Preprocessing modules

Global_commands	signal processing master routines
Select_commands	signal processing master routines
add_noise	Add a controlled amount of noise to the incoming signal
agc	Monitor the recording volume
am_cvt	Conversion of acoustic likelihoods/probabilities
anafft	Fast Fourier analysis and synthesis
asymm_filter	asymmetric filtering
aut2lpc	LPC from autocorrelation coefficients.
autocorr	Calculation of autocorrelation (and crosscorrelation) coefficients
cor_info	Extract information from a corpus file
deriv	Calculation of first and second time derivatives
envelope	Calculate peak and dip envelope of a power spectrum
feedback	Get feedback from the recognizer back into the preprocessing
filter_bank	Design and simulate filter banks
freq_warp	Transformation of a kid FFT power spectrum by including more formants
fun_eval	Mathematical transformations of parameters
histeq	Histogram based normalization of spectral features
idct	Inverse Discrete Cosine Transform
irr_filter	IIR filtering
lin_comb	Linear combinations of the parameters.
lpc2cep	Convertion of lpc filter coefficients to cepstral coefficients
math_fun	Execute some mathematical functions in a (ssp-)standard way,
matlab	Execution of a Matlab-routine in an SSP-script.
meannorm	Normalisation of parameters by subtraction of mean
merge_fr	Merge a number of consecutive frames into one frame
mv_svd_filt	Minimum Variance SVD-filtering of white noise in speech signals
mvg_eval	Evaluate a set of GMMs
n_mean	Calculation of noise mean and variance
noise_masking	Normalisation of the SNR of a (mel) spectrum.
norm_aut	Normalisation of autocorrelation parameters
online_bed	begin/end point detection of speech
ordstat	ordered statistics
pitch	Calculate pitch and voicing evidence
sam2trk	Default processing of a sample file
scale	Scaling of parameters
snr_norm	Normalisation of dynamic range of filterbank energies.
spec_impute	Impute a spectrum
speech_detect	Speech Detection
subsel_fr	Subselect frames
synfft	Fast Fourier analysis and synthesis
time_filter	Filtering of samples
trk2sam	trk2sam turn overlapping frames with samples into a samplefile
varfr	Variable frame rate analysis
vq	Vector Quantization and Labeling
vqmask	Impute a spectrum given a decomposed spectrum
vtl_estim	Estimation of warping factors for vocal tract length (VTL) normalisation.