make_sd.diff

来自「speech signal process tools」· DIFF 代码 · 共 126 行

.\" Copyright (c) 1987,1990 Entropic Speech, Inc.; All rights reserved
.\" @(#)make_sd.1	1.4 3/20/90 ESI
.TH MAKE_SD 1\-ESPS 3/20/90
.ds ]W "\fI\s+4\ze\h'0.05'e\s-4\v'-0.4m'\fP\(*p\v'0.4m'\ Entropic Speech, Inc.
.SH NAME
make_sd \- convert FEA fields to FEA_SD sampled data file
.SH SYNOPSIS
.B make_sd
.BI \-f " field"
[
.BI \-r " range"
] [
.BI \-S " sampling_rate"
] [
.BI \-x " debug_level"
]
.I file.fea
.I file.feasd
.SH DESCRIPTION
.PP
.I Make_sd
makes a FEA_SD sampled data file \fI(file.feasd)\fR of the data in the
named \fIfield\fR in the input file
.I file.fea.  
The data type of the FEA_SD file is that of the field \fIfield\fP (but
complex data are not handled).  If
.I file.fea
is "\-", the standard input is read, and if \fIfile.feasd\fR is "\-",
the standard output is written.
.PP
By default, all elements of the specified field in the first record
are used in making the sampled data file.  If the \fB\-r\fR option is
used, the field elements in all the specified records are catenated
together into one sampled data file.  If only one element from each 
field is to form the sampled data stream, use \fItofeasd\fP
(1\-\s-1ESPS\s+1) instead.  
.PP
The fieldname must be given with the
.B \-f
flag.  This is not optional and there is no default.
.SH OPTIONS
.PP
The following options are supported:
.TP
.BI \-r " record"
.TP
.BI \-r " first:last"
.TP
.BI \-r " first:+incr
Determines the records from which values from
the named \fIfield\fP are extracted. In
the first form, a single integer identifies one record from which 
values of 
.I field 
are extracted.
In the second form, a pair of unsigned integers gives the first and last
points of the range.  If
.I first
is omitted, 1 is used.  If 
.I last 
is omitted, the last record in 
.I file.fea
is assumed.  The third form is equivalent to the first with
.I "last = first + incr."
When more than one record is processed, the element values are catenated
together before being put in the sampled data file.
.TP
.BI \-S " sampling_rate \fR[1\fP]"
Specifies the sampling rate of the data \- this value will be stored
in the \fIrecord_freq\fP generic of the output file.  
.TP
.BI \-x " debug_level"
If
.I debug_level
is nonzero, debugging information is written to the standard error output.
Default is 0 (no debugging output).
.SH EXAMPLES
.PP
\fImake_sd\fR \fB\-f\fR spec_param \fB\-r5\fR speech.fana spec_param.sd
.PP
puts the \fIspec_param\fR field from record 5 of speech.fana 
into the sampled data file \fIspec_param.sd\fR.
.PP
An example of how \fImake_sd\fR
is useful is in computing the cross spectrum from the output
of \fIcross_cor\fR.
To compute 
the cross spectrum corresponding to record 5,
the following command will work.
.PP
\fImake_sd\fR \fB\-f\fR cross_cor \fB\-r5\fR speech.fana \- \fB|\fR fft -o10 \- spec_param.spec
.SH "ESPS PARAMETERS"
No ESPS parameter file is read.
.SH "ESPS COMMON"
ESPS Common processing is disabled.  
.SH DIAGNOSTICS
.PP
.I Make_sd
complains and exits if 
.I file.fea
does not exist or is not a FEA file.
.SH "IMPLEMENTATION NOTE"
.PP
The current version of \fImake_sd\fP is implemented as a shell script
that pipes the output of \fIpplain\fP (1\-\s-1ESPS\s+1) through
\fItestsd\fP (1\-\s-1ESPS\s+1). This is an example of how ESPS
programs can be combined with the "Unix tools approach" \- see the
simplified example on the \fIpplain\fP (1\-\s-1ESPS\s+1) man page.
.SH BUGS
.PP
None known. 
.SH EXPECTED CHANGES
.PP
.PP
A later version will integrate \fItofeasd\fP (1\-\s-1ESPS\s+1) and 
\fImake_sd\fP.  
.SH SEE ALSO
\fItofeasd\fP (1\-\s-1ESPS\s+1), fea_element(1\-ESPS),
pplain(1\-ESPS), testsd(1\-ESPS), auto(1\-ESPS), cross_cor(1\-ESPS)
.SH AUTHOR
.PP
.nf
Manual page by David Burton
Program by Alan Parker