public void SelectVoiceByHints(VoiceGender gender, VoiceAge age, int voiceAlternate, CultureInfo culture)
{
Helpers.ThrowIfNull(culture, nameof(culture));
if (voiceAlternate < 0)
{
throw new ArgumentOutOfRangeException(nameof(voiceAlternate), SR.Get(SRID.PromptBuilderInvalidVariant));
}
if (!VoiceInfo.ValidateGender(gender))
{
throw new ArgumentException(SR.Get(SRID.EnumInvalid, "VoiceGender"), nameof(gender));
}
if (!VoiceInfo.ValidateAge(age))
{
throw new ArgumentException(SR.Get(SRID.EnumInvalid, "VoiceAge"), nameof(age));
}
TTSVoice ttsVoice = VoiceSynthesizer.GetEngine(null, culture, gender, age, voiceAlternate, true);
if (ttsVoice == null)
{
// No match - throw
throw new InvalidOperationException(SR.Get(SRID.SynthesizerSetVoiceNoMatch));
}
VoiceSynthesizer.Voice = ttsVoice;
}
protected async Task <string> GetAmazonSynthSpeech(string text, TTSVoice voicePreference, TTSPitch pitchPreference, string filename = null)
{
AmazonSynthesizeSpeechRequest synthesisRequest = voicePreference.GetAmazonTTSSpeechRequest();
synthesisRequest.TextType = TextType.Ssml;
synthesisRequest.Text = PrepareAmazonSSML(text, pitchPreference);
// Perform the Text-to-Speech request, passing the text input
// with the selected voice parameters and audio file type
AmazonSynthesizeSpeechResponse synthesisResponse = await amazonClient.SynthesizeSpeechAsync(synthesisRequest);
// Write the binary AudioContent of the response to file.
string filepath;
if (string.IsNullOrWhiteSpace(filename))
{
filepath = Path.Combine(TTSFilesPath, $"{Guid.NewGuid()}.mp3");
}
else
{
filepath = Path.Combine(TTSFilesPath, $"{filename}.mp3");
}
using (Stream file = new FileStream(filepath, FileMode.Create))
{
await synthesisResponse.AudioStream.CopyToAsync(file);
await file.FlushAsync();
file.Close();
}
return(filepath);
}
/// <summary>
/// Configures the voice.
/// </summary>
private static void ConfigureVoice()
{
SpWaveFormatEx fFormat = new SpWaveFormatEx();
ttSVoice = new TTSVoice();
//ttSVoice.SPVoice.Voice = ttSVoice.SPVoice.GetVoices("gender=female", "").Item(0);
StartService("AudioSrv", 20 * 1000);
/* set the audio out to the telephony device*/
ttSVoice.MMSysAudioOut.DeviceId = (int)w2cDrv.Device.Info.WaveOutDrvID;
WriteLog("Sound DeviceId: " + w2cDrv.Device.Info.WaveOutDrvID.ToStr());
fFormat.FormatTag = 1;
fFormat.Channels = 1;
fFormat.SamplesPerSec = 8000;
fFormat.AvgBytesPerSec = 16000;
fFormat.BlockAlign = 2;
fFormat.BitsPerSample = 16;
ttSVoice.MMSysAudioOut.Format.SetWaveFormatEx(fFormat);
/* Prevent SAPI from changing the wave format when the device changes*/
ttSVoice.SPVoice.AllowAudioOutputFormatChangesOnNextSet = false;
ttSVoice.SPVoice.AudioOutputStream = ttSVoice.MMSysAudioOut;
}
public static TTSVoice TranslateTTSVoice(this string voiceString)
{
if (ttsVoiceLookup == null)
{
ttsVoiceLookup = new Dictionary <string, TTSVoice>();
for (TTSVoice voice = 0; voice < TTSVoice.MAX; voice++)
{
ttsVoiceLookup.Add(voice.Serialize().ToLowerInvariant(), voice);
}
}
string cleanedString = voiceString.Trim().ToLowerInvariant();
if (ttsVoiceLookup.ContainsKey(cleanedString))
{
return(ttsVoiceLookup[cleanedString]);
}
if (cleanedString == "default")
{
return(TTSVoice.Unassigned);
}
return(TTSVoice.MAX);
}
public void SelectVoice(string name)
{
Helpers.ThrowIfEmptyOrNull(name, "name");
TTSVoice engine = VoiceSynthesizer.GetEngine(name, CultureInfo.CurrentUICulture, VoiceGender.NotSet, VoiceAge.NotSet, 1, true);
