|
<< Click to Display Table of Contents >> Navigation: MultiFocusing > Engine - 3D ZO-MF |
Engine - 3D ZO-MF is the core Zero-Offset Multi-Focusing (ZO-MF) search engine for 3D seismic data. It scans every output bin in your survey and determines the optimal Multi-Focusing parameters — the wavefront curvature attributes that best describe the subsurface reflection geometry at each imaging point. The results are written to a proprietary database file (.kdb) that subsequent modules (Data Enhance - 3D ZO-MF, Imaging - 3D ZO-MF) use to reconstruct and enhance the seismic image.
For each output bin, the engine searches through a quantized table of MF parameter combinations (provided by the MF Tables Parametrization input) and identifies the set that produces the highest coherency of pre-stack seismic data gathered within the specified aperture. The search can be constrained to a velocity corridor — either from a connected picking item or from an external .corr file — to reduce computation time and avoid physically unrealistic solutions.
The module supports restart: bins that have already been computed and stored in the .kdb database are automatically skipped, so interrupted runs can be resumed without losing prior work. It supports multi-threaded (CPU), GPU-accelerated, and distributed (remote) execution modes. Both full 3D grid output (via the Output Binning in the 3D group) and arbitrary 2D line output (via the Arbitrary line geometry in the 2D group) are available, but exactly one must be configured per run.
Connect to the SEG-Y file handle that provides the pre-stack seismic data to be searched. This is the standard read/write SEG-Y connector and is active when Use trace vector on disk is set to false (the default). The engine reads amplitude data from this handle for each candidate bin during the MF parameter search.
The sorted trace header index for the input seismic data (GTraceVectorItem). This index contains source-receiver geometry information (CMP positions, offsets, azimuths) for every trace in the dataset, and is used by the engine to gather the correct traces for each output bin according to the MF aperture. Active when Use trace vector on disk is false.
Input binning of seismic data
The bin grid that describes the geometry of the raw input seismic data (source-receiver midpoints). The engine uses this grid to spatially locate traces relative to each output imaging bin. Connecting a bin grid also populates the map view with the input fold map and topography overlays for quality control.
Link to velocity constrain picking that can be taken from another procedure
An optional velocity corridor item (G3DPickingItem) that restricts the MF parameter search to a physically reasonable range of velocities at each bin. Connecting a picking item prevents the search from exploring unrealistic velocity values and significantly reduces computation time in areas where velocity is well constrained. If neither this item nor the File with picking parameter is provided, the search spans the full velocity range defined in the MF Tables Parametrization.
Service object provides parameters for MF search
This mandatory input (TableQuantizationItem) contains the full lookup table of quantized Multi-Focusing parameter combinations to be tested at each bin. It defines the velocity range, aperture sizes (inline and crossline), the time axis discretization, and all other MF search grid settings. This item is typically generated by the MF Table Parametrization module and must be connected before execution. Changing the connected item automatically updates the default aperture coeff and calculation area preview.
Link to Fresnel Zone factor table, used for calculation of data consists of few acquisitions with different parameters (nominal fold coverage).
Calculation with this table will perform unique MF nominal fold by using spatially varying Fresnel Zone percentage.
An optional spatial matrix (GMatrixItem) that applies a spatially varying Fresnel Zone aperture scaling factor across the survey. This is particularly useful when the dataset combines acquisitions with different nominal fold or shooting geometries, ensuring a consistent effective aperture for the MF search despite variations in source-receiver density. If not connected, a constant aperture defined by the MF Tables Parametrization is used.
An optional spatial map (GMatrixItem) that provides a laterally varying near-surface velocity V0. When connected, the engine uses the local V0 value at each output bin to adjust geometric spreading and travel-time computations. If not connected, a constant V0 derived from the MF Tables Parametrization is applied uniformly across the survey. This input is useful for surveys with significant near-surface velocity variations such as land data over variable weathering layers.
An on-disk trace vector data handle (TVOD format) used as an alternative to the SEG-Y data handle. This input is active only when Use trace vector on disk is set to true. The TVOD format can offer faster random-access trace reading for large 3D datasets compared to SEG-Y. When this mode is used, the SEG-Y data handle and Input trace headers inputs are hidden and replaced by this handle together with the Trace handler read bulk size parameter.
When false (default), the engine reads seismic amplitude data from the connected SEG-Y data handle using the standard trace headers index. When true, data is read from an on-disk trace vector (TVOD) file instead. Switch to TVOD mode for very large 3D surveys where SEG-Y random-access I/O becomes a performance bottleneck. In TVOD mode, the Trace handler read bulk size parameter also becomes available to tune read efficiency.
Controls how many traces are read from the on-disk TVOD file in a single I/O operation. Default: 200000 traces. Larger values reduce the number of disk read operations and can improve throughput on systems with fast storage, at the cost of higher memory usage. Visible only when Use trace vector on disk is true.
