Usage:
The Deghosting steps attempts to attenuate the ghost in pre-stack data by local prediction error filtering. The input data is partitioned temporally and spatially into overlapping blocks of data, and each block is minimised independently. A range of prediction filters are calculated using an averaged autocorrelation for each record, using the specified data time range. A reverse operator is optionally calculated and added to the forward operator before the filtering to approximate zero phase filtering. The prediction filters with increasing gap lengths are applied to the data within each block. The o/p is minimised in the L1 sense.
Input Links:
1) Seismic data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Reference:
-
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Ghost delay range: Specifies the range to determine the extent of the predictive gap lengths to include for attenuating the ghost.
Minimum ghost delay (ms) — Enter the smallest ghost delay time to include.
Minimum ghost delay (ms) — Enter the largest ghost delay time to include.
Allow unfiltered data in output — If checked, a possible output from the minimization within a block is the input data. [In data blocks where there is little evidence of the ghost this is a reasonable outcome].
Filter Design: Specifies the parameters from which the predictive gap operators will be calculated are specified here.
Design window start (ms) — Enter the initial data time to be used to extract data used to calculate the predictive gap operators from the data autocorrelation.
Design window length (ms) — Enter the data window length to be used as above.
Filter length (ms) — Enter the lengths of the filters to calculate in milliseconds (not including the predictive gaps).
Apply moveout – If checked, the moveout velocity may be overridden in the adjacent box. A moveout correction corresponding to the trace offset will then be applied before windowing the data according to the parameters above.
Percentage white noise – Specify amount of white noise (as a % of the autocorrelation zero lag ) to add to stabilize the solution of the Toeplitz matrix in estimation of the filter.
Overlapping deghosting windows: Specify the overlap of blocks is set to 50%, using Hanning tapering between the blocks.
Window length (ms) — Enter the length in milliseconds of each deghosting window.
Window width (traces) — Enter the width in traces of each deghosting window.
Reverse Operator: Specifies if the operation adds in a time-reversed and scaled copy of the operator to act like a zero-phase filter.
Apply reverse Operator — If checked, a reverse operator will be added.
Ghost amplitude — Enter the scaling of the reverse operator to approximate the relative amplitude of the ghost.
Usage:
The FKK Domain Spectral Shaping step is a spectral shaping filter primarily for application to post-stack and 3D migrated seismic data. The operator is applied in the 3D Fourier transformed FKK domain.
The process applies an integration filter to the data with an optional 90 degrees phase shift rotation to simulate acoustic impedance data. Applying such a process to migrated data can cause severe amplitude distortions, excessively amplifying dipping events, signal or noise. Therefore, in replace mode, this process simulates de-migration of the data, application of the integration filter and then re-migration of the data. The result is an FKK filter that is thought to balance the spectrum and preserve low frequencies so as to be superior to many other FKK filters.
The input data must be in the form of frequency slices – convert the data to frequency slices by the use of Frequency Slices in a prior job flow.
After filtering, convert the frequency slices back to standard time traces in a subsequent job flow with the use of Frequency Slices to Time Traces.
Input Links:
1) Frequency slices sorted by Slice Number and CMP Line (mandatory).
Output Links:
1) Frequency slices sorted by Slice Number and CMP Line (mandatory).
Reference:
Lazaratos, S, and David R., 2009, Inversion By Pre-Migration Spectral Shaping, SEG Annual Meeting, 2009.
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Integrating filter application mode: Specify filter application mode.
Replace — If checked, simulates de-migration of the data, application of the integration filter and then re-migration of the data
Apply — If checked, apply the FKK spectral shaping to the data.
Apply pre-migration —If checked, applies the FKK spectral shaping before migration.
Remove — If checked, removes the FKK filter to the data
Migration velocity — Enter constant velocity to be used in the migration step.
Start frequency — Enter the start frequency to be analyzed.
Apply 90 deg. Phase shift — If checked, applies a 90 degrees phase shift to the input data to simulate acoustic impedance.