if (engine == null || name != engine.VoiceInfo.Name)
{
throw new ArgumentException(SR.Get(SRID.SynthesizerSetVoiceNoMatch));
}
VoiceSynthesizer.Voice = engine;
}
public static TTSService GetTTSService(this TTSVoice voice)
{
switch (voice)
{
case TTSVoice.en_AU_Standard_A:
case TTSVoice.en_AU_Standard_B:
case TTSVoice.en_AU_Standard_C:
case TTSVoice.en_AU_Standard_D:
case TTSVoice.en_IN_Standard_A:
case TTSVoice.en_IN_Standard_B:
case TTSVoice.en_IN_Standard_C:
case TTSVoice.en_IN_Standard_D:
case TTSVoice.en_GB_Standard_A:
case TTSVoice.en_GB_Standard_B:
case TTSVoice.en_GB_Standard_C:
case TTSVoice.en_GB_Standard_D:
case TTSVoice.en_GB_Standard_F:
case TTSVoice.en_US_Standard_B:
case TTSVoice.en_US_Standard_C:
case TTSVoice.en_US_Standard_D:
case TTSVoice.en_US_Standard_E:
case TTSVoice.en_US_Standard_G:
case TTSVoice.en_US_Standard_H:
case TTSVoice.en_US_Standard_I:
case TTSVoice.en_US_Standard_J:
return(TTSService.Google);
case TTSVoice.en_AU_Nicole:
case TTSVoice.en_AU_Olivia:
case TTSVoice.en_AU_Russell:
case TTSVoice.en_GB_Amy:
case TTSVoice.en_GB_Emma:
case TTSVoice.en_GB_Brian:
case TTSVoice.en_IN_Aditi:
case TTSVoice.en_IN_Raveena:
case TTSVoice.en_US_Ivy:
case TTSVoice.en_US_Joanna:
case TTSVoice.en_US_Kendra:
case TTSVoice.en_US_Kimberly:
case TTSVoice.en_US_Salli:
case TTSVoice.en_US_Joey:
case TTSVoice.en_US_Justin:
case TTSVoice.en_US_Matthew:
case TTSVoice.en_GB_WLS_Geraint:
return(TTSService.Amazon);
case TTSVoice.Unassigned: goto case TTSVoice.en_US_Joanna;
default:
BGC.Debug.LogError($"Unsupported TTSVoice {voice}");
goto case TTSVoice.en_US_Standard_B;
}
}
protected async Task <string> GetSynthSpeech(string text, TTSVoice voicePreference, TTSPitch pitchPreference)
{
switch (voicePreference.GetTTSService())
{
case TTSService.Amazon:
return(await GetAmazonSynthSpeech(text, voicePreference, pitchPreference));
case TTSService.Google:
return(await GetGoogleSynthSpeech(text, voicePreference, pitchPreference));
default:
communication.SendErrorMessage($"Unsupported TTSVoice for TTSService {voicePreference}");
goto case TTSService.Google;
}
}
protected async Task <string> GetGoogleSynthSpeech(string text, TTSVoice voicePreference, TTSPitch pitchPreference, string filename = null)
{
VoiceSelectionParams voice = voicePreference.GetGoogleVoiceSelectionParams();
AudioConfig config = new AudioConfig
{
AudioEncoding = AudioEncoding.Mp3,
Pitch = pitchPreference.GetSemitoneShift()
};
//TTS
SynthesisInput input = new SynthesisInput
{
Ssml = PrepareGoogleSSML(text)
};
// Perform the Text-to-Speech request, passing the text input
// with the selected voice parameters and audio file type
GoogleSynthesizeSpeechResponse response = await googleClient.SynthesizeSpeechAsync(input, voice, config);
// Write the binary AudioContent of the response to file.
string filepath;
if (string.IsNullOrWhiteSpace(filename))
{
filepath = Path.Combine(TTSFilesPath, $"{Guid.NewGuid()}.mp3");
}
else
{
filepath = Path.Combine(TTSFilesPath, $"{filename}.mp3");
}
using (Stream file = new FileStream(filepath, FileMode.Create))
{
response.AudioContent.WriteTo(file);
}
return(filepath);
}
public async Task <AudioRequest> TTSRequest(TTSVoice voicePreference, TTSPitch pitchPreference, Effect effectsChain, string[] splitTTSText)
{
if (splitTTSText.Any(x => x.Contains('/') || x.Contains('!')))