File consists the velocity corridor.
The search of the optimal MF parameters will be performed only inside this constrain.
If not specified the search will be performed for all velocities.
An optional path to a velocity corridor file (.corr format). This file defines the minimum and maximum velocity at each time-space location within which the MF parameter search is confined. Providing a corridor file is an alternative to connecting the GMFPickingItem input; both serve the same purpose of constraining the search. Using either constraint is strongly recommended for production workflows to reduce computation time and prevent spurious solutions outside the plausible velocity range.
Parameters for 3D output
This group contains settings for a full 3D grid output. Use the 3D group when you want to compute MF parameters across a regular bin grid covering the entire (or a subarea of the) 3D survey. Either the 3D group or the 2D group must be configured, but not both simultaneously.
The path and filename for the output 3D MF database file (.kdb). This file stores the optimal ZO-MF parameter sets for every computed bin in the output bin grid. Subsequent modules — such as Data Enhance - 3D ZO-MF and Imaging - 3D ZO-MF — read this file to perform seismic enhancement and imaging. If a .kdb file with the same name already exists and its internal parameters differ from the current run settings, the engine will prompt you to Cancel, Continue (append to existing database), or Rewrite the file from scratch.
Service object provides location of imaging pint (output geometry)
The output bin grid (GBinGridItem) that defines the regular grid of imaging points for 3D output. Each node of this grid is one output bin for which the MF parameters will be searched and stored. The output grid may be identical to the input binning or may be coarser (e.g., for faster initial runs). This item must be provided when the 3D group is used.
Parameters for arbitrary line calculation
This group contains settings for computing ZO-MF parameters along an arbitrary 2D line (crooked line) through a 3D survey. Use this group instead of the 3D group when you want results only along a single user-defined path, for example to QC the MF search result along a specific inline or an interpreted fault line before running the full 3D computation.
The path and filename for the output 2D (arbitrary line) MF database file (.kdb). This file serves the same role as the 3D storage file but for the bin points defined by the Arbitrary line geometry. Only one storage file (3D or 2D) may be set per execution.
Service object provides location of imaging point (output geometry)
A bin point vector (GBinPointVectorItem) that defines the sequence of imaging points along the arbitrary 2D line. Typically generated by the Create Stack Line module. Each point in this vector becomes one output bin for the ZO-MF parameter search. This item must be provided when the 2D group is used.
This group controls the spatial decimation of velocity analysis results stored in the .kdb database. A finer storage grid (smaller step values) preserves more detail for velocity QC and picking, but increases database size. Coarser values are appropriate when storage space is limited or when only a rough velocity overview is needed.
The inline decimation step for the stored velocity analysis data. Default: 5. A value of 5 means that full velocity spectra (semblance panels) are saved for every 5th inline. Reduce this value for denser velocity QC coverage; increase it to save disk space.
The crossline decimation step for the stored velocity analysis data. Default: 5. Works the same way as Each inLine but controls the crossline direction. Together, these two parameters define a coarse velocity analysis grid (e.g., every 5th inline by every 5th crossline) within the output bin grid.
This group exposes advanced tuning options that affect MF aperture behavior. These settings should only be changed from their defaults if you have a specific technical reason to do so.
A dimensionless multiplier applied to the MF aperture sizes (both inline and crossline) defined in the MF Tables Parametrization. Default: 1.0 (no scaling). Values greater than 1 expand the aperture (more traces gathered per bin, higher sensitivity but slower computation); values less than 1 shrink it. This parameter is located in the Expert parameters group and is automatically initialised from the connected MF Tables Parametrization when it is changed.
Parameters of calculation area
This group defines which bins within the output geometry will be computed. You can limit the computation to a rectangular sub-area of the survey using inline/crossline range limits, apply a regular decimation step to produce a coarse first-pass result, or restrict computation to bins that have at least one near-offset trace within a specified distance (Selection by min offset mode). The calculation area preview map is updated in real time as you change these parameters.
Selects how the calculation area is defined. Three options are available:
None (default) — The inline/crossline range and step parameters below are used to define a rectangular sub-area. This is the most common mode.
Point vector — A connected GPointVectorItem defines the exact set of bins to compute. The X coordinate of each point is interpreted as an inline number and the Y coordinate as a crossline number.
Bins — A connected GBinPointVectorItem directly specifies the bin locations to compute.
First inline to be calculated
The first inline number to include in the calculation. Default: -1 (no limit — start from the first available inline in the output grid). Enter a positive inline number to restrict computation to a sub-area of the survey. Visible only when External geometry constraint is set to None.
Last inline to be calculated
The last inline number to include in the calculation. Default: -1 (no limit — process through the last available inline). Set together with First inline number to define the inline extent of the calculation sub-area.
First xline to be calculated
The first crossline number to include in the calculation. Default: -1 (no limit). Set this together with Last crossline number to define the crossline extent of a sub-area computation.
