public async Task <string> ObtainSpeechPathAsync(string text, string language = "en-US", bool male = true)
{
TextToSpeechClient client = TextToSpeechClient.Create();
var response = await client.SynthesizeSpeechAsync(new SynthesizeSpeechRequest
{
Input = new SynthesisInput
{
Text = text
},
Voice = new VoiceSelectionParams
{
LanguageCode = language,
SsmlGender = male ? SsmlVoiceGender.Male : SsmlVoiceGender.Female
},
AudioConfig = new AudioConfig
{
AudioEncoding = AudioEncoding.Mp3
}
});
var filename = Path.GetTempFileName();
using (Stream output = File.Create(filename))
{
response.AudioContent.WriteTo(output);
_logger.Log($"Audio content written to temp file '{filename}'");
}
return(filename);
}
public async Task GenerateSpeechFile(string text, string path)
{
TextToSpeechClient client = TextToSpeechClient.Create();
SynthesisInput input = new SynthesisInput
{
Text = text
};
VoiceSelectionParams voice = new VoiceSelectionParams
{
LanguageCode = "en-US",
SsmlGender = SsmlVoiceGender.Neutral
};
AudioConfig config = new AudioConfig
{
AudioEncoding = AudioEncoding.Mp3
};
var response = await client.SynthesizeSpeechAsync(new SynthesizeSpeechRequest
{
Input = input,
Voice = voice,
AudioConfig = config
});
using (Stream output = File.Create(path))
{
response.AudioContent.WriteTo(output);
}
}
private async Task GetTweetAudioAsync(string ssml, string voiceName, string languageCode, Stream destination)
{
string args = "-hide_banner -loglevel panic -i pipe:0 -ac 2 -f s16le -ar 48000 pipe:1";
var request = new SynthesizeSpeechRequest
{
Input = new SynthesisInput
{
Ssml = ssml
},
Voice = new VoiceSelectionParams
{
LanguageCode = languageCode,
Name = voiceName
},
AudioConfig = new AudioConfig
{
AudioEncoding = AudioEncoding.OggOpus,
}
};
var response = await _ttsClient.SynthesizeSpeechAsync(request);
using MemoryStream audioContent = new MemoryStream();
response.AudioContent.WriteTo(audioContent);
audioContent.Seek(0, SeekOrigin.Begin);
await(audioContent | Cli.Wrap("ffmpeg.exe").WithArguments(args) | destination).ExecuteAsync();
}
private static async Task SpeakGoogle(ISpeaker speaker, string textToSpeech, string user)
{
textToSpeech = textToSpeech.Replace("\"", "\"\"");
// Instantiate a client
TextToSpeechClient client = TextToSpeechClient.Create();
// Set the text input to be synthesized.
SynthesisInput input = new SynthesisInput
{
//Text = textToSpeech,
Ssml = File.ReadAllText("Speakers/SSML.xml").Replace("{text}", textToSpeech).Replace("{voice}", speaker.Voice).Replace("{posmsg}", speaker.Diction).Replace("{alert}", speaker.Alert),
};
// Build the voice request, select the language code ("en-US"),
// and the SSML voice gender ("neutral").
VoiceSelectionParams voice = new VoiceSelectionParams
{
LanguageCode = speaker.Accent.ToString(),
//SsmlGender = SsmlVoiceGender.Neutral
};
// Select the type of audio file you want returned.
AudioConfig config = new AudioConfig
{
AudioEncoding = AudioEncoding.Mp3
};
// Perform the Text-to-Speech request, passing the text input
// with the selected voice parameters and audio file type
var response = await client.SynthesizeSpeechAsync(new SynthesizeSpeechRequest
{
Input = input,
Voice = voice,
AudioConfig = config
});
// create a temp file with .ps1 extension
var cFile = System.IO.Path.GetTempPath() + Guid.NewGuid() + ".mp3";
// Write the binary AudioContent of the response to an MP3 file.
