private void speakNative(Model.Wrapper wrapper)
{
#if UNITY_STANDALONE_WIN || UNITY_EDITOR_WIN
SpVoice speechSynthesizer = new SpVoice();
SpObjectToken voice = getSapiVoice(wrapper.Voice);
if (voice != null)
{
speechSynthesizer.Voice = voice;
}
speechSynthesizer.Volume = calculateVolume(wrapper.Volume);
speechSynthesizer.Rate = calculateRate(wrapper.Rate);
System.Threading.Thread worker = new System.Threading.Thread(() => speechSynthesizer.Speak(prepareText(wrapper), wrapper.ForceSSML && !Speaker.isAutoClearTags ? SpeechVoiceSpeakFlags.SVSFIsXML : SpeechVoiceSpeakFlags.SVSFIsNotXML));
worker.Start();
do
{
System.Threading.Thread.Sleep(50);
} while (worker.IsAlive && !silence);
if (silence)
{
speechSynthesizer.Skip("Sentence", int.MaxValue);
}
//speechSynthesizer.WaitUntilDone(int.MaxValue);
#endif
}
private void speakToFile(Model.Wrapper wrapper, string outputFile)
{
#if UNITY_STANDALONE_WIN || UNITY_EDITOR_WIN
SpVoice speechSynthesizer = new SpVoice();
SpFileStream spFile = new SpFileStream();
spFile.Format.Type = SpeechAudioFormatType.SAFT48kHz16BitStereo;
spFile.Open(outputFile, SpeechStreamFileMode.SSFMCreateForWrite);
SpObjectToken voice = getSapiVoice(wrapper.Voice);
if (voice != null)
{
speechSynthesizer.Voice = voice;
}
speechSynthesizer.AudioOutputStream = spFile;
speechSynthesizer.Volume = calculateVolume(wrapper.Volume);
speechSynthesizer.Rate = calculateRate(wrapper.Rate);
speechSynthesizer.Speak(prepareText(wrapper), wrapper.ForceSSML && !Speaker.isAutoClearTags ? SpeechVoiceSpeakFlags.SVSFIsXML : SpeechVoiceSpeakFlags.SVSFIsNotXML);
speechSynthesizer.WaitUntilDone(int.MaxValue);
spFile.Close();
#endif
}
private void Initialize()
{
try
{
voice = new SpVoiceClass();
voice.EndStream += Voice_EndStream;
ISpeechObjectTokens objTokens = voice.GetVoices("", "");
const string useVoice = "ScanSoft Mei-Ling_Full_22kHz";
int useIndex = -1;
for (int i = 0; i < objTokens.Count; i++)
{
SpObjectToken sot = objTokens.Item(i);
if (sot.GetDescription(0) == useVoice)
{
useIndex = i;
break;
}
}
if (useIndex == -1)
{
useIndex = 0;
}
voice.Voice = objTokens.Item(useIndex);
}
catch (Exception e)
{
MessageBox.Show("Error:" + e.Message);
}
}
public void Save(string filePath, string text)
{
try
{
SpObjectToken backupSapi = null;
SpFileStream ss = new SpFileStream();
ss.Open(filePath, SpeechStreamFileMode.SSFMCreateForWrite);
sapi.AudioOutputStream = ss;
Thread t = new Thread(() => {
backupSapi = sapi.Voice;
sapi.Voice = SpeakerList[AvatorIdx];
sapi.Rate = Speed;
sapi.Volume = Volume;
sapi.Speak(text);
sapi.Voice = backupSapi;
});
t.SetApartmentState(ApartmentState.STA);
t.Start();
t.Join();
ss.Close();
}
catch (Exception e)
{
throw new Exception(string.Format("保存処理で落ちたっす。{0}", e.Message));
}
}
public void Talk(string text, bool asyncFlag = false)
{
try
{
SpObjectToken backupSapi = null;
Thread t = new Thread(() =>
{
backupSapi = sapi.Voice;
sapi.AudioOutput = OutputDeviceList[DevIdx];
sapi.Voice = SpeakerList[AvatorIdx];
sapi.Rate = Speed;
sapi.Volume = Volume;
sapi.Speak(text);
sapi.Voice = backupSapi;
});
t.SetApartmentState(ApartmentState.STA);
t.Start();
if (!asyncFlag)
{
t.Join();
}
}
catch (Exception e)
{
throw new Exception(string.Format("発声処理で落ちたっす。{0}", e.Message));
}
}
/// <summary>Creates a speaker with the specified voice</summary>
/// <param name="Name"></param>
/// <param name="Color"></param>
/// <param name="Voice"></param>
public Actor(string Name, ConsoleColor Color, SpObjectToken Voice)
{
SAPI = new SpVoice {
Voice = Voice
};
this.Name = Name;
this.Color = Color;
}
//マイクから読み取るため、マイク用のデバイスを指定する