{
List <AudioRequest> audioRequestSegments = new List <AudioRequest>();
//Complex parsing
StringBuilder stringbuilder = new StringBuilder();
foreach (string ttsWord in splitTTSText)
{
if (ttsWord.Contains('/') || ttsWord.Contains('!'))
{
foreach (string ttsWordSegment in SplitStringByCommandRegex(ttsWord))
{
if (ttsWordSegment.StartsWith('/'))
{
//Sound Effect
SoundEffect soundEffect = soundEffectSystem.GetSoundEffectByAlias(ttsWordSegment);
if (soundEffect is null)
{
//Unrecognized, append as is
stringbuilder.Append(ttsWordSegment);
}
else
{
//Output current
if (stringbuilder.Length > 0)
{
string filename = await GetSynthSpeech(stringbuilder.ToString(), voicePreference, pitchPreference);
audioRequestSegments.Add(new AudioFileRequest(filename, effectsChain));
stringbuilder.Clear();
}
audioRequestSegments.Add(new SoundEffectRequest(soundEffect));
}
}
else if (ttsWordSegment.StartsWith('!'))
{
//Command
AudioRequest request = AudioRequest.ParseCommand(ttsWordSegment.ToLower());
if (request is null)
{
//Unrecognized, append as is
stringbuilder.Append(ttsWordSegment);
}
else
{
//Output current
if (stringbuilder.Length > 0)
{
string filename = await GetSynthSpeech(stringbuilder.ToString(), voicePreference, pitchPreference);
audioRequestSegments.Add(new AudioFileRequest(filename, effectsChain));
stringbuilder.Clear();
}
audioRequestSegments.Add(request);
}
}
else
{
stringbuilder.Append(ttsWordSegment);
}
}
if (stringbuilder.Length > 0)
{
stringbuilder.Append(' ');
}
}
else
{
stringbuilder.Append(ttsWord);
stringbuilder.Append(' ');
}
}
if (stringbuilder.Length > 0)
{
string filename = await GetSynthSpeech(stringbuilder.ToString(), voicePreference, pitchPreference);
audioRequestSegments.Add(new AudioFileRequest(filename, effectsChain));
stringbuilder.Clear();
}
return(new ConcatenatedAudioRequest(audioRequestSegments));
}
else
{
//Simple parsing
string ttsSpeech = string.Join(' ', splitTTSText);
string filename = await GetSynthSpeech(ttsSpeech, voicePreference, pitchPreference);
return(new AudioFileRequest(filename, effectsChain));
}
}
public Task <AudioRequest> TTSRequest(TTSVoice voicePreference, TTSPitch pitchPreference, Effect effectsChain, string ttsText) =>
TTSRequest(voicePreference, pitchPreference, effectsChain, ttsText.Split(' ', options: StringSplitOptions.RemoveEmptyEntries));
public static string Serialize(this TTSVoice voice)
{
switch (voice)
{
case TTSVoice.en_AU_Standard_A: return("en-AU-Standard-A");
case TTSVoice.en_AU_Standard_B: return("en-AU-Standard-B");
case TTSVoice.en_AU_Standard_C: return("en-AU-Standard-C");
case TTSVoice.en_AU_Standard_D: return("en-AU-Standard-D");
case TTSVoice.en_IN_Standard_A: return("en-IN-Standard-A");
case TTSVoice.en_IN_Standard_B: return("en-IN-Standard-B");
case TTSVoice.en_IN_Standard_C: return("en-IN-Standard-C");
case TTSVoice.en_IN_Standard_D: return("en-IN-Standard-D");
