Name
Vad -- voice activity detector
Synopsis
vad [options]
Function
Voice Activity Detector. Attempts to trim silence and quiet background sounds from the ends of (fairly high resolution i.e. 16-bit, 44-48kHz) recordings of speech. The algorithm currently uses a simple cepstral power measurement to detect voice, so may be fooled by other things, especially music. The effect can trim only from the front of the audio, so in order to trim from the back, the reverse effect must also be used. E.g.

 
norm vad

to trim from the front,

 
norm reverse vad reverse

to trim from the back, and

 
norm vad reverse vad reverse

to trim from both ends. The use of the norm effect is recommended, but remember that neither reverse nor norm is suitable for use with streamed audio.

Options: Default values are shown in parenthesis.

-t num (7)
The measurement level used to trigger activity detection. This might need to be changed depending on the noise level, signal level and other charactistics of the input audio.

-T num (0.25)
The time constant (in seconds) used to help ignore short bursts of sound.

-s num (1)
The amount of audio (in seconds) to search for quieter/shorter bursts of audio to include prior to the detected trigger point.

-g num (0.25)
Allowed gap (in seconds) between quieter/shorter bursts of audio to include prior to the detected trigger point.

-p num (0)
The amount of audio (in seconds) to preserve before the trigger point and any found quieter/shorter bursts.

Advanced Options: These allow fine tuning of the algorithm's internal parameters.

-b num
The algorithm (internally) uses adaptive noise estimation/reduction in order to detect the start of the wanted audio. This option sets the time for the initial noise estimate.

-N num
Time constant used by the adaptive noise estimator for when the noise level is increasing.

-n num
Time constant used by the adaptive noise estimator for when the noise level is decreasing.

-r num
Amount of noise reduction to use in the detection algorithm (e.g. 0, 0.5, ...).

-f num
Frequency of the algorithm's processing/measurements.

-m num
Measurement duration; by default, twice the measurement period; i.e. with overlap.

-M num
Time constant used to smooth spectral measurements.

-h num
"Brick-wall" frequency of high-pass filter applied at the input to the detector algorithm.

-l num
"Brick-wall" frequency of low-pass filter applied at the input to the detector algorithm.

-H num
"Brick-wall" frequency of high-pass lifter used in the detector algorithm.

-L num
"Brick-wall" frequency of low-pass lifter used in the detector algorithm.

See also the silence effect.


Show TOC