Work on calculation of MFCCs

Issue #492 Minor changes to ensure that Delta and DoubleDelta features are being calculated when set true in the config file.

src/AnalysisPrograms/SpectrogramGenerator/Audio2Sonogram.Entry.cs

@@ -101,7 +101,7 @@ public static void Main(Arguments arguments)
     }
 
     /// <summary>
-    /// In line class used to return results from the static method Audio2Sonogram.GenerateFourSpectrogramImages().
+    /// In line class used to return results from the static method SpectrogramGenerator.GenerateSpectrogramImages().
     /// </summary>
     public class AudioToSonogramResult
     {

src/AnalysisPrograms/SpectrogramGenerator/SpectrogramGenerator.Core.cs

@@ -76,9 +76,6 @@ public static AudioToSonogramResult GenerateSpectrogramImages(
             // Default noiseReductionType = Standard
             var bgNoiseThreshold = config.BgNoiseThreshold;
 
-            // threshold for drawing the difference spectrogram
-            var differenceThreshold = config.DifferenceThreshold;
-
             // EXTRACT ENVELOPE and SPECTROGRAM FROM RECORDING SEGMENT
             var dspOutput1 = DSP_Frames.ExtractEnvelopeAndFfts(recordingSegment, frameSize, frameStep);
 
@@ -186,7 +183,10 @@ public static AudioToSonogramResult GenerateSpectrogramImages(
                 // IMAGE 5) draw difference spectrogram. This is derived from the original decibel spectrogram
                 if (@do.Contains(DifferenceSpectrogram))
                 {
+                    // threshold for drawing the difference spectrogram
                     //var differenceThreshold = configInfo.GetDoubleOrNull("DifferenceThreshold") ?? 3.0;
+                    var differenceThreshold = config.DifferenceThreshold;
+
                     var differenceImage = GetDifferenceSpectrogram(dbSpectrogramData, differenceThreshold);
                     differenceImage = BaseSonogram.GetImageAnnotatedWithLinearHertzScale(
                         differenceImage,
@@ -202,6 +202,7 @@ public static AudioToSonogramResult GenerateSpectrogramImages(
             // The default spectrogram has 64 frequency bands.
             if (@do.Contains(MelScaleSpectrogram))
             {
+                sonoConfig.mfccConfig.DoMelScale = true;
                 images.Add(
                     MelScaleSpectrogram,
                     GetMelScaleSpectrogram(sonoConfig, recordingSegment, sourceRecordingName));
@@ -210,9 +211,20 @@ public static AudioToSonogramResult GenerateSpectrogramImages(
             // IMAGE 7) Cepstral Spectrogram
             if (@do.Contains(CepstralSpectrogram))
             {
+                // TODO at present noise reduction type must be set = Standard.
+                // ... but use the NoiseReductionParameter that is set in the config file.
+                //sonoConfig.NoiseReductionParameter = 0.0;
+                sonoConfig.NoiseReductionType = NoiseReductionType.Standard;
+
+                sonoConfig.mfccConfig.DoMelScale = true;
+
+                // set the default number of cepstral coefficients
+                sonoConfig.mfccConfig.CcCount = 12;
+                sonoConfig.mfccConfig.IncludeDelta = config.IncludeDelta;
+                sonoConfig.mfccConfig.IncludeDoubleDelta = config.IncludeDoubleDelta;
                 images.Add(
                     CepstralSpectrogram,
-                    GetCepstralSpectrogram(sonoConfig, recordingSegment, sourceRecordingName));
+                    GetCepstrogram(sonoConfig, recordingSegment, sourceRecordingName));
             }
 
