|
<< Click to Display Table of Contents >> Navigation: MultiFocusing > Data Enhance - 3D ZO-MF(new) |
Data Enhance - 3D ZO-MF(new) enhances and interpolates 3D pre-stack seismic data using a Zero-Offset Multi-Focusing (ZO-MF) database. For each output bin, the module retrieves the ZO-MF wave-field parameters computed by the 3D ZO-MF engine, selects neighbouring input traces from the raw seismic data, and applies multi-focusing weighted summation (partial stacking) to produce a high signal-to-noise enhanced gather. This is the updated (new) version of the 3D ZO-MF enhancement module, offering additional Fresnel zone aperture control, improved fold management, and more flexible on-disk trace geometry handling compared to the previous version.
Use this module after running the 3D ZO-MF engine to convert the computed multi-focusing parameters into enhanced pre-stack gathers. The output is written to a GSD file and supports both direct-write and append modes, making it suitable for large surveys processed in chunks or on distributed computing nodes. The module supports GPU acceleration for maximum throughput on large 3D surveys.
The path to the output GSD file where the enhanced pre-stack gathers will be written. This file is created by the module. If the file already exists and contains previously computed chunks (from an interrupted run), the module will detect this and ask whether to continue from where it left off or overwrite the file.
The 3D ZO-MF database file (.kdb) produced by the 3D ZO-MF engine. This file contains the multi-focusing wave-field parameters — including focusing weights and time-quantization tables — computed for each bin in the survey. This input is mandatory; the module cannot run without a valid storage file.
When set to false (default), the input trace geometry is provided via the Input trace headers and SEG-Y data handle connectors. When set to true, the input geometry is read from a TVOD trace-vector file on disk, as specified by the Input traces data handle connector. Use the on-disk mode for very large surveys where the trace header index does not fit in memory.
When set to false (default), the output geometry is provided via the Output trace headers connector. When set to true, the output geometry is read from a TVOD file on disk, via the Output traces data handle connector. Enable this option together with the input on-disk mode for fully disk-based geometry handling on very large 3D surveys.
Connection to the raw input seismic data in SEG-Y format. This connector is active when Use input headers from disk is set to false. The module reads input traces from this data handle during processing — the number of samples in the SEG-Y data must not exceed the number of samples recorded in the ZO-MF storage file.
The trace header index (geometry) of the raw input seismic dataset, defining the bin, source, and receiver positions of each input trace. Active when Use input headers from disk is false. This index is used to locate input traces within the aperture of each output bin during enhancement.
A TVOD (Trace Vector On Disk) handle to the input trace geometry stored as a disk-based index file. Active when Use input headers from disk is true. Use this connector for large surveys where the full trace index does not fit in RAM.
The trace header index defining the geometry (bin positions and trace layout) of the desired output enhanced gathers. Active when Use output headers from disk is false. The output bins defined here determine which locations in the survey will be enhanced and written to the output file.
A TVOD handle to the output geometry stored on disk. Active when Use output headers from disk is true. Provides the output bin and trace layout from a disk-resident index file, enabling processing of surveys too large to hold the output geometry in memory.
An optional picking corridor item that constrains the multi-focusing summation to a time window around a picked horizon or event. When connected, the module constructs a picking corridor bounding box from the ZO-MF time-quantization table and uses it to restrict the wave-field summation to the region of interest, improving the enhancement quality near picked reflectors.
Controls how many traces are read from the TVOD file in a single I/O operation when the on-disk mode is active. Default: 200000 traces. Larger values improve I/O efficiency but require more memory. Reduce this value if the process runs out of memory during the initial geometry loading phase. This parameter is visible only when one of the on-disk header modes is enabled.
An optional 2D map (GMatrixItem) of the near-surface replacement velocity V0, sampled at each bin location. When connected, the multi-focusing interpolator uses spatially varying V0 values instead of a constant reference velocity, which improves the accuracy of the wave-field summation in areas with strong lateral velocity variations near the surface.
When enabled, the output gathers will be NMO-corrected (normal moveout applied) in addition to the multi-focusing enhancement. Default: false. Enable this option when the downstream processing step requires flattened gathers (for example, before AVO analysis or stacking without a separate NMO step).
When enabled, the module applies a multi-focusing emulation mute to the output traces. Default: false. This mute suppresses shallow early-arrival energy on the enhanced output gathers in the same way the ZO-MF engine would naturally mute it, ensuring consistency between the enhanced and original data mute patterns.
