|
<< Click to Display Table of Contents >> Navigation: Stack > Horizontal velocity picking |
Note: This module is deprecated and retained for backward compatibility only. New workflows should not use this module.
The Horizontal velocity picking module performs velocity analysis along interpreted horizons in the lateral (spatial) direction. Instead of analyzing velocity variation with depth at a single CMP location (as in conventional vertical velocity analysis), this module computes a semblance spectrum that shows how the optimal stacking velocity varies laterally across the survey area along a selected horizon interface. This is useful for detecting lateral velocity anomalies such as gas clouds, shallow channels, or fault-related velocity changes that affect the quality of post-stack imaging.
The module reads a pre-stack stack volume and a set of interpreted horizons, then calculates a horizontal semblance cube by scanning a range of velocities and measuring trace coherency along each horizon at each bin location. The resulting horizontal semblance panels are displayed interactively in inline and crossline views, enabling the user to pick the best velocity at each location either manually or automatically. Picked velocities are used to update the velocity model laterally.
The module provides four custom actions: Calculate semblances computes the semblance cube for the selected bin; Import horizon loads a horizon from an external source into the active picking session; Auto picking automatically picks the semblance maximum across all bins; and Calculate current inline/crossline recalculates and refreshes the display for the currently selected line.
The main seismic data connection for the module, providing access to the input trace stream, sorted headers, and SEG-Y file handle. This is the standard sequence data item inherited from the processing pipeline.
A handle to the SEG-Y file that contains the input seismic data. The module reads the sample interval (dt) and the total number of samples from this handle to configure the semblance computation grid.
The sorted trace header index that allows the module to efficiently locate and extract inline and crossline subsets from the full stack volume. This index must be pre-sorted by CMP or bin location.
The post-stack seismic volume on which horizontal semblance is computed. The module slices this volume into inline and crossline gathers at the selected bin location and measures trace coherency along the provided horizon surface at each velocity trial.
The existing vertical velocity model (picks from a conventional velocity analysis module). This provides the starting velocity field that the horizontal picking session will refine laterally. The vertical velocity picks define the background velocity used for NMO correction before horizontal semblance is computed.
One or more interpreted horizon surfaces along which horizontal semblance is evaluated. The module extracts a time value from each horizon at every bin location and reads the corresponding semblance value at that time from the pre-computed semblance cube. At least one horizon must be loaded for the module to operate. The active horizon is selected using the Current horizon parameter.
Path to a KDB-format binary file used to cache the computed semblance cube on disk. When this file is specified and the semblance cube has already been computed, subsequent interactive sessions can reload the cached data instantly without recomputing from scratch. Leave empty to recompute every time. Providing a storage file is strongly recommended for large datasets.
A constant velocity value (in m/s) used to replace the background velocity model when computing horizontal semblance at the selected bin. Set this to 0 (default) to use the vertical velocity picks from the connected velocity picking item instead of a constant value. Use a non-zero constant velocity when you want to test the semblance response at a specific uniform velocity, for example to QC a flat-layer assumption.
Controls the velocity bandwidth (in m/s) displayed in the horizontal semblance panel, centered on the background velocity. Default is 1200 m/s. A wider value shows a broader range of lateral velocity variation in the display, which is useful when large velocity contrasts are expected across the survey. Narrow this value to zoom in on a smaller velocity range when picking near-homogeneous layers.
The zero-based index of the horizon (from the Input horizons list) on which horizontal semblance analysis is performed. Default is 0 (the first horizon). Increment this value to switch the analysis to a deeper horizon without re-running the full calculation.
A time offset (in seconds) applied to the horizon when extracting semblance values from the cube. Default is 0.05 s (50 ms). Use this to shift the analysis window up or down relative to the picked horizon time, for example to account for a known static shift or to evaluate semblance slightly above or below the reflector.
When enabled (default: on), the semblance values extracted along the horizon are normalized to the range [0, 1] before display. Normalization makes it easier to visually compare semblance peaks across bins with very different absolute amplitudes. Disable this option when you need to preserve absolute semblance magnitudes for quantitative comparison between locations.
This parameter group controls how the semblance cube is computed. Expand it to access the velocity scan range, semblance quality, and smoothing settings described below.
The minimum velocity (in m/s) at which the semblance scan begins. Default is 1000 m/s. Set this to the lowest geologically plausible stacking velocity for the zone of interest. Values below the true velocity will produce low semblance and appear as weak, unfocused energy on the semblance panel.