Override trace spacing: Specify if ones wishes to override trace space as defined in input seismic file.
Specify trace spacing — If checked, specify trace spacing entering the corresponding value.
Specigy line spacing (3D) — If checked, specify line spacing entering the corresponding value. This option is only valid for three-dimensional dataset.
Usage:
The Predictive Deconvolution step is a Wiener-Levinson algorithm for applying a predictive multi-gate deconvolution to your data. You choose the percent pre-whitening, filter length, number of operators, the overlap of the operator design windows, start time of each operator design window, and the design window lengths. For the predictive deconvolution method, you must specify the predictive length of your wavelet. You may also apply a linear moveout to your deconvolution design windows to allow a sliding window whose start time varies with offset.
Input Links:
1) Seismic data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Reference:
-
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Number of operators — Enter the number of deconvolutin operators to be used.
Overlap between operators (ms) — Enter the overlap in milliseconds between operators. This is the amount of the previous and/or next window to include in the calculation of the inverse filter for the current window.
Apply reverse operator — If checked, applies the reverse deconvolution operator to the data.
Ghost amplitude — Enter the amplitude value for the deconvolved ghost.
Operator — Sequential number of the operator to be used.
Prewhiten % — Enter the amount of white noise to add. The zero lag of the autocorrelation function is increased by this amount to induce stability in the matrix solution.
Operator length (ms) — Enter the length of the operator to be calculated and applied in milliseconds.
Moveout — If checked, a linear moveout will be applied to the deconvolution design window. The window start time will shift by: delta time = offset / velocity.
Velocity — Enter the velocity to be applied in the linear moveout.
Window start (ms) — Enter the start time in milliseconds of the deconvolution design window.
Window length (ms) — Enter the length in milliseconds of the deconvolution design window.
Usage:
The Predicitve Multichannel Deconvolution step is a Wiener-Levinson algorithm for applying a multi-channel predictive multi-gate deconvolution to your data. You choose the percent pre-whitening, filter length, number of operators, the overlap of the operator design windows, start time of each operator design window, and the design window lengths. For the predictive deconvolution method, you must specify the predictive length of your wavelet. You may also apply a linear moveout to your deconvolution design windows to allow a sliding window whose start time varies with offset.
Input Links:
1) Seismic data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Reference:
-
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Number of operators — Enter the number of deconvolutin operators to be used.
Overlap between operators (ms) — Enter the overlap in milliseconds between operators. This is the amount of the previous and/or next window to include in the calculation of the inverse filter for the current window.
Apply reverse operator — If checked, applies the reverse deconvolution operator to the data.
Ghost amplitude — Enter the amplitude value for the deconvolved ghost.
Range limit traces used in operator design — Limit the number of input traces used in the operator design based on a trace header key.
Range limit header field — Select the trace header key to be used in the trace limit.
Minimum range value — Enter the minimum range value.
Maximum range value — Enter the maximum range value.
Absolute value — If checked, consider the absolute value of the selected trace header key.
Operator — Sequential number of the operator to be used.
Prewhiten % — Enter the amount of white noise to add. The zero lag of the autocorrelation function is increased by this amount to induce stability in the matrix solution.
Operator length (ms) — Enter the length of the operator to be calculated and applied in milliseconds.
Moveout — If checked, a linear moveout will be applied to the deconvolution design window. The window start time will shift by: delta time = offset / velocity.
Velocity — Enter the velocity to be applied in the linear moveout.
Window start (ms) — Enter the start time in milliseconds of the deconvolution design window.
Window length (ms) — Enter the length in milliseconds of the deconvolution design window.
Usage:
The Q Filter step applies a frequency dependent “Q” attenuation compensation filter or an attenuation modeling filter to the input file. For both attenuation and modeling options, phase only or phase and amplitude filters may be applied.
Input Links:
1) Seismic data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Reference:
Hargreaves, N. D., Calvert A. J., 1991, Inverse Q filtering by Fourier Transform, Geophysics, 56, p. 519.