A group of parameters inherited from the ZO-MF imaging engine that govern how the multi-focusing wave-field summation selects and weights contributions from different azimuths and directions. Key sub-parameters include: Directions (number of summation azimuth sectors), From angle / To angle (azimuth range in degrees, default -90 to +90), Correlation threshold (minimum semblance required for a contribution to be accepted, default 10%), S/N enhance (additional signal-to-noise weighting), Angle distance selection, and Radial distance selection. These settings must match those used when the ZO-MF storage was computed.
Selects how the stretch mute is applied to the enhanced gathers. Default: Use mute factor. In the Use mute factor mode, a single time value (in ms) defines the mute threshold applied uniformly. In the Use mute function mode, a time-variable mute function (derived from the ZO-MF parameters) is applied instead, providing more accurate muting that adapts to the local moveout.
The mute time threshold in milliseconds. Default: 1000 ms. Active when Mute type is set to Use mute factor. Traces whose time shift during multi-focusing summation exceeds this value will be muted in the output. Increase this value to retain more of the far-offset data; decrease it to remove heavily stretched contributions.
Selects how the Common-Offset aperture is applied during enhancement. Default: XY aperture. In XY aperture mode, traces are selected based on their CMP distance from the output bin in the horizontal plane. In Offset aperture mode, traces are selected based on their source-receiver offset. Use XY aperture for standard 3D surveys; use Offset aperture when offset-class regularization is required.
When enabled, output traces that contain only zero values (bins with no valid input data in their aperture) are removed from the output file. Default: false. Enable this option to reduce the output file size and avoid zero-filled trace placeholders in areas not covered by the input data.
A parameter group controlling the spatial interpolation method used to compute MF weights at output bin locations that do not coincide exactly with MF storage bin centres. The parameters in this group apply to the MF weight interpolation step, not to the seismic trace amplitudes themselves.
The spatial interpolation algorithm for MF weights. Default: Triangulation. Triangulation is faster and works well when the MF storage grid is dense and regular. Kriging is more accurate when the MF storage grid is irregular or sparse, and it honours the spatial correlation structure of the data. When Kriging is selected, the covariance type, range, and number of points parameters below become active.
The covariance model used by the Kriging interpolator. Default: Exponential. The Exponential model is a good general-purpose choice for most seismic surveys. The Spherical model has a finite range of influence and is suitable when the MF parameters are known to be uncorrelated beyond a certain distance. The Gaussian model produces the smoothest interpolation and is appropriate when MF parameters vary very gradually across the survey.
The spatial correlation range for Kriging, in metres. Default: 100000 m. This defines the maximum distance over which MF storage bins influence the interpolated weight at an output bin location. The default value is intentionally large to ensure all available MF bins contribute. Reduce this value only if the MF parameters are known to vary significantly over short distances in your survey area.
The maximum number of nearest MF storage bins used in each Kriging weight estimate. Default: 15. Using more points increases accuracy but also increases computation time. Values between 10 and 20 are appropriate for most surveys.
A parameter group that controls the spatial aperture used to gather input traces for the multi-focusing summation. It is divided into two sub-groups: MF trace selection (which governs how MF storage bins are selected) and Raw data (which governs how raw input seismic traces are selected). Setting these distances correctly is critical to obtaining a well-focused, noise-free enhancement output.
Controls which ZO-MF storage bins contribute their wave-field parameters to the enhancement of each output bin. Only storage bins within the specified distance from the output bin CMP are used.
The maximum lateral distance (in metres) between an output bin and a ZO-MF storage bin for the storage bin to contribute its multi-focusing parameters to the enhancement. Default: 50 m. Minimum: 0. Maximum: 10000 m. Set this to approximately half the bin spacing of the ZO-MF storage grid to ensure that the closest storage bin always provides its parameters. Increasing this value beyond the storage bin spacing allows multiple storage bins to contribute, which may improve quality in low-fold areas.
The maximum number of ZO-MF storage bins whose wave-field parameters are used for each output bin, even if more bins fall within the distance limit. Default: 8. Minimum: 1. Increasing this value allows the module to blend parameters from more storage bins, which can smooth the enhancement over larger areas but increases computation time.
A sub-group of parameters that control how raw input seismic traces are selected for the multi-focusing summation aperture. The parameters here govern the spatial extent, fold limit, azimuth range, and Fresnel zone aperture scaling used when gathering input traces around each output bin.
