A command to get the cepstral peak prominence (CPP) of the selected PowerCepstrogram.

The returned value is the average of the cepstral peak prominences of the individual frames.

Settings

Subtract trend before smoothing

determines whether the smoothing should be performed on the Cepstrogram after the trend of each PowerCepstrum frame has been removed. Because, in general, the trends in the analysis frames will not be equal, the value of CPPS will be different whether or not smoothing is performed. If no smoothing is going on the result should be independent of this setting.

Time averaging window (s)

determines the width of the averaging window in the time domain. The result of the smoothing will be that in the new smoothed PowerCepstrogram each cepstral value is the average of the cepstral values within the averaging window that was positioned symmetrically around the center of this frame in the selected PowerCepstrogram. By chosing a value of zero, you can prevent any smoothing in the time dimension.

Quefrency averaging window (s)

determines the width of the averaging window in the quefrency domain. By chosing a value of zero, you can prevent any smoothing in the quefrency dimension.

Peak search pitch range (Hz)

determine the limits of the quefrency range where a peak is searched for. The lower quefrency is determined as 1 / pitchCeiling and this value is in general more critical than the value of the upper quefrency which equals 1 / pitchFloor. A pitchCeiling of 300 Hz will correspond to a lower quefrency of 1/300≈0.0033 seconds.

Tolerance

Interpolation

determines how the amplitude and position of a peak are determined.

Trend line quefrency range (s)

the quefrency range for which the amplitudes (in dB) will be modelled by a straight line. The lower value for this range in the Hillenbrand et al. (1994) article was chosen as 0.001 s in order to reduce the effect of very low quefrency data on the straight line fit. In our analysis this value is not so critical if we use the robust fitting procedure. If you choose the "Least squares" fit method then it matters more.

Trend type

defines how to model the cepstrum background. We can model it with a straight line as was done in Hillenbrand et al. (1994). The slope of this line will generally be negative because the background amplitudes get weaker for higher quefrencies. Or, we could use an exponential model in which the background cepstral amplitudes decay in a non-linear fashion.

Fit method

defines how the line that models the cepstrum background is calculated. The default method, "Robust slow", corresponds to Theil's robust line fit. The "Robust" method corresponds to the incomplete theil regression and is computationally faster but somewhat less precise. To be compatible with the past, a standard least squares line fit can also be chosen but it is much less precise than the other two because a least squares fit is much more influenced by the peak cepstral values than the other two.