The maximum velocity (in m/s) at which the semblance scan ends. Default is 6000 m/s. Set this to the highest velocity expected in the target zone. Choosing an excessively high end velocity increases computation time without adding information in most shallow surveys.
The step size (in m/s) between successive velocity trials in the semblance scan. Default is 50 m/s. Smaller increments produce a finer, smoother semblance panel that resolves closely spaced velocity events, but increase computation time proportionally. Use 25 m/s or finer when the target requires high velocity precision (for example, near salt flanks or in areas of rapid lateral velocity change).
The length of the time window (in seconds) over which individual semblance traces are smoothed before display. Default is 0.05 s (50 ms). Smoothing suppresses short-period noise and produces a cleaner semblance image, making peaks easier to pick. Increasing this window produces more stable but spatially smeared semblance peaks; reducing it preserves sharper peak definition at the cost of more noise.
The maximum allowable NMO stretch ratio (as a fraction, where 1.0 = 100%). Default is 0.5 (50%). Traces whose NMO stretch at a given offset and velocity exceeds this threshold are muted before semblance calculation. This prevents highly stretched, low-frequency energy at far offsets from artificially inflating semblance values at incorrect velocities. Typical values are 0.3 to 0.6. Use smaller values (stricter muting) for shallow fast layers where stretch is severe.
The number of samples over which the semblance panel is smoothed in the velocity direction (Y axis of the semblance display). Default is 10 samples. Increasing this value blurs velocity boundaries laterally on the semblance panel, which helps suppress noise-driven spurious peaks when the signal-to-noise ratio is low. Reduce it when you need to resolve closely spaced velocity events.
The time window length (in seconds) used to normalize the semblance panel along the time axis. Default is 0.1 s (100 ms). This running normalization compensates for amplitude variations with depth so that semblance peaks at all time levels appear with similar brightness, regardless of the absolute energy of the reflections. Use a larger window in data with strong amplitude contrasts between shallow and deep events.
Controls the spatial density at which semblance is calculated, affecting both visual quality and computation speed. Options are High, Normal, and Low (default). Use High for final QC picking in areas of complex geology where fine spatial detail is important. Use Low during initial exploration and for large datasets where fast interactive response is preferred.
The number of adjacent bins (in each direction) merged into a super-gather for semblance calculation. Default is 3, meaning a 7x7 bin neighborhood (radius of 3 bins) is combined. Larger super-gather radii improve semblance signal-to-noise ratio by increasing the fold used in the coherency calculation, at the cost of reduced lateral resolution. Use a smaller radius (1 or 2) where spatial velocity contrasts are sharp; use a larger radius (4 or more) in low-fold or noisy surveys.
This parameter group controls the behavior of the interactive and automatic picking tools. Expand it to configure the manual picking window, automatic picking window, pick type, magnet behavior, and picking threshold described below.
The half-width of the time search window (in seconds) used when placing a manual pick on the horizontal semblance panel. Default is 0.02 s (20 ms). When the user clicks on the semblance display, the module searches within this time window around the click position to snap the pick to the nearest local semblance maximum (if the magnet option is enabled). Increase this value when reflections have low frequency and broader semblance peaks.
The search window (in seconds) used by the automatic picking algorithm when propagating picks laterally across bins. Default is 0.06 s (60 ms). The auto-picker finds the semblance maximum within this time window relative to the pick on the previous bin. A wider window allows the autopicker to follow steeply dipping or rapidly changing velocity trends; a narrower window constrains picks to more gently varying events.
Determines the polarity of the seismic event tracked by the picking engine. Options are Positive phase (default) and Negative phase. Choose the polarity that corresponds to the reflector being tracked on the stack. In standard zero-phase data with SEG normal polarity, a hard kick (impedance increase) appears as a positive peak. If you are tracking a trough (impedance decrease), select Negative phase to ensure the picker locks on to the correct event.
Controls how a manual click on the semblance panel is snapped to an event. Options are Nearest (default) and Max energy. Nearest snaps the pick to the closest local semblance maximum within the picking window. Max energy snaps to the absolute maximum semblance value within the window, regardless of proximity to the click. Use Nearest for precise interactive control; use Max energy when you want to rapidly capture the strongest event within the window.
The minimum normalized semblance value (from 0 to 1) that a bin must exceed before a pick is accepted during automatic picking. Default is 0.5 (50%). Bins where the semblance peak falls below this threshold are left unpicked. Raise the threshold to restrict automatic picks to high-confidence locations with strong reflector coherency. Lower it in areas of poor signal-to-noise ratio where even weak semblance peaks carry useful velocity information.