             // IMAGE 8) Octave-frequency scale Spectrogram
@@ -415,8 +427,9 @@ public static Image<Rgb24> GetMelScaleSpectrogram(
             string sourceRecordingName)
         {
             // TODO at present noise reduction type must be set = Standard.
+            //sonoConfig.NoiseReductionParameter = 3.0;
             sonoConfig.NoiseReductionType = NoiseReductionType.Standard;
-            sonoConfig.NoiseReductionParameter = 3.0;
+
             var melFreqGram = new SpectrogramMelScale(sonoConfig, recording.WavReader);
             var image = melFreqGram.GetImage();
             var titleBar = BaseSonogram.DrawTitleBarOfGrayScaleSpectrogram(
@@ -431,18 +444,15 @@ public static Image<Rgb24> GetMelScaleSpectrogram(
             return image;
         }
 
-        public static Image<Rgb24> GetCepstralSpectrogram(
+        public static Image<Rgb24> GetCepstrogram(
             SonogramConfig sonoConfig,
             AudioRecording recording,
             string sourceRecordingName)
         {
-            // TODO at present noise reduction type must be set = Standard.
-            sonoConfig.NoiseReductionType = NoiseReductionType.Standard;
-            sonoConfig.NoiseReductionParameter = 3.0;
-            var cepgram = new SpectrogramCepstral(sonoConfig, recording.WavReader);
-            var image = cepgram.GetImage();
+            var cepstrogram = new SpectrogramCepstral(sonoConfig, recording.WavReader);
+            var image = cepstrogram.GetImage();
             var titleBar = BaseSonogram.DrawTitleBarOfGrayScaleSpectrogram(
-                    "CEPSTRO-GRAM " + sourceRecordingName,
+                    "CEPSTROGRAM " + sourceRecordingName,
                     image.Width,
                     ImageTags[CepstralSpectrogram]);
             var startTime = TimeSpan.Zero;

src/AnalysisPrograms/SpectrogramGenerator/SpectrogramGeneratorConfig.cs

@@ -12,13 +12,23 @@ public class SpectrogramGeneratorConfig : AnalyzerConfig
 #pragma warning disable SA1623 // Property summary documentation should match accessors
         public int WaveformHeight { get; set; } = 100;
 
-        public double BgNoiseThreshold { get; set; } = 3.0;
+        public double BgNoiseThreshold { get; set; } = 0.0;
 
         /// <summary>
         /// DIFFERENCE SPECTROGRAM - PARAMETER (in decibels).
         /// </summary>
         public double DifferenceThreshold { get; set; } = 3.0;
 
+        /// <summary>
+        /// CEPSTROGRAM - PARAMETER.
+        /// </summary>
+        public bool IncludeDelta { get; set; } = false;
+
+        /// <summary>
+        /// CEPSTROGRAM - PARAMETER.
+        /// </summary>
+        public bool IncludeDoubleDelta { get; set; } = false;
+
         /// <summary>
         /// LOCAL CONTRAST NORMALIZATION PARAMETERS.
         /// </summary>

src/AudioAnalysisTools/DSP/MFCCStuff.cs

@@ -343,8 +343,9 @@ public static double InverseHerzTranform(double m, double c, double div)
             return binBounds;
         }
 
+        /*
         /// <summary>
-        /// Does MelFilterBank for passed sonogram matrix.
+        /// Calculates the MelFilterBank for passed sonogram matrix.
         /// IMPORTANT !!!!! Assumes that min freq of passed sonogram matrix = 0 Hz and maxFreq = Nyquist.
         /// Uses Greg's MelIntegral.
         /// </summary>
@@ -413,9 +414,10 @@ public static double InverseHerzTranform(double m, double c, double div)
 
             return outData;
         }
+        */
 
         /// <summary>
-        /// Does mel conversion for sonogram for any frequency band given by minFreq and maxFreq.
+        /// Does conversion from linear frequency scale to mel-scale for any frequency band given by minFreq and maxFreq.
         /// Uses Greg's MelIntegral
         /// The first step is to calculate the number of filters for the required frequency sub-band.
         /// </summary>
@@ -429,8 +431,7 @@ public static double InverseHerzTranform(double m, double c, double div)
             double freqRange = maxFreq - minFreq;
             if (freqRange <= 0)
             {
-                Log.WriteLine("Speech.MelFilterBank(): WARNING!!!! Freq range = zero");
-                throw new Exception("Speech.LinearFilterBank(): WARNING!!!! Freq range = zero. Check values of min & max freq.");
+                throw new Exception("FATAL ERROR: Speech.LinearFilterBank(): Freq range = zero. Check values of min & max freq.");
             }
 