Hargreaves, N., 1992, Similarity and the inverse Q filter: Some simple algorithms for inverse Q filtering, Geophysics, 57, p. 994.
Wang, Y., 2002, A stable and efficient approach of inverse Q filtering, Geophysics, 67, p. 657.
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Operator type: Specify the type of operator.
Inverse Q - compensation — If checked, attenuation will be removed from the input seismic traces.
Forward Q - modelling — If checked, the input seismic traces will have attenuation is to be introduced to.
Compensation type: Specify the compensation type.
Phase only — If selected, an allpass, phase-only filter is to be applied for forward or inverse Q filtering. This is a computationally efficient option.
Phase and amplitude — If selected, a phase and amplitude type filter is to be applied for forward or inverse Q filtering. This option is computer intensive.
Constant Q — Enter the value of Q for forward or inverse Q filtering.
Reference frequency (Hz) — Enter the reference frequency for forward or inverse Q filtering. There will be slight frequency dependent phase shifts with respect to this frequency.
Gain limit (dB) — Enter the maximum gain limit per frequency in the case of inverse Q filtering.
Usage:
The Signature Deconvolution step removes a seismic signature from your data traces by calculating the inverse of a supplied signature trace and then filtering your data by this signature inverse. This is a common technique for use in processing marine data to remove the signature of the airguns.
Input Links:
1) Seismic data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Deconvolution paramters: Specify generic parameters of the deconvolution operator.
Pre-whitening percent — Enter the prewhitening multiplier. The zero lag of the autocorrelation function is increased by this amount to induce stability in the matrix solution.
Inverse filter length (ms) — Enter the length of the filter to be calculated and applied in milliseconds.
Input signature start time (ms) — Enter the start time of the input signatures in milliseconds.
Input signature length (ms) — Enter the length of the input signatures in milliseconds.
Time shift for output (ms) – Enter the time shift for the output trace following signature deconvolution
Signature Input Options: Specify the data input source of the signature trace or traces.
One signature from an auxiliary file — This option inputs one signature trace from an auxiliary input data and uses that trace as the signature trace for the entire data set.
One signature per record from an auxiliary data file — This option inputs one signature trace per record from an auxiliary input data set and uses that trace as the signature trace for all traces in the record in the signature deconvolution calculation.
One signature per trace from an auxiliary data file — This option inputs one signature trace from an auxiliary input data set per data trace and uses that trace as the signature trace for the corresponding data trace in the signature deconvolution calculation.
One signature per record in the data file — This option uses one trace from each data record and uses that trace as the signature trace for all seismic traces in the record in the signature deconvolution calculation.
Signature trace number — Enter the trace number to use as the signature trace.
Sweep trace number — Enter the number of traces in the sweep.
First trace to kill — If you demultiplexed the data set with the auxiliary traces to recover the signature trace, you may wish to kill the these auxiliary traces. Enter the first trace number to kill.
Last trace to kill — If you demultiplexed the data set with the auxiliary traces to recover the signature trace, you may wish to kill the these auxiliary traces. Enter the last trace number to kill.
Usage:
The Spectral Whitening step is a multi-banded spectral whitening method for balancing the energy of the selected frequency bands of your data. You specify the beginning and ending frequencies in the pass zone, the number of bands within this zone, and the length of the AGC operator. The individual band-pass filter cut off values are linearly interpolated based on the low and high pass frequencies and the number of bands that you choose. Each band is given a Butterworth third order slope with the midpoints of each slope being the half power points. Within each band every trace is AGC’ed with respect to the maximum amplitude in the window. All AGC’ed bands are then summed. Optionally, you may specify the Low Cut / Low Pass / High Pass / High Cut (LC,LP,HP,HC) filter points and associated weights for summing of up to ten filter bands. For optimal balancing of the energy of your frequency bands, your filters should be designed such that the slopes of your filters exactly overlap as is illustrated below. Such a design optimally flattens the spectrum of your data without distorting the spectrum.
