Possible bug in calculation of ACtivity and event count Summary indices

[towsey]

Describe the bug

In the following method:
public static SummaryActivity CalculateActivity(double[] dBarray, TimeSpan frameStepDuration, double dbThreshold)
at line 84 of ActivityAndCover.cs, the dbThreshold argument is not used. Instead the default value of 6dB is used.
That means the user appears to have but actually has no control over this value.

Expected behavior

User should be able to set value of threshold and expect it to be used when calling this method.

Additional Details

This same method calculates both activity and event count.
The calculation of events seems unnecessarily complicated and could be simplified.

[towsey]

The calculation of summary indices derived from the waveform depends on a calculation of the envelope. Currently we take the max value in any frame. But the following is a better approach as it removes DC bias.
dB re full scale = 20 * log10((max-min)/65536)
This can be changed at this same time.

[atruskie]

We use normalised values so the 65536 constant is incorrect there. Otherwise sounds good.

[towsey]

Also at line 144 in DSP_Frames.cs, the FFT window is set to Hamming whereas elsewhere the window has been set to Hanning.
string windowName = FFT.KeyHammingWindow;
The window name should default to Hanning.

[towsey]

It is now apparent that we are using the Hamming window in many places and Hanning in others.