A stride factor for sub-sampling the input raw traces when gathering them for the aperture. Default: 1 (use every trace). Minimum: 1. Setting this to 2 uses every second input trace, halving the computation cost. Use values greater than 1 only when processing very dense surveys and speed is more important than maximum enhancement quality.
The maximum lateral distance (in metres) between a raw input trace CMP and the output bin for the trace to be included in the enhancement aperture. Default: 150 m. Minimum: 0. Maximum: 100000 m. This is the primary aperture parameter for raw trace selection. Set it to at least the bin spacing of the input data, and typically 2–4 times the bin spacing for a good enhancement fold. Increasing this value broadens the aperture, which improves signal quality but may reduce lateral resolution.
The maximum source-to-receiver distance (in metres) used to filter raw input traces when building the enhancement aperture. Default: 100000 m (effectively no limit). This parameter allows the exclusion of very far-offset traces that may degrade the enhancement. Leave at the default unless you need to restrict the maximum source-receiver separation for quality reasons.
When enabled (default: true), the spatial aperture for raw trace selection is applied symmetrically around the output bin. When disabled, the aperture may be asymmetric (for example, only gathering traces from one azimuthal quadrant). Leave enabled for standard 3D surveys to ensure balanced fold from all directions.
An optional 2D map (GMatrixItem) of Fresnel zone scale factors. When the FZ matrix or picking parameter is set to matrix, this map provides spatially varying multipliers that scale the CMP aperture radius (Max distance to CMP) at each output bin location. Values greater than 1 expand the aperture; values less than 1 shrink it. Use this to account for spatially varying Fresnel zone sizes across the survey.
Controls which aperture distance(s) are scaled by the Fresnel zone factor. Default: cmp and cs-cr. none — no Fresnel zone scaling is applied. cmp — only the CMP aperture is scaled. cs-cr — only the source-receiver distance is scaled. cmp and cs-cr — both distances are scaled. This parameter is only relevant when the FZ matrix or picking is not set to none.
An optional velocity model (GVelocityModelAbstractItem) used as the source of Fresnel zone scale factors when FZ matrix or picking is set to model. The module samples the maximum value from this model at each output bin location and uses it as the aperture scale factor, enabling model-based spatially varying Fresnel zone corrections.
Selects the source of the Fresnel zone aperture scaling factors. Default: none (no Fresnel zone scaling). Set to matrix to use the FZ factor matrix map. Set to model to derive scale factors from the FZ abstract item velocity model. Fresnel zone scaling is an advanced feature that allows the aperture to vary spatially to match the physical Fresnel zone size, improving lateral resolution in the enhanced output.
When enabled, MF storage bins whose minimum source-receiver offset in the input data exceeds the Min offset calculation threshold are excluded from the calculation. Default: false. This is useful when certain survey areas lack near-offset coverage and the ZO-MF parameters computed there are unreliable.
The minimum source-receiver offset threshold in metres. Active when Use min offset to calculate is enabled. Storage bins where all input traces have offsets greater than this value are skipped. Default: 100000 m. Set this to a realistic near-offset limit for your acquisition (for example, 50–200 m) to exclude bins with no near-offset coverage.
A group of parameters that restrict the enhancement to a sub-area of the survey, defined by inline and crossline ranges. All four limits default to -1, which means no restriction (the full survey is processed). Use this to process a test patch before committing to a full survey run, or to reprocess a specific sub-area without rewriting the rest of the output file.
The first inline number to include in the calculation. Default: -1 (no lower limit — process from the first available inline). Set to a positive value to start processing from a specific inline.
The last inline number to include in the calculation. Default: -1 (no upper limit — process through the last available inline).
The first crossline number to include in the calculation. Default: -1 (no lower limit).
The last crossline number to include in the calculation. Default: -1 (no upper limit).
When enabled (default: true), the correlation window used to align traces before multi-focusing summation adapts dynamically to the local moveout, providing more accurate alignment of reflected events at all offsets and times. When disabled, a fixed-length window (defined by Interpolation window length) is applied. Leave enabled for best results in most surveys. Disable only if the dynamic window causes instability in areas of very complex moveout.
The half-length of the correlation window (in seconds) used during the multi-focusing trace alignment and summation. Default: 0.010 s (10 ms). A longer window improves the stability of the coherency estimate but reduces temporal resolution. Typically set to 1–2 dominant periods of the data. Valid range: 0–1000 s.