Last xline to be calculated
The last crossline number to include in the calculation. Default: -1 (no limit). Together with First crossline number, this defines the crossline bounds of the calculation area.
Determines how the set of bins to be computed is selected within the inline/crossline bounds. Two options are available:
Constant grid (default) — Computes bins on a regular inline/crossline grid determined by the Step along inLine and Step along xLine parameters, subject to the Calculation rule. Use this mode for systematic full-survey or sub-area runs.
Selection by min offset — Selects bins based on the distance to the nearest source location. Only bins that have a source within the Min offset calculation distance are included. This is useful for surveys where some output bins are poorly illuminated (no near-offset traces) and should be excluded from the search.
The maximum distance (m) from a bin to the nearest source within which a bin is considered adequately illuminated. Default: 100000 m (effectively no constraint). Decrease this value to exclude bins that are far from any source shot point, which is common near the edges of an irregular acquisition pattern. Visible only when Selection type is set to Selection by min offset.
A spatial tolerance (m) used when merging nearby source locations for the min-offset selection computation. Default: 0 m. Sources within this distance of each other are treated as a single source for the purpose of determining bin illumination. Increase this value if your source point spacing is irregular and causes gaps in the selection map. Visible only when Selection type is set to Selection by min offset.
The desired minimum source-to-bin distance (m) used in the min-offset selection mode. Default: 1000 m. This parameter defines the target coverage radius for selecting bins when working with the Selection by min offset mode. Bins within this radius of a source point are prioritised for computation. Visible only when Selection type is set to Selection by min offset.
How calculation will be performed in case calculation steps bigger than:
1. Logical AND – calculation performed along inlines and crosslines with determined steps.
2. Logical OR – calculation performed in interception points with determined steps.
Controls which bins are computed when step sizes greater than 1 are used. Default: Logical AND. With Logical AND, a bin is computed only if it falls on both the inline step grid AND the crossline step grid (i.e., at grid intersections only — a sparser, faster computation). With Logical OR, a bin is computed if it falls on either the inline step grid OR the crossline step grid (produces a cross-hatch pattern of computed lines). Visible only when Selection type is set to Constant grid.
Step between calculated inlines
Default: 1
The inline decimation step for the computation. Default: 1 (compute every inline). Set to 2 to compute every other inline, 5 to compute every 5th inline, etc. Useful for producing a quick first-pass result over the full survey before running at full density. Visible only when Selection type is set to Constant grid.
Step between calculated xlines
Default: 1
The crossline decimation step for the computation. Default: 1 (compute every crossline). Works the same way as Step along inLine number but in the crossline direction. Visible only when Selection type is set to Constant grid.
Start calculation time
Default: 0
The start of the time interval (in seconds) over which the MF parameter search is performed. Default: 0 s. Set this to a positive value to skip shallow data and begin the search at a deeper reflector interval, saving computation time when you are only interested in a specific depth range.
End calculation time
Default: 0
The end of the time interval (in seconds) over which the MF parameter search is performed. Default: 6 s. Reduce this value to limit computation to shallower data. The search is only performed between First Time and Last Time.
When true, the engine will re-compute bins that were previously marked for recalculation via the interactive map view (using the left-mouse-button drag selection on the Bins to recalculate overlay). Default: false. This flag works together with the restart mechanism: with it false, all previously CALCULATED bins are always skipped regardless of the selection; with it true, only bins explicitly flagged in the recalculation selection will be re-run. Use this to selectively recompute problem areas without restarting the full survey.
A read-only output parameter reporting the fraction of output bins that are skipped (not computed) relative to the total number of candidate bins. A value close to 0 means nearly all bins are being computed; a value close to 1 indicates heavy sub-sampling (large step values or tight selection constraints). Use this as a quick sanity check to verify the calculation area settings before launching a long production run.
A read-only output parameter reporting the total count of bins that will be (or have been) computed in this run. This count updates after the calculation area is refreshed. It gives you an estimate of the computational workload before starting execution.
Manually refreshes the Calculation area map view and recalculates the Sparse factor and Number of calculation bins output values. While the map updates automatically when parameters change, run this action explicitly after connecting or changing geometry inputs if the map does not refresh. This also clears any cached min-offset selector so that the selection is recomputed from scratch using current data.
Marks the bins currently highlighted in the Bins to recalculate map overlay as NOT_CALCULATED in the .kdb database. The next execution run will then re-compute those bins from scratch. Use the left-mouse-button drag in the map view to select the bins you want to invalidate before running this action. This is the standard workflow for selectively reprocessing a specific part of the survey.
Marks the bins currently highlighted in the Bins to recalculate map overlay as CALCULATED in the .kdb database, so they will be skipped on the next execution run. Use this action to permanently exclude specific bins from future computation — for example, areas that are known to contain only noise or are outside the zone of interest.