public static Signal ObterSinalMono(string arquivoAudio)
{
var audioDecoder = new WaveDecoder(arquivoAudio);
using (var originalSignal = audioDecoder.Decode())
{
var monoFilter = new MonoFilter();
return(monoFilter.Apply(originalSignal));
}
}
public override void Awake()
{
base.Awake();
filters = new MonoFilter[NUM_FILTERS];
bufferCopy = new float[MAX_BUFFER_LENGTH];
controlBuffer = new float[MAX_BUFFER_LENGTH];
//primary stereo filter
filters[0] = new MonoFilter(frequency[0], resonance);
filters[1] = new MonoFilter(frequency[0], resonance);
//secondary stereo filter (for BP/notch)
filters[2] = new MonoFilter(frequency[1], resonance);
filters[3] = new MonoFilter(frequency[1], resonance);
}
/// <summary>
/// This method should be implemented by inheriting classes to implement the
/// actual feature extraction, transforming the input image into a list of features.
/// </summary>
///
protected override IList <MelFrequencyCepstrumCoefficientDescriptor> InnerTransform(Signal signal)
{
if (signal.NumberOfChannels > 1)
{
signal = new MonoFilter().Apply(signal);
}
float[] w_fSig = new float[signal.Length];
signal.CopyTo(w_fSig);
int w_nfr = (int)(signal.Length / m_fshift + 1);
var w_mfcc = new MelFrequencyCepstrumCoefficientDescriptor[w_nfr];
for (int w_fr = 0; w_fr < w_nfr; w_fr++)
{
int w_start = (int)System.Math.Round(w_fr * m_fshift);
int w_end = System.Math.Min(signal.Length, w_start + m_wlen);
Int16[] w_frame = new Int16[w_end - w_start];
int w_j = 0;
for (int w_i = w_start; w_i < w_end; w_i++, w_j++)
{
w_frame[w_j] = (Int16)(w_fSig[w_i] * 32768f);
}
int w_len = w_frame.Length;
if (w_len < m_wlen)
{
Array.Resize(ref w_frame, m_wlen);
for (int w_i = w_len; w_i < m_wlen; w_i++)
{
w_frame[w_i] = 0;
}
}
double[] w_s2mfc = frame2s2mfc(w_frame);
w_mfcc[w_fr] = new MelFrequencyCepstrumCoefficientDescriptor(w_fr, w_s2mfc);
}
return(w_mfcc);
}
protected string ParseReferenceStringRight(MonoFilter f)
{
ReferenceFilterValue rfv = (ReferenceFilterValue)f.Right;
return(ParseReferenceFilterValue(rfv));
}