private SpObjectToken CreateMicrofon()
{
SpeechLib.SpObjectTokenCategory objAudioTokenCategory = new SpeechLib.SpObjectTokenCategory();
objAudioTokenCategory.SetId(@"HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech Server\v11.0\AudioInput", false);
SpObjectToken objAudioToken = new SpObjectToken();
objAudioToken.SetId(objAudioTokenCategory.Default, @"HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech Server\v11.0\AudioInput", false);
//return null;
return(objAudioToken);
}
private SpObjectToken CreateMicrofon()
{
SpObjectTokenCategory objAudioTokenCategory = new SpObjectTokenCategory();
objAudioTokenCategory.SetId(@"HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech\AudioInput");
SpObjectToken objAudioToken = new SpObjectToken();
objAudioToken.SetId(objAudioTokenCategory.Default, @"HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech\AudioInput");
return(objAudioToken);
}
private void Voice_VoiceChange(int StreamNumber, object StreamPosition, SpObjectToken VoiceObjectToken)
{
int index = VoiceCombo.FindStringExact(VoiceObjectToken.GetDescription(0));
if (index != -1)
{
VoiceCombo.SelectedIndex = index;
}
AddEventMessage("VoiceChange: Voice = " + VoiceObjectToken.GetDescription(0));
}
public void initialize_audio()
{
SpObjectTokenCategory category = new SpObjectTokenCategory();
category.SetId(SpeechStringConstants.SpeechCategoryAudioIn, false);
SpObjectToken token = new SpObjectToken();
token.SetId(category.EnumerateTokens().Item(0).Id, "", false);
//token.SetId(category.Default, "", false);
context.Recognizer.AudioInput = token;
}
public override int runCommand(XdmNode node)
{
Speech sp = Speech.instance;
//SpeechSynthesizer sp = new SpeechSynthesizer();
/*
* Store values
*/
SpObjectToken prevVoice = sp.speech.Voice;
int prevVolume = sp.speech.Volume;
int prevRate = sp.speech.Rate;
foreach (string key in getProperties().Keys)
{
string value = getProperties()[key];
switch (key)
{
case "voice":
ISpeechObjectToken voice = sp.getVoice(value);
if (voice != null)
{
sp.speech.Voice = (SpObjectToken)voice;
}
break;
case "volume":
int volume = Int32.Parse(value);
if (volume > -1 && volume < 101)
{
sp.speech.Volume = volume;
}
break;
case "rate":
int rate = Int32.Parse(value);
if (rate > -11 && rate < 11)
{
sp.speech.Rate = rate;
}
break;
}
}
runChilds(node);
if (commands.Count > 0)
{
sp.speech.Voice = prevVoice;
sp.speech.Volume = prevVolume;
sp.speech.Rate = prevRate;
}
return(0);
}
static void Main(string[] args)
{
try
{
SpVoice sapi = new SpVoice();
SpObjectTokenCategory sapiCat = new SpObjectTokenCategory();
Dictionary <string, SpObjectToken> TokerPool = new Dictionary <string, SpObjectToken>();
SpFileStream Sfs = null;
// See https://qiita.com/7shi/items/7781516d6746e29c03b4
sapiCat.SetId(@"HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech_OneCore\Voices", false);
// Narrator voices
foreach (SpObjectToken token in sapiCat.EnumerateTokens())
{
if (!TokerPool.ContainsKey(token.GetAttribute("name")))
{
TokerPool.Add(token.GetAttribute("name"), token);
}
}
// SAPI voices
foreach (SpObjectToken token in sapi.GetVoices("", ""))
{
if (!TokerPool.ContainsKey(token.GetAttribute("name")))
{
TokerPool.Add(token.GetAttribute("name"), token);
}
}
SpeakerList = TokerPool.Select((val, idx) => new { Key = idx, Value = val.Value }).ToDictionary(s => s.Key, s => s.Value);
if (!Opt(args))
{
return;
}
if (file != "")
{
Sfs = new SpFileStream();
Sfs.Open(file, SpeechStreamFileMode.SSFMCreateForWrite);
sapi.AudioOutputStream = Sfs;
}
Thread t = new Thread(() => {
SpObjectToken backupSapi = sapi.Voice;
