to_file(eng, "Hello there!", "out.wav");

For low-latency applications, the CereVoice Engine API supports output via a callback function, allowing audio to be received incrementally as soon as TTS output is available. The callback API provides word and phonetic boundary information for animation, and markers for application pass-through control.

For more information please contact us.

SDK package includes

  • CereVoice Engine TTS library, written in ANSI C for portability and speed
  • CereVoice Audio library for playback of generated TTS*
  • Python interface to CereVoice Engine*
  • CereProc Heather, Scottish-accented female voice, with vocal gesture library and patented Emotional Synthesis (additional voices from the CereProc Voice Library are available)
  • User Guide
  • Full API documentation
  • Example applications written in C++ and Python, demonstrating simple and callback API usage

Technical details

  • Supported Platforms:
    Microsoft Windows (XP, Vista, 7, 8, 8.1, 2003 Server, 2008 Server, 2012 Server, 32 and 64 bit versions)
    Mac OS X (Intel CPU)
    Linux (32 and 64 bit)
    Embedded Linux (including Arm, Blackfin, MIPS CPUs)
    iPhone OS
    Android
  • Supported Standards:
    SSML version 1.0/1.1, VoiceXML 2.0 (includes say-as entity support for IVR integration)
    CereVoice XML for speech tuning and emotional synthesis
    CereVoice User Lexicon format
    CereVoice Abbreviation format
    8/16/22khz linear PCM/wave audio output
  • Minimum Requirements:
    400Mhz CPU
    20Mb available RAM
    150Mb disk space
  • Supported Languages:
    English
    French
    Spanish
    Italian
    German
    Portuguese (European and Brazilian)
    Japanese
    Dutch
    Catalan

New in version 3.1

  • Voice quality, pronunciation, and text processing enhancements
  • Performance and latency improvements on mobile and embedded platforms
  • Japanese language support

New in version 3

  • Simpler, more powerful CereVoice Engine API
  • New integrated text processing engine
  • High performance, low latency synthesis with the callback API
  • Audio compression improvements for crisper, clearer voice quality
  • Improved letter-to-sound system allowing more accurate pronunciations
  • Lower memory footprint for mobile and embedded systems
  • iPhone OS support
  • New languages (French, Spanish, Italian) and voices (Scottish male, US male)

For more information please contact us.

* Not available on iPhone OS