using (Stream output = File.Create(cFile))
response.AudioContent.WriteTo(output);
Sounds.RandomTrollSound();
SpeakerCore.PreSpeech(user);
SpeakerCore.ExecuteMP3File(cFile);
await AutomaticTranslator.Translate(textToSpeech);
}
public async Task <string> ToAudio(string transcript)
{
if (toSpeechClient == null)
{
return("");
}
var BUFFER_SIZE = 2048;
var timeStamp = DateTime.Now.ToString("MM-dd-yyyy_HH_mm_ss");
var outputFileName = $@".\vocalized\{timeStamp}.mp3";
try
{
using (var fileStream = File.Create(outputFileName))
{
var response = await toSpeechClient.SynthesizeSpeechAsync(new SynthesizeSpeechRequest
{
Input = new SynthesisInput
{
Text = transcript
},
// Note: voices can also be specified by name
Voice = new VoiceSelectionParams
{
Name = settingsService.settings.googleSettings.Voice.Name,
LanguageCode = settingsService.settings.generalSettings.TextInputLanguage
},
AudioConfig = new AudioConfig
{
AudioEncoding = AudioEncoding.Mp3
}
});
var ms = new MemoryStream(response.AudioContent.ToByteArray());
byte[] buffer = new byte[BUFFER_SIZE];
int readBytes;
using (BufferedStream bufferedInput = new BufferedStream(ms))
{
while ((readBytes = await bufferedInput.ReadAsync(buffer, 0, buffer.Length)) > 0)
{
await fileStream.WriteAsync(buffer, 0, readBytes);
}
}
}
return(outputFileName);
}
catch (Exception e)
{
Console.WriteLine("Exception caught: " + e);
MessageBox.Show("Exception caught: " + e);
}
return(outputFileName);
}
/// <summary>
/// リクエストを送信する
/// </summary>
/// <param name="text">音声合成を行う対象の文</param>
public static void CreateRequest(string text)
{
var request = new SynthesizeSpeechRequest
{
Input = new SynthesisInput {
Text = text
},
AudioConfig = _audioConfig,
Voice = _voiceSelectionParams
};
_stopwatch.Restart();
// リクエストを非同期で送信し，返ってきた後に再生するメソッドに投げる
Task.Run(async() => { SetAudioClip(await _client.SynthesizeSpeechAsync(request)); });
}
/// <summary>Snippet for SynthesizeSpeechAsync</summary>
public async Task SynthesizeSpeechAsync()
{
// Snippet: SynthesizeSpeechAsync(SynthesisInput, VoiceSelectionParams, AudioConfig, CallSettings)
// Additional: SynthesizeSpeechAsync(SynthesisInput, VoiceSelectionParams, AudioConfig, CancellationToken)
// Create client
TextToSpeechClient textToSpeechClient = await TextToSpeechClient.CreateAsync();
// Initialize request argument(s)
SynthesisInput input = new SynthesisInput();
VoiceSelectionParams voice = new VoiceSelectionParams();
AudioConfig audioConfig = new AudioConfig();
// Make the request
SynthesizeSpeechResponse response = await textToSpeechClient.SynthesizeSpeechAsync(input, voice, audioConfig);
// End snippet
}
public async Task <Stream> TextToAudioStreamAsync(string text, string language = "fr-FR", SsmlVoiceGender voiceGender = SsmlVoiceGender.Male, CancellationToken token = default)
{
var request = new SynthesizeSpeechRequest
{
AudioConfig = new AudioConfig
{
AudioEncoding = AudioEncoding.OggOpus,
},
Input = new SynthesisInput
{
Text = text
},
Voice = new VoiceSelectionParams
{
LanguageCode = language,
SsmlGender = voiceGender,
},
};
var response = await ttsClient.SynthesizeSpeechAsync(request, token);
using var opusStream = new MemoryStream();
response.AudioContent.WriteTo(opusStream);
opusStream.Position = 0;
var opusDecoder = new OpusDecoder(48000, 2);
var oggIn = new OpusOggReadStream(opusDecoder, opusStream);
var pcmStream = new MemoryStream();
while (oggIn.HasNextPacket)
{
short[] packet = oggIn.DecodeNextPacket();
if (packet != null)
{
for (int i = 0; i < packet.Length; i++)
{
byte[] bytes = BitConverter.GetBytes(packet[i]);
pcmStream.Write(bytes, 0, bytes.Length);
}
}
}
pcmStream.Position = 0;
return(pcmStream);
}
private async Task QueueText(string text, int index, GoogleSpeechRequest request, GoogleSpeechResult result)
{
var response = await _client.SynthesizeSpeechAsync(
input : new SynthesisInput
{
Text = text
},
voice : request.VoiceSelection,
audioConfig : request.AudioConfig
);
result.DigestBytes(index, response.AudioContent);
if (request.CallBack != null)
{
await request.CallBack(text.Length);