case TTSVoice.en_GB_Standard_A: return("en-GB-Standard-A");
case TTSVoice.en_GB_Standard_B: return("en-GB-Standard-B");
case TTSVoice.en_GB_Standard_C: return("en-GB-Standard-C");
case TTSVoice.en_GB_Standard_D: return("en-GB-Standard-D");
case TTSVoice.en_GB_Standard_F: return("en-GB-Standard-F");
case TTSVoice.en_US_Standard_B: return("en-US-Standard-B");
case TTSVoice.en_US_Standard_C: return("en-US-Standard-C");
case TTSVoice.en_US_Standard_D: return("en-US-Standard-D");
case TTSVoice.en_US_Standard_E: return("en-US-Standard-E");
case TTSVoice.en_US_Standard_G: return("en-US-Standard-G");
case TTSVoice.en_US_Standard_H: return("en-US-Standard-H");
case TTSVoice.en_US_Standard_I: return("en-US-Standard-I");
case TTSVoice.en_US_Standard_J: return("en-US-Standard-J");
case TTSVoice.en_AU_Nicole: return("Nicole");
case TTSVoice.en_AU_Olivia: return("Olivia");
case TTSVoice.en_AU_Russell: return("Russell");
case TTSVoice.en_GB_Amy: return("Amy");
case TTSVoice.en_GB_Emma: return("Emma");
case TTSVoice.en_GB_Brian: return("Brian");
case TTSVoice.en_IN_Aditi: return("Aditi");
case TTSVoice.en_IN_Raveena: return("Raveena");
case TTSVoice.en_US_Ivy: return("Ivy");
case TTSVoice.en_US_Joanna: return("Joanna");
case TTSVoice.en_US_Kendra: return("Kendra");
case TTSVoice.en_US_Kimberly: return("Kimberly");
case TTSVoice.en_US_Salli: return("Salli");
case TTSVoice.en_US_Joey: return("Joey");
case TTSVoice.en_US_Justin: return("Justin");
case TTSVoice.en_US_Matthew: return("Matthew");
case TTSVoice.en_GB_WLS_Geraint: return("Geraint");
case TTSVoice.Unassigned: return("Unassigned");
default:
BGC.Debug.LogError($"Unsupported TTSVoice {voice}");
goto case TTSVoice.en_US_Standard_B;
}
}
public static Amazon.Polly.Model.SynthesizeSpeechRequest GetAmazonTTSSpeechRequest(this TTSVoice voice)
{
Amazon.Polly.Model.SynthesizeSpeechRequest synthesisRequest = new Amazon.Polly.Model.SynthesizeSpeechRequest
{
OutputFormat = OutputFormat.Mp3,
Engine = Engine.Standard,
LexiconNames = awsLexicons
};
switch (voice)
{
case TTSVoice.en_AU_Standard_A:
case TTSVoice.en_AU_Standard_B:
case TTSVoice.en_AU_Standard_C:
case TTSVoice.en_AU_Standard_D:
case TTSVoice.en_IN_Standard_A:
case TTSVoice.en_IN_Standard_B:
case TTSVoice.en_IN_Standard_C:
case TTSVoice.en_IN_Standard_D:
case TTSVoice.en_GB_Standard_A:
case TTSVoice.en_GB_Standard_B:
case TTSVoice.en_GB_Standard_C:
case TTSVoice.en_GB_Standard_D:
case TTSVoice.en_GB_Standard_F:
case TTSVoice.en_US_Standard_B:
case TTSVoice.en_US_Standard_C:
case TTSVoice.en_US_Standard_D:
case TTSVoice.en_US_Standard_E:
case TTSVoice.en_US_Standard_G:
case TTSVoice.en_US_Standard_H:
case TTSVoice.en_US_Standard_I:
case TTSVoice.en_US_Standard_J:
BGC.Debug.LogError($"Tried to get Amazon VoiceId from Google TTS Voice {voice}");
goto case TTSVoice.en_GB_Brian;
case TTSVoice.en_AU_Nicole:
synthesisRequest.VoiceId = VoiceId.Nicole;
break;
//Olivia is unsupported