Input Links:
1) Seismic data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Number of bands — Enter the number of frequency bands. Each trace is split into this number of frequency bands which are AGC'ed and then summed to create the output trace.
Start frequency (Hz) — Enter the low frequency pass band limit in hertz (Hz). This is the lowest frequency in the pass band.
End frequency (Hz) — Enter the high frequency pass band limit in hertz (Hz). This is the highest frequency in the pass band. It may not exceed one-half the Nyquist frequency.
AGC length — Enter the AGC operator length. This is the length of the AGC operator, which is applied to each frequency band.
Operator — Sequential number of the operator to be used.
Low Cut — Enter the filter low frequency cutoff point in hertz (Hz).
Low Pass — Enter the filter low frequency pass point in hertz (Hz).
High Pass — Enter the filter high frequency pass point in hertz (Hz).
High Cut — Enter the filter high frequency cutoff point in hertz (Hz).
Weights — Enter the filter band relative weight.
Usage:
The Spiking Deconvolution step is a Wiener-Levinson algorithm for applying a spiking multi-gate deconvolution to your data. You choose the percent pre-whitening, filter length, number of operators, the overlap of the operator design windows, start time of each operator design window, and the design window lengths. For the predictive deconvolution method, you must specify the predictive length of your wavelet. You may also apply a linear moveout to your deconvolution design windows to allow a sliding window whose start time varies with offset.
Input Links:
1) Seismic data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Reference:
-
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Number of operators — Enter the number of deconvolutin operators to be used.
Overlap between operators (ms) — Enter the overlap in milliseconds between operators. This is the amount of the previous and/or next window to include in the calculation of the inverse filter for the current window.
Apply reverse operator — If checked, applies the reverse deconvolution operator to the data.
Ghost amplitude — Enter the amplitude value for the deconvolved ghost.
Operator — Sequential number of the operator to be used.
Prewhiten % — Enter the amount of white noise to add. The zero lag of the autocorrelation function is increased by this amount to induce stability in the matrix solution.
Operator length (ms) — Enter the length of the operator to be calculated and applied in milliseconds.
Moveout — If checked, a linear moveout will be applied to the deconvolution design window. The window start time will shift by: delta time = offset / velocity.
Velocity — Enter the velocity to be applied in the linear moveout.
Window start (ms) — Enter the start time in milliseconds of the deconvolution design window.
Window length (ms) — Enter the length in milliseconds of the deconvolution design window.
Usage:
The Spiking Multichannel Deconvolution step is a Wiener-Levinson algorithm for applying a multi-channel spiking multi-gate deconvolution to your data. You choose the percent pre-whitening, filter length, number of operators, the overlap of the operator design windows, start time of each operator design window, and the design window lengths. For the predictive deconvolution method, you must specify the predictive length of your wavelet. You may also apply a linear moveout to your deconvolution design windows to allow a sliding window whose start time varies with offset.
Input Links:
1) Seismic data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Reference:
-
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Number of operators — Enter the number of deconvolutin operators to be used.
Overlap between operators (ms) — Enter the overlap in milliseconds between operators. This is the amount of the previous and/or next window to include in the calculation of the inverse filter for the current window.
Apply reverse operator — If checked, applies the reverse deconvolution operator to the data.
Ghost amplitude — Enter the amplitude value for the deconvolved ghost.
Range limit traces used in operator design — Limit the number of input traces used in the operator design based on a trace header key.
Range limit header field — Select the trace header key to be used in the trace limit.
Minimum range value — Enter the minimum range value.
Maximum range value — Enter the maximum range value.
Absolute value — If checked, consider the absolute value of the selected trace header key.
Operator — Sequential number of the operator to be used.
Prewhiten % — Enter the amount of white noise to add. The zero lag of the autocorrelation function is increased by this amount to induce stability in the matrix solution.
Operator length (ms) — Enter the length of the operator to be calculated and applied in milliseconds.