}
}
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, $"{filename}.mp3");
}
using (Stream file = new FileStream(filepath, FileMode.Create))
{
response.AudioContent.WriteTo(file);
}
return(filepath);
}
/// <summary>Snippet for SynthesizeSpeechAsync</summary>
public async Task SynthesizeSpeechRequestObjectAsync()
{
// Snippet: SynthesizeSpeechAsync(SynthesizeSpeechRequest, CallSettings)
// Additional: SynthesizeSpeechAsync(SynthesizeSpeechRequest, CancellationToken)
// Create client
TextToSpeechClient textToSpeechClient = await TextToSpeechClient.CreateAsync();
// Initialize request argument(s)
SynthesizeSpeechRequest request = new SynthesizeSpeechRequest
{
Input = new SynthesisInput(),
Voice = new VoiceSelectionParams(),
AudioConfig = new AudioConfig(),
EnableTimePointing =
{
SynthesizeSpeechRequest.Types.TimepointType.Unspecified,
},
};
// Make the request
SynthesizeSpeechResponse response = await textToSpeechClient.SynthesizeSpeechAsync(request);
// End snippet
}
async Task Run()
{
using (var cts = new CancellationTokenSource())
{
Console.InputEncoding = Encoding.Unicode;
Console.OutputEncoding = Encoding.Unicode;
Console.CancelKeyPress += (s, e) =>
{
e.Cancel = true;
cts.Cancel();
};
try
{
using var input = new StreamReader(Console.OpenStandardInput(), Encoding.Unicode);
using var api = PickBestApi(input, PortAudioHostApi.SupportedHostApis, cts.Token);
using var device = GetOutputDevice(input, api.Devices, cts.Token);
Console.Write("Saisir le code de langue voulu (défaut : fr-FR): ");
if (!ReadLine(input, cts.Token, out string language))
{
return;
}
if (string.IsNullOrWhiteSpace(language))
{
language = "fr-FR";
}
Console.WriteLine("Choisir le type de voix (femme/homme):");
Console.WriteLine("[0] - Non spécifié");
Console.WriteLine("[1] - Homme");
Console.WriteLine("[2] - Femme");
Console.WriteLine("[2] - Neutre");
if (!ReadLine(input, cts.Token, out string voiceGenderStr) || !int.TryParse(voiceGenderStr, out int voiceGenderInt))
{
return;
}
var voiceGender = (SsmlVoiceGender)voiceGenderInt;
Console.WriteLine($"Prêt à parler sur {device.Name}");
while (!cts.IsCancellationRequested)
{
Console.Write("->");
// Handle user input
if (!ReadLine(input, cts.Token, out string line) || line is null)
{
return;
}
Console.WriteLine("Récupération du son pour le texte : " + line);
Console.WriteLine("Son récupéré, lecture...");
using var audioStream = await TextToAudioStreamAsync(line, language, voiceGender, cts.Token);
using var pump = new PortAudioDevicePump(
device,
2,
new PortAudioSampleFormat(PortAudioSampleFormat.PortAudioNumberFormat.Signed, 2),
device.DefaultLowOutputLatency,
48000,
(buffer, offset, count) => audioStream.Read(buffer, offset, count));
using var handle = new ManualResetEventSlim(false);
pump.StreamFinished += FinishedHandler;
pump.Start();
handle.Wait();
pump.StreamFinished -= FinishedHandler;
void FinishedHandler(object sender, EventArgs eventArgs) => handle.Set();
Console.WriteLine("Lecture terminée");
}
}
catch (OperationCanceledException)
{
}
}
async Task <Stream> TextToAudioStreamAsync(string text, string language, SsmlVoiceGender voiceGender, CancellationToken token)
{
var request = new SynthesizeSpeechRequest
{
AudioConfig = new AudioConfig
{
AudioEncoding = AudioEncoding.OggOpus,
},
Input = new SynthesisInput
{
Text = text
},
Voice = new VoiceSelectionParams
{
LanguageCode = language,
SsmlGender = voiceGender,
},
};
var response = await ttsClient.SynthesizeSpeechAsync(request, token);
using (var opusStream = new MemoryStream())
{
response.AudioContent.WriteTo(opusStream);
opusStream.Position = 0;
var opusDecoder = new OpusDecoder(48000, 2);
var oggIn = new OpusOggReadStream(opusDecoder, opusStream);
var pcmStream = new MemoryStream();
while (oggIn.HasNextPacket)
{
short[] packet = oggIn.DecodeNextPacket();
if (packet != null)
{
for (int i = 0; i < packet.Length; i++)
{
var bytes = BitConverter.GetBytes(packet[i]);
pcmStream.Write(bytes, 0, bytes.Length);
}
}
}
pcmStream.Position = 0;
return(pcmStream);
}
}
}