filter2.1

来自「speech signal process tools」· 1 代码 · 共 276 行

.\" Copyright (c) 1992 Entropic Research Laboratory, Inc.; All rights reserved
.\" @(#)filter2.1	1.9 9/1/98 ERL
.ds ]W (c) 1992 Entropic Research Laboratory, Inc.
.TH FILTER2 1\-ESPS 11/19/92
.SH NAME
.nf
filter2 \- Performs digital filtering on a sampled data file.
filter \- same as \fIfilter2\fR
.fi
.SH SYNOPSIS
.B filter[2]
[
.BI \-P " param_file"
]
[
.BI \-p " range"
]
[
.BI \-r " range"
]
[
.BI \-d " data_type"
]
[
.BI \-f " filter_string"
]
[
.BI \-F " filt_file"
]
[
.BI \-x " debug_level"
]
[
.B \-z 
]
.I input.sd output.sd
.SH DESCRIPTION
.PP
The program \fIfilter2\fR performs a digital filtering operation on 
the input sampled data FEA_SD(5\-\s-1ESPS\s+1) file, 
\fIinput.sd\fR, and produces an output sampled data file, \fIoutput.sd\fR.
The output sampled data file is of the type FEA_SD(5\-\s-1ESPS\s+1).
.PP
If \fIinput.sd\fR is replaced by "-", the standard input is  read.
If  \fIoutput.sd\fR  is  replaced  by  "-", the standard output is
written.  If \fIinput.sd\fR is  missing  from  the  command  line,
\fIfilter2\fR gets its name from ESPS Common.
.PP
This program may implement either finite impulse response (FIR) or
infinite impulse response (IIR) filters.  The filter data may come
from a FEA_FILT file specified by the \fB\-F\fR option, or it may also be
supplied in the parameter file by the \fB\-P\fR option.  
Currently, only real filters are supported.
.PP
In the IIR system, the filter is realized by cascading the second order 
sub-systems of pairs of poles and zeros and their complex conjugate
counterparts.  Each sub-system is implemented in the direct form II.  
This is to insure the numerical stability of the system.
.PP
In the 
case when only the numerator and denominator coefficients of H(z) are
available, the program will implement the IIR system in the direct 
form using these coefficients. However, this implementation is 
numerically unstable.
.PP
In FIR system, the filter is realized in the direct form.
.PP
All data samples occurring before the first sample in the input file 
are assumed to be zero.
.PP
.SH OPTIONS
The following options are supported:
.TP
.BI \-P " param_file"
uses the parameter file 
\fIparam_file\fR instead of the default, which is \fIparams\fR.
.TP
.BI \-r " first:last"
.TP
.BI \-r " first:+incr"
Determines the range of points from input file for filtering.  In
the first form, a pair of unsigned integers gives the first and last points
of the range.  If \fIfirst\fR is omitted, 1 is used.  If \fIlast\fR is
omitted, the last point in the file is used.  The second form is equivalent
to the first with \fIlast = first + incr\fR.
.TP
.BI \-p " "
Same as the \fB-r\fR option.
.TP
.BI \-d  " data_type \fR[input data type]\fR"
This option specifies the data type of \fIoutput.sd\fR.  Available data types
are DOUBLE, FLOAT, LONG, SHORT, BYTE, DOUBLE_CPLX, FLOAT_CPLX, LONG_CPLX,
SHORT_CPLX, and BYTE_CPLX.  By default, output data type is the same as
the input data type.  
.TP
.BI \-f " filter_string \fR[filter]\fR"
If the coefficients are being read from the parameter file, then
\fIfilt_string\fR is the body of the name of the variable that contains
filter coefficients.  See the \fBESPS_PARAMETERS\fR section.  For example,
if this option is applied as \fI-f myfilter\fR, then the value of the
\fImyfilter_psiz\fR variable in the parameter file will be taken as the
number of poles.
.IP
If the coefficients are being read from a FEA_FILT file, then
\fIfilter_string\fR is the number of the filter record to use.
The default name in this case is 1, the first record in the file.
.TP
.BI \-F " filt_file"
Read the coefficients from the
.IR FEA_FILT (5\-ESPS)
file \fIfilt_file\fR
rather than from the parameter file.
.TP
.BI \-x " debug_level \fR[0]\fR"
If \fIdebug_level\fR is positive, \fIfilter\fR prints debugging 
messages and other information on the standard error
output.  The messages proliferate as the \fIdebug_level\fR
increases.  If \fIdebug_level\fR is 0 (the default), no messages are
printed.
.TP
.BI \-z " "
By specifying \fB\-z\fR,
the output \fIstart_time\fR generic value is reduced by the value of the
\fIdelay_samples\fR generic header value of input filter file, 
divided by the sampling rate of input sampled data. 
This often helps time align a filtered
signal with the input signal.  If \fIdelay_samples\fR is not defined 
in the input file header, a value of 0 is assumed.
.SH ESPS PARAMETERS
The following parameters are read from the parameter file:
.TP
.I start - integer
The first point in the input sampled data file that is processed.
The samples are assumed to be numbered starting with one so that
setting \fIstart = 1\fR will cause processing to begin with the first sample.
.TP
.I nan - integer
The number of points in the sampled data file to process.  A value 0 indicates
to process the entire file.
.TP
.I filter_file_name - string
The FEA_FILT file name containing the filter data.  
.PP
