pppeak - pitch tracking for EGG data
pppeak [-a val] [-c ch] [-i] [-l wsize] [-o offset] [-t thresh] [-v] [-w] [-z coef] filename1[.wav] filename2[.pps]
pppeak() analyzes the data in wavefile filename1 and writes a
pitch period file into filename2. The pitch period algorithm is a
peak detection method; the first peak above the specified threshold after a
positive-going zero crossing is identified as the beginning of a pitch period,
and a pitch mark will be written to the output file at this location.
An offset may be added to this location. Before a peak is detected,
the wavefile data is filtered to remove DC and noise as follows:
Y(t) = x(t) - x(t-1) + Z * Y(t-1)
where Z is a coefficient (default 1.0)
specified as an option by -z
The user may specify the threshold and also change the direction of the
peak detection (search for negative-going peaks).
Note - this program was written to provide pitch tracking for EGG data
only. The peak detection algorithm will work very poorly on speech or
other periodic waveforms.
sets the automatic threshold level, used when an absolute threshold (see -t)
is not specified. val must be a fraction between 0.0 and 1.0; the
value sets the threshold, with respect to the maximum peak in the file.
The default value is 0.15 (15% of maximum).
selects the active channel. The default is channel 1 for a multichannel file.
inverts the input data, so the pitch mark is aligned with a negative peak.
The default is a positive peak.
specifies the size of the window in ms. default is 10 ms.
adds an offset in number of samples to the pitch mark location
selects the threshold level for detecting a peak; the value may be 0 to 32767.
If not specified, the program will select a threshold based on the maximum
peak value and the -a parameter (default is 15% of maximum value)
specifies a high pass filtering window applied to the data before the
verbose mode (prints out state of internal parameters)
coefficient for filter (see above)