The half-length of the stabilization window (in seconds) used to normalize the enhancement operator and prevent division by very small values. Default: 0.010 s (10 ms). Increasing this value provides more robust enhancement in low signal-to-noise areas but may slightly reduce amplitude fidelity. Valid range: 0–1000 s.
Controls how results are written to the output GSD file. Default: Direct. In Direct mode, each enhanced gather is written to a fixed location in the output file based on its bin index, enabling random access and restart capability. In Append mode, results are appended sequentially to the output file. Use Direct mode for standard processing. Use Append mode when combining results from multiple separate processing runs.
The maximum allowable time shift (in seconds) applied to individual traces during the multi-focusing trace alignment step. Default: 99999 s (effectively unlimited). Setting a finite limit prevents large incorrect shifts from corrupting the output in areas where the ZO-MF parameters are inaccurate. For example, setting this to 0.05 s (50 ms) limits the alignment to physically reasonable shifts for typical seismic reflection surveys.
A container for experimental and research parameters. These settings are for advanced testing purposes and should not be changed in standard processing workflows.
An optional path to a plain text (.txt) file listing inline/crossline pairs whose computed results in the output file should be marked as not-yet-calculated, forcing them to be recalculated. This is a research tool for selectively invalidating and re-running specific bins without reprocessing the entire survey. Leave blank in normal processing.
A group of parameters that configure the SEG-Y data cache used during processing. These settings control how much of the input SEG-Y file is buffered in memory at once, affecting the balance between I/O performance and memory consumption. The default values are appropriate for most systems.
Selects whether the multi-focusing summation is performed on the CPU or GPU. GPU execution can significantly accelerate the enhancement on large 3D surveys when CUDA-capable graphics cards are available. Select CPU if no GPU is present or if GPU results need to be verified against the CPU implementation.
Options for distributing the enhancement computation across multiple nodes in a cluster. When enabled, the module splits the output bins among the available nodes and each node processes its assigned chunk independently, writing results to the shared output file. This is the recommended approach for production-scale 3D surveys.
The number of output bins processed in each distributed work chunk. Larger values reduce the number of inter-node communications and improve throughput, but increase the memory requirement per node. Smaller values improve load balancing for surveys with uneven fold distribution.
Sets the maximum number of CPU threads used per node during distributed execution. Use this to reserve system resources for other processes running on the same node, or to prevent oversubscription on systems with limited RAM per core.
An optional text suffix appended to the job name in the distributed computing queue. Use this to identify and distinguish jobs from different processing runs in the cluster job manager.
When enabled, allows the user to specify a custom CPU core affinity mask for the processing threads, pinning them to specific physical cores. This is an advanced performance tuning option for multi-socket servers and should normally be left disabled.
The CPU core affinity mask specifying which cores the processing threads may use. Active only when Set custom affinity is enabled.
The number of CPU threads used for the multi-threading parallelization of the enhancement computation on each node. The module automatically selects the most efficient threading strategy (outer-loop or inner-loop parallelization) based on the gather fold relative to the thread count. Set this to the number of physical cores available for seismic processing on the node.
A group of options that allow external scripts to be executed before and after the enhancement run. Use these to automate pre- and post-processing steps such as file staging, notifications, or quality control checks.
Path to a script that is executed before the enhancement begins. Leave blank if no pre-run script is needed.
Path to a script that is executed after the enhancement completes. Leave blank if no post-run script is needed.
When enabled, this module is bypassed during flow execution. The module is skipped without processing. Use this to temporarily disable the enhancement step during flow testing without removing it from the processing sequence.
The primary output of this module is the enhanced pre-stack seismic gather file written to the path specified in Output file name (GSD format). Each output gather corresponds to one output bin from the geometry defined by the output trace headers. The output gathers have the same time sampling and number of samples as the input data, with amplitudes enhanced by multi-focusing weighted summation. No data connector outputs are registered from this module; all results are written directly to the GSD file.
Clicking this action reloads the ZO-MF storage file and refreshes the map displays showing the input bins, output bins, and the calculation area (restricted by the inline/crossline limits). Run this action after connecting the storage file or changing the calculation area parameters to update the visual overview before starting the enhancement. The map displays show: Calculation area (bins that will be processed, shown in blue), Output bins (all output bin locations, shown in green), and Input bins (bins present in the MF storage, shown in violet).