Moveout — If checked, a linear moveout will be applied to the deconvolution design window. The window start time will shift by: delta time = offset / velocity.
Velocity — Enter the velocity to be applied in the linear moveout.
Window start (ms) — Enter the start time in milliseconds of the deconvolution design window.
Window length (ms) — Enter the length in milliseconds of the deconvolution design window.
Usage:
The Surface Consistent Decomposition step uses a Gauss-Seidel iterative method to perform a four-component surface-consistent decomposition of the amplitude spectra of the input seismic data into source, receiver, offset, and CMP components. Output component spectra can then be used as input to Surface Consistent Deconvolution step.
Input Links:
1) Seismic data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Reference:
-
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Design window: Specify window where surface-consitent decomposition will be performed.
Use fixed window to built pilot — If checked, use a fixed time window to build a pilot.
Design window start time (ms) — Enter start time of the pilot window in millieseconds.
Design window end time (ms) — Enter the end time of the pilot window in milliseconds.
Apply linear moveout to design window — If checked, design window is defined as an offset-dependent linear moveout.
Linear moveout velocity — Enter the linear moveout velocity.
Offset limit traces in design window — If checked, traces for the analysis are limited in terms of offset.
Minimum offset — Enter the minimum offset to be considered.
Maximum offset — Enter the maximum offset to be considered.
Surface consistent spectral decomposition: Specify consistent spectra decomposition parameters.
RMS error change for termination — Enter the minimum RMS error for the iterative procedure to be considered finished.
Midpoint smoothing radius (bin) — Enter the radius (in number of bins) of the midpoint smoothing operator.
Compute source spectra — Select yes to compute source spectra.
Compute receiver spectra — Select yes to compute receiver spectra.
Compute midpoint spectra — Select yes to compute midpoint spectra.
Compute offset spectra — Select yes to compute offset spectra.
Residual spectra output: Specify type of output.
Output source spectra — Select yes to output source spectra.
Output receiver spectra — Select yes to output receiver spectra.
Output midpoint spectra — Select yes to output midpoint spectra.
Output offset spectra — Select yes to output offset spectra.
Usage:
The Surface Consistent Deconvolution step uses a Gauss-Seidel iterative method to perform a four-component surface-consistent decomposition of the amplitude spectra of the input seismic data into source, receiver, offset, and CMP components. Deconvolution operators are constructed from the source and receiver components of this decomposition. The user specifies whether the deconvolution will be of the spiking or predictive type, the design window, operator length, and percent white noise.
Input Links:
1) data in any sort order (mandatory).
Output Links:
1) Seismic data in any sort order (mandatory).
Reference:
Cary, P. M., and Lorentz, G. A., 1993, Four-component surface-consistent deconvolution: Geophysics, 58, 383-392.
Example Flowchart:
Step Parameter Dialog:
Parameter Description:
Deconvolution parameters: Specifyvtype of deconvolution to perform: Spiking or Predictive.
Spiking — If selected, use a spiking deconvolution.
Predictive — If selected, use a predictive or gapped deconvolution.
Prediction length (ms) — Enter the prediction length in milliseconds.
Pre-whitening percent — Enter the pre-whitening multiplier. The zero lag of the autocorrelation function is increased by this amount to induce stability in the matrix solution.
Inverse filter length (ms) — Enter the length of the filter to be calculated and applied in milliseconds.
Deconvolution operator application: Specify type of deconvolution operators to be applied to the data.
Compute and apply source operators — Select yes, to compute and apply source deconvolution operators.
Compute and apply receiver operators — Select yes, to compute and apply receiver deconvolution operators.
Compute and apply midpoint operators — Select yes, to compute and apply midpoint deconvolution operators.
Compute and apply offset operators — Select yes, to compute and apply offset deconvolution operators.
Output type: Specify type of output.
Output deconvolved trace data — If selected, deconvolution operators are applied to the output data.
Output deconvolution operators — If selected, deconvolution operators are the output and can be used in other processing steps. No deconvolution is applied to input data.