private void btAnalyse_Click(object sender, EventArgs e)
{
WorkProgress wp = new WorkProgress();
wp.PBar = pbProg;
if (ckTrim.Checked)
{
TrimWaves((int)tbTrimThres.Value);
}
if (ckNormalize.Checked && cbNormType.SelectedIndex >= 0)
{
NormalizeWaves(cbNormType.SelectedIndex);
}
if (ckOverlap.Checked)
{
DetectBestOverlap((int)tbOverlapSamples.Value, wp);
}
/* if (AnalyzeBitwise(wp))
* {
* MessageBox.Show("Compared samples are BITWISE IDENTICAL\nNo more analysis needed", "Good", MessageBoxButtons.OK, MessageBoxIcon.Information);
* return;
* }*/
AnalyzeTime(wp);
AnalyzeSpectrum(wp);
}
public void DetectBestOverlap(int iInterval, WorkProgress iWP)
{
float CrossCorr = 0, BestCrossCorr = 0;
int bestI = 0;
lbCurStep.Text = "Detecting best overlap interval";
float[] data1 = file1.GetData()[cbChan1.SelectedIndex];
float[] data2 = file2.GetData()[cbChan2.SelectedIndex];
for (int i = -iInterval; i < iInterval; i++)
{
CrossCorr = MathLib.CrossCorrelation(data1, data2, i);
if (CrossCorr > BestCrossCorr)
{
BestCrossCorr = CrossCorr;
bestI = i;
lbBestOverlap.Text = i.ToString();
lbBestOverlap.Refresh();
}
iWP.UpdateProgress(100 * (i + iInterval) / 2 / iInterval);
}
if (bestI == 0)
{
return;
}
if (bestI > 0)
{
file2.SkipFirstSamples((uint)bestI);
}
else
{
file1.SkipFirstSamples((uint)Math.Abs(bestI));
}
lbInfo1.Text = file1.NSamples.ToString() + " samples";
lbInfo2.Text = file2.NSamples.ToString() + " samples";
w1.Refresh();
w2.Refresh();
}
public void AnalyzeTime(WorkProgress iWP)
{
lbCurStep.Text = "Time Domain Analysis";
float[] difArr;
float[] data1 = file1.GetData()[cbChan1.SelectedIndex];
float[] data2 = file2.GetData()[cbChan2.SelectedIndex];
iWP.UpdateProgress(0);
UInt32 len = Math.Min(file1.NSamples, file2.NSamples);
uint skip = 100;
UInt32 c = 0;
difArr = new float[len / skip + 1];
spTimeDiff.ForceYPeak = 100;
for (UInt32 i = 0; i < len; i += skip, c++)
{
float avg = 0;
for (UInt32 n = 0; n < skip && i + n < len; n++)
{
avg += Math.Abs(data1[i + n] - data2[i + n]);
}
avg /= skip;
difArr[c] = avg * 100 / file1.MaxAllowableValue;
if (c % 10 == 0)
{
iWP.UpdateProgress((int)(i * 100 / len));
spTimeDiff.FFTData = difArr;
spTimeDiff.Refresh();
}
}
spTimeDiff.FFTData = difArr;
spTimeDiff.Refresh();
lbAvgErr.Text = Math.Round(MathLib.ArrayAverage(difArr, difArr.Length), 5).ToString();
}
public bool AnalyzeBitwise(WorkProgress iWP)
{
lbCurStep.Text = "Bitwise Analysis";
iWP.UpdateProgress(0);
bool identical = true;
UInt32 len = Math.Min(file1.NSamples, file2.NSamples);
float[] data1 = file1.GetData()[cbChan1.SelectedIndex];
float[] data2 = file2.GetData()[cbChan2.SelectedIndex];
for (UInt32 i = 0; i < len; i++)
{
if (data1[i] != data2[i])
{
identical = false;
break;
;
}
}
iWP.UpdateProgress(100);
return(identical);
}
public void AnalyzeSpectrum(WorkProgress iWP)
{
lbCurStep.Text = "Frequency Domain Analysis";
iWP.UpdateProgress(0);
float[] data1 = file1.GetData()[cbChan1.SelectedIndex];
float[] data2 = file2.GetData()[cbChan2.SelectedIndex];
float[] globDiffArr = new float[MathLib.FFT_Samples];
UInt32 pos, steps;
UInt32 len = Math.Min(file1.NSamples, file2.NSamples);
UInt32 SkipSum;
switch (cbFFTQual.SelectedIndex)
{
case 0:
MathLib.SetQuality(MathLib.FFT_Quality.Low);
break;
case 1:
MathLib.SetQuality(MathLib.FFT_Quality.Average);
break;
case 2:
MathLib.SetQuality(MathLib.FFT_Quality.Good);
break;
case 3:
MathLib.SetQuality(MathLib.FFT_Quality.Best);
;
break;
default:
MathLib.SetQuality(MathLib.FFT_Quality.Average);
break;
}
switch (cbFFTPrecision.SelectedIndex)
{
case 0:
SkipSum = 1000;
break;
case 1:
SkipSum = 500;
break;
case 2:
SkipSum = 100;
break;
case 3:
SkipSum = 50;
break;
default:
SkipSum = 100;
break;
}
spDiff.OnlyPositive = false;
spDiff.ForceYPeak = 200;
spErr.ForceYPeak = 100;
float[] ErrorArr = new float[(long)Math.Ceiling((decimal)len / SkipSum)];
for (pos = 0, steps = 0; pos < len - MathLib.FFT_Samples - 1; pos += SkipSum, steps++)
{
float[] ft1 = MathLib.FFT(data1, pos);
float[] ft2 = MathLib.FFT(data2, pos);
sp1.FFTData = ft1;
sp2.FFTData = ft2;
float[] difArr = MathLib.SpectrumAttenuation_dB(ft1, ft2);
MathLib.ArraySum(ref globDiffArr, difArr);
float AbsErr = 0;
for (UInt32 i = 0; i < difArr.Length; i++)
{
AbsErr += Math.Abs(difArr[i]);
}
AbsErr /= difArr.Length;
ErrorArr[steps] = AbsErr;
spErr.FFTData = ErrorArr;
spErr.Refresh();
spDiff.FFTData = difArr;
spDiff.Refresh();
if (steps % 10 == 0)
{
iWP.UpdateProgress((int)(pos * 100 / len));
}
}
MathLib.ArrayRatio(ref globDiffArr, steps - 1);
spDiff.FFTData = globDiffArr;
spDiff.Refresh();
lbFreqAvgErr.Text = MathLib.ArrayAverage(ErrorArr, ErrorArr.Length).ToString();
}
