---
name: "sphinx_cont_seg"
section: "1"
platform: "linux"
package: "sphinxbase-utils"
version: "0.8+5prealpha+1-21+b1"
release: "Debian 13 (Trixie)"
platforms: ["linux"]
url: "https://new.carta.tech/sphinx_cont_seg"
---

# sphinx\_cont\_seg(1)

Segment a waveform file into non-silence regions

## NAME

sphinx\_cont\_seg - Segment a waveform file into non-silence regions

## SYNOPSIS

**sphinx\_cont\_seg** \[ *options* \]...

## DESCRIPTION

This program reads an input file and segments it into individual non-silence regions. It can process either file or read data from microphone. Use following arguments:

- **-adcdev**

  of audio device to use for input.

- **-alpha**

  Preemphasis parameter

- **-argfile**

  file giving extra arguments.

- **-dither**

  Add 1/2-bit noise

- **-doublebw**

  Use double bandwidth filters (same center freq)

- **-frate**

  Frame rate

- **-infile**

  of audio file to use for input.

- **-input\_endian**

  Endianness of input data, big or little, ignored if NIST or MS Wav

- **-lifter**

  Length of sin-curve for liftering, or 0 for no liftering.

- **-logspec**

  Write out logspectral files instead of cepstra

- **-lowerf**

  Lower edge of filters

- **-ncep**

  Number of cep coefficients

- **-nfft**

  Size of FFT

- **-nfilt**

  Number of filter banks

- **-remove\_dc**

  Remove DC offset from each frame

- **-remove\_noise**

  Remove noise with spectral subtraction in mel-energies

- **-remove\_silence**

  Enables VAD, removes silence frames from processing

- **-round\_filters**

  Round mel filter frequencies to DFT points

- **-samprate**

  Sampling rate

- **-seed**

  Seed for random number generator; if less than zero, pick our own

- **-singlefile**

  a single cleaned file.

- **-smoothspec**

  Write out cepstral-smoothed logspectral files

- **-transform**

  Which type of transform to use to calculate cepstra (legacy, dct, or htk)

- **-unit\_area**

  Normalize mel filters to unit area

- **-upperf**

  Upper edge of filters

- **-vad\_postspeech**

  Num of silence frames to keep after from speech to silence.

- **-vad\_prespeech**

  Num of speech frames to keep before silence to speech.

- **-vad\_startspeech**

  Num of speech frames to trigger vad from silence to speech.

- **-vad\_threshold**

  Threshold for decision between noise and silence frames. Log-ratio between signal level and noise level.

- **-verbose**

  Show input filenames

- **-warp\_params**

  defining the warping function

- **-warp\_type**

  Warping function type (or shape)

- **-wlen**

  Hamming window length

## AUTHOR

Written by M. K. Ravishankar \<rkm@cs.cmu.edu\>.  This (rather lousy) manual page by David Huggins-Daines \<dhuggins@cs.cmu.edu\>

## COPYRIGHT

Copyright © 1999-2001 Carnegie Mellon University.  See the file *COPYING* included with this package for more information.\