sapi.Voice = SpeakerList[AvatorIdx];
sapi.Rate = Speed;
sapi.Volume = Volume;
sapi.Speak(text);
sapi.Voice = backupSapi;
});
t.SetApartmentState(ApartmentState.STA);
t.Start();
t.Join();
if (file != "")
{
Sfs.Close();
}
}
catch (Exception e)
{
Console.WriteLine("err:{0},{1}", e.Message, e.InnerException == null ? "" : e.InnerException.Message);
}
}
//button5生成读音
private void button5_Click(object sender, RoutedEventArgs e)
{
//数据库全部读出
WordLogic[] wordList = WordLogic.FindAll();
SpFileStream SpFileStream = new SpFileStream();
String fileName;
SpeechVoiceSpeakFlags SpFlags = SpeechVoiceSpeakFlags.SVSFlagsAsync;
SpeechStreamFileMode SpFileMode = SpeechStreamFileMode.SSFMCreateForWrite;
SpVoice voice = new SpVoice();
// prepare voice
SpObjectTokenCategory aotc = new SpObjectTokenCategory();
aotc.SetId(SpeechLib.SpeechStringConstants.SpeechCategoryVoices);
foreach (ISpeechObjectToken token in aotc.EnumerateTokens())
{
if (token.GetDescription() == "VW Julie")
{
englishToken = (SpObjectToken)token;
}
else if (token.GetDescription() == "VW Hui")
{
chineseToken = (SpObjectToken)token;
}
}
voice.Voice = englishToken;
voice.Rate = -4;
String outFolderPath = Directory.GetParent("../../TestCases/") + @"\单词音频\";
if (!Directory.Exists(outFolderPath))
{
Directory.CreateDirectory(outFolderPath);
}
for (int i = 0; i < wordList.Length; i++)
{
String word = wordList[i].word;
if (word != null)
{
word = word.Trim();
}
if (String.IsNullOrEmpty(word))
{
// 遇到无效内容，退出
continue;
}
word = convert(word);
fileName = outFolderPath + word + ".wav";
SpFileStream.Open(fileName, SpFileMode, false);
voice.AudioOutputStream = SpFileStream;
voice.Speak(word, SpFlags);
voice.WaitUntilDone(Timeout.Infinite);
SpFileStream.Close();
}
MessageBox.Show("音频生成完毕！");
}
public SpeechService(ServiceCreationInfo info)
: base("speech", info)
{
mVoice = new SpVoice();
// Select voice
string voiceName = null;
try
{
voiceName = info.Configuration.Voice;
}
catch (RuntimeBinderException) {}
if (!string.IsNullOrEmpty(voiceName))
{
SpObjectToken voiceToken = null;
CultureInfo culture = new CultureInfo("en-US");
foreach (var voice in mVoice.GetVoices())
{
var token = voice as SpObjectToken;
if (token == null)
{
continue;
}
if (culture.CompareInfo.IndexOf(token.Id, voiceName, CompareOptions.IgnoreCase) < 0)
{
continue;
}
voiceToken = token;
}
if (voiceToken != null)
{
mVoice.Voice = voiceToken;
}
}
// Select output. Why isn't this default any longer?
var enumerator = new MMDeviceEnumerator();
MMDevice endpoint = enumerator.GetDefaultAudioEndpoint(DataFlow.Render, Role.Console);
if (endpoint != null)
{
foreach (var output in mVoice.GetAudioOutputs())
{
var token = output as SpObjectToken;
if (token == null)
{
continue;
}
if (token.Id.IndexOf(endpoint.ID) < 0)
{
continue;
}
mVoice.AudioOutput = token;
break;
}
}
mVoiceCommands = new Dictionary <string, DeviceBase.VoiceCommand>();
mInput = new AudioInput();
mRecognizer = new SpeechRecognitionEngine();
mRecognizer.SpeechRecognized += new EventHandler <SpeechRecognizedEventArgs>(OnSpeechRecognized);
mRecognizer.RecognizerUpdateReached += new EventHandler <RecognizerUpdateReachedEventArgs>(OnUpdateRecognizer);
mRecognizer.RecognizeCompleted += new EventHandler <RecognizeCompletedEventArgs>(OnRecognizeCompleted);
var grammar = new Grammar(new GrammarBuilder(new Choices(new string[] { "computer" })));
mRecognizer.LoadGrammar(grammar);
var speechFormat = new SpeechAudioFormatInfo(44100, AudioBitsPerSample.Sixteen, AudioChannel.Mono);
mRecognizer.SetInputToAudioStream(mInput.mStream, speechFormat);
mRecognizer.RecognizeAsync(RecognizeMode.Multiple);
}
