我正在尝试编写一个应用程序来收听我的计算机音频并使用 Google 语音识别对其进行转录。
我已经能够使用录制系统声音WasapiLoopbackCapture
并且我已经能够使用带有测试文件的谷歌流识别api,但我无法将两者合并在一起。
当我将音频从WasapiLoopbackCapture
google 流式传输时,它不会返回任何结果。
private WaveInEvent StartListening()
{
var waveIn = new WaveInEvent
{
DeviceNumber = 0,
WaveFormat = new WaveFormat(SampleRate, ChannelCount)
};
waveIn.DataAvailable += (sender, args) =>
_microphoneBuffer.Add(ByteString.CopyFrom(args.Buffer, 0, args.BytesRecorded));
waveIn.StartRecording();
return waveIn;
}
并调整它以使用 WasapiLoopbackCapture:
private IDisposable StartListening()
{
var waveIn = new WasapiLoopbackCapture();
//var waveIn = new WaveInEvent
//{
// DeviceNumber = 0,
// WaveFormat = new WaveFormat(SampleRate, ChannelCount)
//};
SampleRate = waveIn.WaveFormat.SampleRate;
ChannelCount = waveIn.WaveFormat.Channels;
BytesPerSecond = SampleRate * ChannelCount * BytesPerSample;
Console.WriteLine(SampleRate);
Console.WriteLine(BytesPerSecond);
waveIn.DataAvailable += (sender, args) =>
_microphoneBuffer.Add(ByteString.CopyFrom(args.Buffer, 0, args.BytesRecorded));
waveIn.StartRecording();
return waveIn;
}
但它不会返回任何转录文本。
我已将输入流保存到文件中,并且播放正常 - 所以声音到了,我猜是从 WasapiLoopback 接收到的 waveFormat 与谷歌喜欢的不兼容 - 我尝试了一些转换并且不能让它工作。
我已经查看了有关堆栈溢出的以下主题,但仍然无法使其正常工作: 重新采样 WasapiLoopbackCapture Naudio - Convert 32 bit wav to 16 bit wav
并尝试将它们结合起来:
private IDisposable StartListening()
{
var waveIn = new WasapiLoopbackCapture();
//var waveIn = new WaveInEvent
//{
//DeviceNumber = 0,
//WaveFormat = new WaveFormat(SampleRate, ChannelCount)
//};
// SampleRate = waveIn.WaveFormat.SampleRate;
// ChannelCount = waveIn.WaveFormat.Channels;
// BytesPerSecond = waveIn.WaveFormat.AverageBytesPerSecond;// SampleRate * ChannelCount * BytesPerSample;
var target = new WaveFormat(SampleRate, 16, 1);
var writer = new WaveFileWriter(@"c:\temp\xx.wav", waveIn.WaveFormat);
Console.WriteLine(SampleRate);
Console.WriteLine(BytesPerSecond);
var stop = false;
waveIn.DataAvailable += (sender, args) =>
{
var a = args;
byte[] newArray16Bit = new byte[args.BytesRecorded / 2];
short two;
float value;
for (int i = 0, j = 0; i < args.BytesRecorded; i += 4, j += 2)
{
value = (BitConverter.ToSingle(args.Buffer, i));
two = (short)(value * short.MaxValue);
newArray16Bit[j] = (byte)(two & 0xFF);
newArray16Bit[j + 1] = (byte)((two >> 8) & 0xFF);
}
var resampleStream = new NAudio.Wave.Compression.AcmStream(new WaveFormat(waveIn.WaveFormat.SampleRate
,16,waveIn.WaveFormat.Channels), target);
Buffer.BlockCopy(newArray16Bit, 0, resampleStream.SourceBuffer, 0, a.BytesRecorded/2);
int sourceBytesConverted = 0;
var bytes = resampleStream.Convert(a.BytesRecorded/2, out sourceBytesConverted);
var converted = new byte[bytes];
Buffer.BlockCopy(resampleStream.DestBuffer, 9, converted,0, bytes);
a = new WaveInEventArgs(converted,bytes);
_microphoneBuffer.Add(ByteString.CopyFrom(a.Buffer, 0, a.BytesRecorded));
if (writer != null)
{
writer.Write(a.Buffer, 0, a.BytesRecorded);
if (writer.Position > waveIn.WaveFormat.AverageBytesPerSecond * 5)
{
stop = true;
writer.Dispose();
writer = null;
Console.WriteLine("Saved file");
}
}
};
waveIn.StartRecording();
return waveIn;
}
但它不起作用。
我不确定这是否是正确的道路。
一个修复的代码示例将不胜感激
我尝试转换比特率等。但无法让它工作。