Microsoft Server Speech Platform 10.1 Released (SR and TTS in 26 languages)


Want to build your speech application with speech recognitions and synthesis?


Install the package below, you can now use SR and TTS in 26 language easily. 



The 26 languages are:













































































































Catalan


Spain


ca-ES


Chinese (Hong Kong)


China


zh-HK


Chinese (Simplified)


China


zh-CHS


Chinese (Traditional)


Taiwan


zh-TW


Danish


Denmark


da-DK


Dutch


Netherlands


nl-NL


English


Australia


en-AU


English


Canada


en-CA


English


India


en-IN


English


United Kingdom


en-GB


English


United States


en-US


Finnish


Finland


fi-Fl


French


Canada


fr-CA


French


France


fr-FR


German


Germany


de-DE


Italian


Italy


it-IT


Japanese


Japan


ja-JP


Korean


Korean


ko-KR


Norwegian (Bokmal)


Norway


nb-NO


Polish


Poland


pl-PL


Portuguese


Brazil


pt-BR


Portuguese


Portugal


pt-PT


Russian


Russia


ru-RU


Spanish


Spain


es-ES


Spanish


Mexico


es-MX


Swedish


Sweden


sv-SE


 


Feel free to try the TTS in the 26 languages, post your feedbacks about them here.


 


 

Comments (30)

  1. Denis Hamel says:

    French Canadian Harmonie, need some more tuning.  Does not sound that great compare to acapela or RealSpeak

  2. myblog1 says:

    Hi, Denis :

    Thanks for the feedback. Do you have some specific comments from intelligiblity, prosody naturalness or sound quality perspective?

  3. dede says:

    How do I create SR and TTS for my language? Beside our national language, we have many region languages. It would be nice if we can preserve the languages. thanks for guide.

  4. myblog1 says:

    Hi, DeDE:

    For your region language, do they have different lexicon or pronunication system?

    Seems that you want to create dialect voice, there is not public tools to do so yet. But I think this is a direction of TTS in the future.

  5. Server says:

    Wait this is for Server OS only? I installed it on Windows 7 but no voices show up in the Text to Speech control panel.

  6. Server says:

    Why not make these server only voices available on the desktop?

  7. myblog1 says:

    Hi, Server:

    These voices are optimized for server scenerios.

    For example, the output format of the voice is 8k16bit.

    But you can use the voices on desktop by writing an app with sdk but not with the text-to-speech control panel.

  8. dede says:

    yes, it is difference lexicon and pronunciation system per region languages.

    Since it own region has they languages and culture, just like mini world.

    Hope we can get the sdk. thanks for explaining. keep a hope. 🙂

  9. michaelplevy says:

    Can you explain what this package is? How does it relate to UCMA? Is it a new package, and if so why is it version 10.1?

  10. Michael says:

    I think this is great news for us. However, I must admit, I'm confused. Can you please do a post that explains the difference between this new "Speech Platform", the UCMA 2.0 SDK, and the Microsoft.speech and System.speech namespaces?

    I think this new release takes the Speech features from UCMA and the Recognizer and TTS engines from OCS and makes them available for reuse. So, with this I can build server speech apps without OCS. Is this correct?

  11. myblog1 says:

    Hi, Michael:

    UCMA SDK is with OCS. Now speech platform is a standalone SDK

  12. Iggy says:

    Does this require us to buy a separate license? I have a TTS synthesis application running on a server and would like to replace the Desktop Speech to Speech Server.

  13. myblog1 says:

    hi, Iggy:

    From technology point of view, you can certainly do this.

    For license issue, I am not a lawyer, but you can check the license agreement to see if it matches your scenerio and need.

  14. Arturito says:

    Can't get it going.

    1. Running Windows 7 Pro x64 with Visual Studio 2008.

    2. Installed Runtime x86, SDK x86 and SR en-GB.

    3. In VS set platform target to x86 for All Configuration

    Copied and pasted the code from documentation file:

    using Microsoft.Speech.Recognition;

       private void Form1_Load(object sender, EventArgs e)

       {

         // Create a new SpeechRecognitionEngine instance.

         sre = new SpeechRecognitionEngine();

         // Create a simple grammar that recognizes “red”, “green”, or “blue”.

