NAME
spec - module to perform spectral decomposition of input
time data into descrete amplitude or phase spectra by either
a descrete Fourier transform or Maximum Entropy Method.
SYNOPSIS
spec [ -Nntap ] [ -Ootap ] [ -Sstap ] [ -Fftap ] [ -nsins ]
[ -neine ] [ -rsirs ] [ -reire ] [ -siwst ] [ -eiwed ] [
-fminmin freq ] [ -fmaxmax freq ] [ -fintfreq interval ] [
-opbias ] [ -trbias ] [ -taper% cosine taper ] [ -G ] [
-minmin polynomial order ] [ -maxmax polynomial order ] [
-phase ] [ -W ] [ -MEM ] [ -ssam ] [ -V ] [ -? ]
DESCRIPTION
spec computes amplitude or phase spectra for the input data
as per the input parameters. The conventional Fourier
transform is used (rather than the Fast Fourier Transform)
so that the frequency sample increment can be arbitrarily
specified. The output is either amplitude or phase. The pro-
gram will also produce an amplitude spectrum via Maximum
Entropy Method. Note that MEM does not produce a phase
response. Output data consists of amplitude or phase spec-
tra in USP format. The user can extract the values and plot
with USP program sis_xy and xgraph. The sample interval is
in the output data line header and can be read as Hertz. For
MEM the last sample of the time series is the optimum order
for the polynomial 'best' fitting the spectra.
spec gets both its data and its parameters from command line
arguments. These arguments specify the input, output, the
scaling information output file, the optional 1st spectral
maxima output file, the start and end records and traces,
start and end times for windows, frequency end points, fre-
quency increment ,window tapering, phase unwrapping,
minimum order of the polynomial and maximum order of a poly-
nomial when using Maximum Entropy Method, trace bias removal
and operator bias removal options and verbose printout, if
desired.
Command line arguments
-N ntap
Enter the input data set name or file immediately after
typing -N. If piping in this command line entry will
be omitted. The input file should include the complete
path name if the file resides in a different directory.
Example -N/b/vsp/dummy tells the program to look for
file 'dummy' in directory /b/vsp
-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).
-S stap
Enter the scaling output data set name or file immedi-
ately after typing -S. This output is used as input to
the USP routine scspec to remove the wavelet overprint
from the tuning cube. This is mainly used when import-
ing the cube to Landmark or IESX and want the data
scaled in this fashion. If displaying the output in
xsd use a maximum scalar of 1.0 on your frequency
slices and you won't need this applied using scspec.
Be aware that the maximum spectral ampltiude at a given
frequency can easily be coming from a noisy area of
your tuning cube if appropriate triming of the dataset
has not been done beforehand to limit edge effects of
previous processing.
-F ftap
Enter the data set name for the statistics of the first
spectral maxima file. The first sample contains the
frequency at which the first spectral maxima occurs.
The second sample contains the amplitude at the first
spectral maxima. This information is critical to
determining the maximum frequency useful for thickness
estimations locally.
-rs irs
Enter the start record number. Default value is the
first record.
-re ire
Enter the end record number. Default value is last
record.
-ns ns
Enter start trace number. Default is first trace of
record.
-ne ne
Enter end trace number. Default value is last trace of
record.
-s iwst
Enter start time of window for harmonic analysis.
Default is the start of the trace.
-e iwed
Enter end of window for harmonic analysis. Default is
end of trace.
-taper taper
Enter percent cosine taper to apply to window. 100% is
full cosine taper and 0% is no taper. Default is 10%.
-G Enter the command line argument '-G' to apply a Gaus-
sian taper. This option overrides the cosine taper and
needs no percent value input.
-fmin min freq
Enter beginning frequency for harmonic analysis.
Default is 0 Hz.
-fmax max freq
Enter ending frequency for harmonic analysis. Default
is Nyquist frequency.
-fint freq interval
Enter frequency increment in Hz. Default is 1 hz.
-min min polynomial order
Enter minimum allowable order of the polynomial modeled
to fit the ME amplitude spectrum. if -max and -min are
equal then only that value is used for all traces.
Default is 2.
-max max polynomial order
Enter maximum allowable order of the polynomial modeled
to fit the ME amplitude spectrum. if -max and -min are
equal then only that value is used for all traces.
Default is 2.
-phase
Enter the command line argument '-phase' to output
phase. The analysis window center is the phase refer-
ence.
-W Enter the command line argument '-W' to leave phase
wrapped. The default is to unwrap the phase.
-MEM Enter the command line argument '-MEM' to enable the
Maximum Entropy Method. The last sample in the output
is the order polynomial used to calculate the spectra
for that trace.
-trbias
Enter the command line argument '-trbias' to remove
the trace bias prior to decomposition. This used to be
the default in this code as written by the original
authors. I am not sure what they were thinking as
assumptions of zero mean are really not valid for short
time series which are often the case with this applica-
tion. I believe the notion for this option came from
thinking about the FFT of a stationary time series
where zero mean is an assumption that makes sense. It
is a dangerous option to invoke here however, espe-
cially if you are after thickness estimates in the
presence of thinning units. I am leaving this capabil-
ity as an option in the program purely for historical
reasons. Also a user may want to recreate an old
result and may need this turned on for that purpose.
Please be very careful with your analysis and give this
a lot of theoretical thought if you choose to use this
option [Garossino].
-opbias
Enter the command line argument '-opbias' to remove the
operator bias during the decomposition calculation.
This, along with the trace bias removal above, used to
be the default in this routine [prior to June 28,
1999]. After much thought and consultation with
Gutowski and Vassiliou I can find no theoretical reason
for this operation. It can do nothing but harm to your
output thickness estimates. I would NOT use this
option and recommend AGAINST using it. I am leaving
the capability as an option merely for historical com-
pleteness. A user may need to turn this on to recreate
an old result. If you choose to try this option you
may well be faced with explaining what you think your
are doing.
-ssam
Enter the command line argument '-ssam' to get a prin-
tout to stderr at the start of every input record.
This allows the SSAM user to monitor how the applica-
tion is running.
-V Enter the command line argument '-V' to get additional
printout.
-? Enter the command line argument '-?', -h or -help to
get online help. The program terminates after the help
screen is printed.
DISCUSSION:
For an excellent primer on this process point your web
browser to
http://eptgweb/~usp/TechTransfer/OnlineTraining/Spec_Tutorial/SpecDoc.html
BUGS
No bugs known at present.
SEE ALSO
timfreq, freqtim, samps, fk, specal, ampspec, fft2da, fftxy,
rwspec, mesa
AUTHOR(S):
[James M. Gridley], with much help from a variety of folks
and in cooperation with Greg Partyka. Also, Maximum Entropy
Method was developed by Paul Gutowski. [Paul G.A. Garossino
- June/99], Complete rewrite to move table calculations out
of the trace loop and make output compatible with rwspec.
Also converted trace bias and operator bias removal to
options rather than the default.
COPYRIGHT
copyright 2001, Amoco Production Company
All Rights Reserved
an affiliate of BP America Inc.
Man(1) output converted with
man2html