case TTSVoice.en_AU_Olivia:
synthesisRequest.VoiceId = VoiceId.Nicole;
break;
case TTSVoice.en_AU_Russell:
synthesisRequest.VoiceId = VoiceId.Russell;
break;
case TTSVoice.en_GB_Amy:
synthesisRequest.VoiceId = VoiceId.Amy;
break;
case TTSVoice.en_GB_Emma:
synthesisRequest.VoiceId = VoiceId.Emma;
break;
case TTSVoice.en_GB_Brian:
synthesisRequest.VoiceId = VoiceId.Brian;
break;
case TTSVoice.en_IN_Aditi:
synthesisRequest.VoiceId = VoiceId.Aditi;
synthesisRequest.LanguageCode = LanguageCode.EnIN;
break;
case TTSVoice.en_IN_Raveena:
synthesisRequest.VoiceId = VoiceId.Raveena;
synthesisRequest.LanguageCode = LanguageCode.EnIN;
break;
case TTSVoice.en_US_Ivy:
synthesisRequest.VoiceId = VoiceId.Ivy;
break;
case TTSVoice.en_US_Joanna:
synthesisRequest.VoiceId = VoiceId.Joanna;
break;
case TTSVoice.en_US_Kendra:
synthesisRequest.VoiceId = VoiceId.Kendra;
break;
case TTSVoice.en_US_Kimberly:
synthesisRequest.VoiceId = VoiceId.Kimberly;
break;
case TTSVoice.en_US_Salli:
synthesisRequest.VoiceId = VoiceId.Salli;
break;
case TTSVoice.en_US_Joey:
synthesisRequest.VoiceId = VoiceId.Joey;
break;
case TTSVoice.en_US_Justin:
synthesisRequest.VoiceId = VoiceId.Justin;
break;
case TTSVoice.en_US_Matthew:
synthesisRequest.VoiceId = VoiceId.Matthew;
break;
case TTSVoice.en_GB_WLS_Geraint:
synthesisRequest.VoiceId = VoiceId.Geraint;
break;
case TTSVoice.Unassigned: goto case TTSVoice.en_US_Joanna;
default:
BGC.Debug.LogError($"TTS Voice not supported {voice}");
goto case TTSVoice.Unassigned;
}
return(synthesisRequest);
}
public static VoiceSelectionParams GetGoogleVoiceSelectionParams(this TTSVoice voice)
{
switch (voice)
{
case TTSVoice.Unassigned:
return(new VoiceSelectionParams
{
LanguageCode = "en-US",
SsmlGender = SsmlVoiceGender.Neutral
});
case TTSVoice.en_AU_Standard_A:
case TTSVoice.en_AU_Standard_B:
case TTSVoice.en_AU_Standard_C:
case TTSVoice.en_AU_Standard_D:
return(new VoiceSelectionParams
{
Name = voice.GetGoogleTTSVoiceString(),
LanguageCode = "en-AU",
SsmlGender = SsmlVoiceGender.Neutral
});
case TTSVoice.en_IN_Standard_A:
case TTSVoice.en_IN_Standard_B:
case TTSVoice.en_IN_Standard_C:
case TTSVoice.en_IN_Standard_D:
return(new VoiceSelectionParams
{
Name = voice.GetGoogleTTSVoiceString(),
LanguageCode = "en-IN",
SsmlGender = SsmlVoiceGender.Neutral
});
case TTSVoice.en_GB_Standard_A:
case TTSVoice.en_GB_Standard_B:
case TTSVoice.en_GB_Standard_C:
case TTSVoice.en_GB_Standard_D:
case TTSVoice.en_GB_Standard_F:
return(new VoiceSelectionParams
{
Name = voice.GetGoogleTTSVoiceString(),
LanguageCode = "en-GB",
SsmlGender = SsmlVoiceGender.Neutral
});
case TTSVoice.en_US_Standard_B:
case TTSVoice.en_US_Standard_C:
case TTSVoice.en_US_Standard_D:
case TTSVoice.en_US_Standard_E:
case TTSVoice.en_US_Standard_G:
case TTSVoice.en_US_Standard_H:
case TTSVoice.en_US_Standard_I:
case TTSVoice.en_US_Standard_J:
return(new VoiceSelectionParams
{
