NAME
linespec - Maximum entropy and Welch FFT spectral trace
analysis.
SYNOPSIS
linespec [ -Nntap ] [ -Ootap ] [ -S ] [ -Ggraphfile ] [
-rsirs ] [ -reire ] [ -tsits ] [ -teite ] [ -tiiti ] [
-Mnumc ] [ -Fnfreq ] [ -fft ] [ -rootipows ] [ -log ] [
-mmseg ] [ -gflmpfl ] [ -gfhmpfh ] [ -nop ] [ -V ] [ '-?' ]
DESCRIPTION
linespec is useful for identifying and characterizing line
spectral noise in seismic records or stacks.
Intended Uses of linespec:
Program linespec is meant to be used by those attempting the
detection and characterization of line spectral noise events
and of spectral variance in data sets. See also program see-
lines which is useful for parameter selection when using
linespec.
ADVISEMENT: In order to prevent certain technical problems,
it is recommended that data be scaled so that the maximum
absolute floating point value is somewhere between 1.0 and
1000.0 . Data that is scaled for plotting is USUALLY scaled
sufficently. The usual symptoms of problems due to this
cause are floating point error messages and/or unplotable
graph files due to NAN answers produced. We advise the user
to investigate MEM methods as MEM is not the same 'tame'
beast that we encounter when using Fourier methods. What you
do not know may definately hurt you, in this case, and this
is the price that is extracted for the potential benefits of
a more powerful tool. The Welch method, on the other hand,
is relatively safe. But there is a price to pay in that the
user gets only (typically) a coarse grained analysis (few
frequencies) with only slightly better statistical power
compared to regular Fourier methods. The defaults included
in this program were selected after considerable experimen-
tation and practical experience with (mostly) land data hav-
ing 750 to 3000 samples. You may very well save a signifi-
cant amount of time by using the defaults unless you have
strong counter indications. Some of the parameters may seem
like overkill, but are a response to the fact that line
spectral noise events often have a very narrow peak width
and their true relationship to signal (which usually has
some notches but few spikes) is not realistically determined
by coarser sampling.
Command line arguments
-N ntap
Enter the input data set name or file immediately after
typing -N, unless the input is from a pipe, in which
case the -N entry must be omitted. This input file
should include the complete path name if the file
resides in a different directory.
-O otap
Enter the output data set name or file immediately
after typing -O. This output file is not required when
piping the output to another process. The output data
set also requires the full path name (see above).
Default is linespec.out. See possible pipe issues
below.
-S Enter the command line argument -S if an outout of the
spectra in trace form is to be generated.
-G graphfile
Enter the graphical output data set name immediately
after typing -G. Default is graph.out . Use full path
name if necessary. This is an ASCII suitable for plot
with program XGRAPH.
-rs irs
Enter the first record number to include in the
analysis. Default is first record of the data set.
-re ire
Enter the last record number to include in the
analysis. Default is last record of the data set.
-ts its
Enter the first trace number to include in the
analysis. Default is first trace of each record.
-te ite
Enter the last trace number to include in the analysis.
Default is last trace of each record.
-ti iti
Enter the trace increment between traces included in
the analysis. Default is 1.
-M numc
Enter the order of the MEM approximation desired.
About one sixth of this number of line events can be
resolved easily. Too low a number decreases resolution
and too high a number causes confusion as the method
becomes overly sensitive to system and numerical noise.
Comparisons between analyses are difficult (impossi-
ble?) if this number is changed from analysis to
analysis. Default is 64.
-F nfreq
Enter the number of frequencies between 0 and Nyquist
(inclusive) to be output (MEM). Large numbers produce
VERY BIG output graph files and can make run times
longer than you may like (due to I/O). Unlike Fourier
methods, you can find spectral values for any frequency
from zero to Nyquist, and therefore as many frequency
values as you wish (limited only by storage and execu-
tion time constraints). Default is 1001.
-fft Enter the command line argument -fft to get additional
graphical output file containing a Welch FFT spectral
analysis.
-root ipows
If ipows is not zero, the ipow th root of the amplitude
spectra are calculated and used for output. Must be
integer value. Default is 0 (no root taken).
-log Use this flag if logarithm of amplitude is to be taken
before amplitude specta are output. Formula used is:
a(out) = log10( a + 0.1 )
The constant is to prevent trying to take log10(0.0).
For any
two frequency values, the difference in Decibells is
20.0 times
the difference produced in the output of the program.
-m mseg
Half the number of samples in a Welch FFT window.
Should be an integer power of 2. This is also the
number of frequencies generated (between 0 and Nyquist
Hz) in the analysis output. Default is 128.
-gfl mpfl
Enter the lowest frequency (in Hz) to be output from
analysis. Default is 0.
-gfh mpfh
Enter the highest frequency (in Hz) to be output from
analysis. Default is Nyquist frequency.
-nop Use this flag to supress output of the ASCII MEM graph
file (useful when output in trace form only is
desired).
-V Enter the command line argument -V to get additional
printout.
'-?' Enter the command line argument '-?' to get on line
help. Note the single quotes must be typed by the user
to prevent the shell from interpreting the -?. The
program terminates after the help screen is printed.
BUGS
AUTHOR
Dennis Bjerstedt, Calgary, 1992.
COPYRIGHT
copyright 2001, Amoco Production Company
All Rights Reserved
an affiliate of BP America Inc.
Man(1) output converted with
man2html