             double melNyquist = Mel(nyquist);
@@ -514,7 +515,7 @@ public static double InverseHerzTranform(double m, double c, double div)
         //********************************************************************************************************************
         //********************************************************************************************************************
         //********************************************************************************************************************
-        //******************************* CEPTRA COEFFICIENTS USING DCT AND COSINES
+        //******************************* CALCULATION OF CEPTRAL COEFFICIENTS USING DCT AND COSINES
 
         public static double[,] Cepstra(double[,] spectra, int coeffCount)
         {

src/AudioAnalysisTools/StandardSpectrograms/SpectrogramMelScale.cs

@@ -60,10 +60,13 @@ public SpectrogramMelScale(AmplitudeSonogram sg, int minHz, int maxHz)
 
         /// <summary>
         /// Converts amplitude matrix to mel-frequency scale spectrogram.
+        /// IMPORTANT NOTE: The conversion to Mel-scale is done BEFORE noise reduction.
+        ///                 And conversion to decibels is done after noise reduction.
         /// </summary>
         /// <param name="amplitudeM">Matrix of amplitude values.</param>
         public override void Make(double[,] amplitudeM)
         {
+            // call static method to convert amplitude spectrogram to melscale.
             var m = MakeMelScaleSpectrogram(this.Configuration, amplitudeM, this.SampleRate);
 
             //(iii) NOISE REDUCTION
@@ -79,21 +82,24 @@ public override void Make(double[,] amplitudeM)
         //##################################################################################################################################
 
         /// <summary>
+        /// Converts an amplitude spectrogram to mel-scale spectrogram.
         /// NOTE!!!! The decibel array has been normalised in 0 - 1.
         /// </summary>
-        public static double[,] MakeMelScaleSpectrogram(SonogramConfig config, double[,] matrix, int sampleRate)
+        public static double[,] MakeMelScaleSpectrogram(SonogramConfig config, double[,] amplitudeM, int sampleRate)
         {
-            double[,] m = matrix;
+            double[,] m = amplitudeM;
             int nyquist = sampleRate / 2;
             double epsilon = config.epsilon;
 
             //(i) APPLY FILTER BANK
             //number of Hz bands = 2^N +1. Subtract DC bin
             int fftBinCount = config.FreqBinCount;
 
+            // This config has the default values set for calculating MFCCs.
             // Mel band count is set to 64 by default in BaseSonogramConfig class at line 158.
+            // Coefficient count is set to 12 by default.
             int bandCount = config.mfccConfig.FilterbankCount;
-            Log.WriteIfVerbose("ApplyFilterBank(): Dim prior to filter bank  =" + matrix.GetLength(1));
+            Log.WriteIfVerbose("ApplyFilterBank(): Dim prior to filter bank  =" + amplitudeM.GetLength(1));
 
             //error check that filterBankCount < Number of FFT bins
             if (bandCount > fftBinCount)
@@ -104,12 +110,13 @@ public override void Make(double[,] amplitudeM)
             }
 
             //this is the filter count for full bandwidth 0-Nyquist. This number is trimmed proportionately to fit the required bandwidth.
+            // The last two arguments set the frequency band.
             m = MFCCStuff.MelFilterBank(m, bandCount, nyquist, 0, nyquist);
 
             Log.WriteIfVerbose("\tDim after filter bank=" + m.GetLength(1) + " (Max filter bank=" + bandCount + ")");
 
             //(ii) CONVERT AMPLITUDES TO DECIBELS
-            m = MFCCStuff.DecibelSpectra(m, config.WindowPower, sampleRate, epsilon); //from spectrogram
+            m = MFCCStuff.DecibelSpectra(m, config.WindowPower, sampleRate, epsilon);
             return m;
         }