ATT Natural Voice SAPI 5 - English Voices
Using AT&T Natural Voices® Text-to-Speech
Aloud and includes AT&T Natural Voices. Text To Wav is a Text-To-Speech software using SAPI4.0, SAPI5.1 or Microsoft. Elearning Translation Services.
- Free Text to Speech Natural Voices - SAPI 4 & SAPI 5. Microsoft Simplified Chinese voice (Male) 1.71 MB Site1 Site2: Microsoft SAPI 5.1 TTS Engine (with 3 English.
- Thus, if a voice synthesizer. The following is a list of commercial SAPI 5 voices with samples. Cepstral offers high quality voices in English.
As of v2.6.1, QDS supports AT&T Natural Voices®* or any other SAPI 5**-compliant speech engine installed on your computer to provide text-to-speech functionality. Included in the ACASI installation are several voices:
oCrystal (female, U.S. English)
Microsoft Sapi 5 Download
oMel (male, U.S. English)
oMike (male, U.S. English)
oRosa (female, Spanish).
To see a listing of all voice fonts present on your system, go to Tools List Text-to-speech Voices in the ACASI module. This opens a list box that contains all SAPI 5-compliant voices installed on the system, displayed with the corresponding vendor name and language.
Additional languages and voices are available for purchase from Wizzard Natural Voices:
You can adjust pronunciation of words by building a custom dictionary for AT&T Natural Voices or modify other aspects of the AT&T Natural Voices speech engine by using the SSML (Speech Synthesis Markup Language).
The voice setting is language dependent, so you will select a unique voice for each language version of your questionnaire. You can select voice under ACASI Build Options, Audio Tab and also select to have ACASI read response categories (text buttons). By default, QDS will assume that TTS should be used if no Audio Files are specified.
* licensed from Wizzard Software
** SAPI 5 is a Microsoft acronym for Speech Application Programming Interface.
Re-posting info from #106 (thanks @delta003):
Assuming there is a internet connection, we can just use Google TTS API
(e.g. http://translate.google.com/translate_tts?tl=en&q=Integrate%20text%20to%20speech)
It should be fast enough and I think it covers all languages as Google Translate
The link is the example of HTTP request. You just make HTTP to translate.google.com/translate_tts setting tl=language (two letters shortcut for language) and query=what_do_you_want_to_say (make sure you escape this string). HTTP response have Content-Type:audio/mpeg so from response you can get audio file. Then you just pass that file to system to play it (I'm don't really know C# but this should be easy). Here are some useful links:
http://stackoverflow.com/questions/184683/play-audio-from-a-stream-using-c-sharp
http://stackoverflow.com/questions/7757881/c-sharp-download-text-to-speech-from-google-translate-comes-with-problems
http://stackoverflow.com/questions/9243560/how-can-i-use-google-text-to-speech-api-in-windows-form
I don't have access to Windows now, so I can't implement this, but if you want I can write you a class that implements IAudioService and use this approach. But you'll have to check it and probably debug it. Just ping me here if you need that. Also, I'll have access to windows in ~month so if this issue is still open I'll develop a feature.
I would like to split out the IAudioService as it's currently in breach of the
single responsibility principle. There should really be an interface and
class for playing sounds and an interface and multiple classes for speech.
Then I can just swap out one speech service for another, and even nest them
so we can fall back to Microsoft TTS if there is no active internet
connection, for example.