         Choices colors = new Choices();

         colors.Add("red");

         colors.Add("green");

         colors.Add("blue");

         GrammarBuilder gb = new GrammarBuilder();

         gb.Append(colors);

         // Create the actual Grammar instance, and then load it into the speech recognizer.

         Grammar g = new Grammar(gb);

         sre.LoadGrammar(g);

         // Register a handler for the SpeechRecognized event.

         sre.SpeechRecognized += new EventHandler<SpeechRecognizedEventArgs>(sre_SpeechRecognized);

         sre.SetInputToDefaultAudioDevice();

         sre.RecognizeAsync(RecognizeMode.Multiple);

       }

       // Simple handler for the SpeechRecognized event.

       void sre_SpeechRecognized(object sender, SpeechRecognizedEventArgs e)

       {

         MessageBox.Show(e.Result.Text);

       }

       SpeechRecognitionEngine sre;

    Program compiles but doesn't recognize my voice.

  15. Alonso says:

    Hi, thanks for your post!

    Well, i have been trying to use this in C#, i'm importing the COM Microsoft Speech Object Library version 10.0 (guess that's the one that belongs to this SDK). I'm programming under Windows 7 OS, with VS 2010.

    Ok, so, i'm just trying to use the next code:

    SpVoice voz = new SpVoice();

    voz.Speak("Hi there");

    But, it throws an exception: HRESULT: 0x8004503A

    Don't know what i'm doing wrong, could you please help me?

    Thanks!

  16. myblog1 says:

    do you have sound card on your windows 7?

  17. Alonso says:

    Don't know if the question is for me, but, if that's the case, yeap, i have sound card

  18. John says:

    Do you know if there are male voices available in the runtime? If not, are there any plans by the MS Speech team to add male voices in the future?

  19. myblog1 says:

    Hi, John:

    Currenlty no male voice.

    We will consider the male voice in plan.

    Thanks

  20. Xu Hin says:

    What exactly do the language packs indicate? If I am using the Chinese language pack, does it mean that all english text is translated into chinese and converted to speech or does it pretty much speak English text in a Chinese "accent"?

  21. myblog1 says:

    Hi, Xu:

    Language pack means it reads the text in that language like a native.

    Chinese language pack means a Chinese TTS.

  22. Xu Hin says:

    @myblog – I am not sure if I understand correctly. If I use English language text and use a Chinese language pack, does it read the English text in the Chinese accent? What's the correct way of using the non English language packs?

  23. John says:

    >> We will consider the male voice in plan.

    Thanks. Do you have an estimate on when this support will be added to the runtime? (Don't need an exact date, just want to know if this might take months or years)

  24. myblog1 says:

    Hi, Xu:

    Our Chinese language pack can read English word inside it.  

    Usually you use a language pack if your incoming text is mainly in that language.

  25. myblog1 says:

    Hi, John:

    Sorry, I cannot give you the ETA of it for now. But your feedback is heard.  

  26. Everton says:

    Hi,

    I'm starting to research about SR platforms and can't get it running for Portuguese (Brazil). Where can I find some samples? Tried all available in samples/winui/speech Windows SDK folder but no success.

  27. myblog1 says:

    Hi, Everton:

    Have you looked at the SDK package?

    For the samples in Windows SDK, they are for desktop speech API.

    Here is the difference between desktop Speech API and server Speech API:

    1. They have different GUIDs for COM object.

    2. They have different name space. (System.Speech for Desktop,  Microsoft.Speech for Server)

  28. Eliseo Calderon says:

    Is there a way to process an AMR (Narrowband) audio as an input for the speech recognizer? It seems I can only use PCM. Since AMR is smaller I was hoping I could use it since the speech I am recognizing comes from a recording made in a phone.

  29. koti says:

    i installed speech run-time in my system and Japanese language pack also but when i tried to convert Japanese text in voice then the engine is not speaking synthesizing to voice so can u please tell how exactly i need to follow. and is there any case like i must have Japanese voice then only engine will synthesize the text?

  30. Adalberto says:

    Can I use the new Microsoft Speech Platform with MsAgents ? It works fine ?