If \fIfilter_file_name\fR file exists, the rest of the parameters are ignored.
.TP
.I filter_string_psiz - integer
The number of poles in the transfer function for the filter without their
complex conjugate counterparts.
\fIfilter_string\fR is a string of characters specified by the \fB-f\fR 
option.
.TP
.I filter_string_zsiz - integer
The number of zeros in the transfer function for the filter without their
complex conjugate counterparts.
\fIfilter_string\fR is a string of characters specified by the \fB-f\fR 
option.
.TP
.I filter_string_poles - float array
An array of size \fI2*filter_string_psiz\fR representing 
\fIfilter_string_psiz\fR poles without their complex conjugate counterparts.
Each pole is represented by 2 numbers, the first and second numbers
represent the real and imaginary parts of the pole, respectively.  For
example, a system with four poles \fI[0.9, 0.1i], [0.9, -0.1i], [0.8, 0.2i],\fR
\fI[0.8, -0.2i]\fR, can be stored in the parameter file by setting
\fIfilter_string_poles = {0.9, 0.1, 0.8, 0.2}\fR. \fIfilter_string\fR is a string of characters specified by the \fB-f\fR option.
.TP
.I filter_string_zeros - float array
An array of size \fI2*filter_string_zsiz\fR representing 
\fIfilter_string_zsiz\fR poles without their complex conjugate counterparts.
Each pole is represented by 2 numbers, the first and second numbers
represent the real and imaginary parts of the pole, respectively.  \fIfilter_string\fR is a string of characters specified by the \fB-f\fR option.
.TP
.I filter_string_gain - float
The overall gain for the pole-zero system of the IIR filter implemented by
cascading its second order IIR sub-systems.  This variable is ignored in
the FIR system.    If \fIfilter_string_gain\fR is not set, and if
the numerator coefficient array is specified, its value is set to equal 
to the first element of the numerator coefficient array.  Otherwise, the
default value is 1.
\fIfilter_string\fR is a string of characters specified by the 
\fB-f\fR option.  
.PP
In the IIR system, the pole/zero parameter data is sufficient for the
filtering operation. If the above pole/zero and gain parameters exist in 
the parameter file, the following parameters are ignored.  
Otherwise, the following numerator
and denominator coefficient parameters must be supplied.
.TP
.I filter_string_dsiz - integer
The number of denominator coefficients in the transfer function for the filter.
\fIfilter_string\fR is a string of characters specified by the \fB-f\fR 
option.
.TP
.I filter_string_den - float array
The denominator coefficients.  They are specified in order starting with
b0.  \fIfilter_string\fR is a string of characters specified by the \fB-f\fR option.
.PP
In the FIR system, the following parameters are necessary to define the
filtering operation.
.TP
.I filter_string_nsiz - integer
The number of numerator coefficients in the transfer function for the filter.
\fIfilter_string\fR is a string of characters specified by the \fB-f\fR 
option.
.TP
.I filter_string_num - float array
The numerator coefficients.  They are specified in order starting with
a0.  \fIfilter_string\fR is a string of characters specified by the \fB-f\fR option.
.PP
.SH ESPS COMMON
.PP
If the \fB-r\fR option is not used and ESPS Common processing is enabled,
the following items are read from the ESPS Common File provided that no
input file is given on the command line and provided that standard input
is not used.
.IP
.I "filename - string"
.IP
This is the name of the input file.
.IP
.I "start - integer"
.IP
This is the starting point in the input file for filtering.
.sp
.I "nan - integer"
.IP
This is the number of points in the input file for filtering.
A value of zero means the last point in the file.
.sp
.PP
The following items are written into the ESPS Common file, provided
ESPS Common processing is enabled and the output file is not <stdout>.
.IP
.I "start - integer"
.IP
A value of 1 is written to the Common file.
.sp
.I "nan - integer"
.IP
The number of points in the output file.
.sp
.I "prog - string"
.IP
This is the name of the program (\fIfilter2\fR in this case).
.sp
.SH ESPS HEADER
The file header of \fIoutput.sd\fR will contain mostly the same 
information as is contained in that of \fIinput.sd\fR,
.PP
The \fIstart_time\fR generic header item contains the starting time
in seconds of the first point in the output file.  \fIstart_time\fR maybe
changed by the \fB-z\fR option.  This helps time align a filtered signal
with the input signal since the filter is causal.
.PP
.SH EXAMPLES
Filters input sampled data file with a filter file \fIBworth.filt\fR.

	\fIfilter2 -F Bworth.filt input.sd output.sd\fR

.PP
.SH REFERENCES
Leland B. Jackson, \fIDigital Filters and Signal Processing\fP,
Kluwer Academic Publishers, 1986
.PP
.SH SEE ALSO
.PP
.nf
\fIblock_filter2\fP(3\-ESPS), \fInotch_filt\fP(1\-ESPS), \fIFEA_FILT\fP(5\-ESPS), 
\fIatofilt\fP(1\-ESPS), \fIwmse_filt\fP(1\-ESPS), \fIiir_filt\fP(1\-ESPS), 
\fIsfconvert\fP(1\-ESPS)
.fi
.PP
.SH AUTHOR
.PP
Program and man page by Derek Lin