Name = voice.GetGoogleTTSVoiceString(),
LanguageCode = "en-US",
SsmlGender = SsmlVoiceGender.Neutral
});
case TTSVoice.en_AU_Nicole:
case TTSVoice.en_AU_Olivia:
case TTSVoice.en_AU_Russell:
case TTSVoice.en_GB_Amy:
case TTSVoice.en_GB_Emma:
case TTSVoice.en_GB_Brian:
case TTSVoice.en_IN_Aditi:
case TTSVoice.en_IN_Raveena:
case TTSVoice.en_US_Ivy:
case TTSVoice.en_US_Joanna:
case TTSVoice.en_US_Kendra:
case TTSVoice.en_US_Kimberly:
case TTSVoice.en_US_Salli:
case TTSVoice.en_US_Joey:
case TTSVoice.en_US_Justin:
case TTSVoice.en_US_Matthew:
case TTSVoice.en_GB_WLS_Geraint:
BGC.Debug.LogError($"Tried to get Google VoiceSelectionParams from AWS TTS Voice {voice}");
goto case TTSVoice.Unassigned;
default:
BGC.Debug.LogError($"TTS Voice not supported {voice}");
goto case TTSVoice.Unassigned;
}
}
private static string GetGoogleTTSVoiceString(this TTSVoice voice)
{
switch (voice)
{
case TTSVoice.en_AU_Standard_A: return("en-AU-Standard-A");
case TTSVoice.en_AU_Standard_B: return("en-AU-Standard-B");
case TTSVoice.en_AU_Standard_C: return("en-AU-Standard-C");
case TTSVoice.en_AU_Standard_D: return("en-AU-Standard-D");
case TTSVoice.en_IN_Standard_A: return("en-IN-Standard-A");
case TTSVoice.en_IN_Standard_B: return("en-IN-Standard-B");
case TTSVoice.en_IN_Standard_C: return("en-IN-Standard-C");
case TTSVoice.en_IN_Standard_D: return("en-IN-Standard-D");
case TTSVoice.en_GB_Standard_A: return("en-GB-Standard-A");
case TTSVoice.en_GB_Standard_B: return("en-GB-Standard-B");
case TTSVoice.en_GB_Standard_C: return("en-GB-Standard-C");
case TTSVoice.en_GB_Standard_D: return("en-GB-Standard-D");
case TTSVoice.en_GB_Standard_F: return("en-GB-Standard-F");
case TTSVoice.en_US_Standard_B: return("en-US-Standard-B");
case TTSVoice.en_US_Standard_C: return("en-US-Standard-C");
case TTSVoice.en_US_Standard_D: return("en-US-Standard-D");
case TTSVoice.en_US_Standard_E: return("en-US-Standard-E");
case TTSVoice.en_US_Standard_G: return("en-US-Standard-G");
case TTSVoice.en_US_Standard_H: return("en-US-Standard-H");
case TTSVoice.en_US_Standard_I: return("en-US-Standard-I");
case TTSVoice.en_US_Standard_J: return("en-US-Standard-J");
case TTSVoice.en_AU_Nicole:
case TTSVoice.en_AU_Olivia:
case TTSVoice.en_AU_Russell:
case TTSVoice.en_GB_Amy:
case TTSVoice.en_GB_Emma:
case TTSVoice.en_GB_Brian:
case TTSVoice.en_IN_Aditi:
case TTSVoice.en_IN_Raveena:
case TTSVoice.en_US_Ivy:
case TTSVoice.en_US_Joanna:
case TTSVoice.en_US_Kendra:
case TTSVoice.en_US_Kimberly:
case TTSVoice.en_US_Salli:
case TTSVoice.en_US_Joey:
case TTSVoice.en_US_Justin:
case TTSVoice.en_US_Matthew:
case TTSVoice.en_GB_WLS_Geraint:
BGC.Debug.LogError($"Tried to get Google Voice string from AWS TTS Voice {voice}");
goto case TTSVoice.en_US_Standard_B;
case TTSVoice.Unassigned: goto case TTSVoice.en_US_Standard_B;
default:
BGC.Debug.LogError($"TTS Voice not supported {voice}");
goto case TTSVoice.en_US_Standard_B;
}
}
