Windows 10 IoT Core Speech and Speech-Synthesis Test

Just tested the speech capabilities of Windows 10 IoT Core for my entry:

It works very well. My grammar file still needs some improvements.


Very nice!

What are you using for a microphone? USB mic? Or something plugged into an I/O?

1 Like

@ devhammer - Usb headset. I used a simple mic as well. Works good, but not as good as a usb mic.

EDIT: Works very well with this cheap microphone and a cheap USB sound-card:



I am going to open up the mic and put that without the plastics into my device.

:clap: Very cool demo. Does it support other voices as well?

Until now I have not seen anything compelling enough to sway my preference of conventional Rasbian towards Win 10 IoT Core, but I think this demo certainly goes a long way towards doing that. Thanks for sharing this.

1 Like

@ KiwiSaner -Thanks :slight_smile: Yes, there are two voices (by default) available:

Microsoft Mark Mobile
Microsoft Zira Mobile

using Speech Synthesis Markup Language (SSML), switching voices and gender should be possible.

EDIT: Just tested, assigning a language works without SSML as well

SpeechSynthesizer synt = new SpeechSynthesizer();


                                foreach(var vcs in SpeechSynthesizer.AllVoices)

                                var voice = SpeechSynthesizer.AllVoices.FirstOrDefault(v => v.DisplayName.Contains("Zira"));

                                if (voice != null)
                                    synt.Voice = voice;

                                SpeechSynthesisStream syntStream =
                                    await synt.SynthesizeTextToStreamAsync("Calling your mum. Just a moment.");
                                mediaElement.SetSource(syntStream, syntStream.ContentType);

@ AWSOMEDEVSIGNER - Very nice demo. :clap:

1 Like

@ .Peter. - Thank you Peter :slight_smile: