|
<< Click to Display Table of Contents >> Navigation: Velocity > Time variant smoothing (3D) |
This module is deprecated. It is provided for legacy workflows only.
The Time Variant Smoothing (3D) module applies a spatially and temporally adaptive smoothing filter to a 3D velocity or attribute volume. The smoothing window length grows linearly with time, so that shallow intervals receive minimal smoothing and deeper (noisier) intervals are smoothed over a longer window. This approach preserves near-surface velocity detail while suppressing noise and oscillations at depth. The module supports two smoothing methods: Median (robust to outliers) and Average (conventional boxcar). The output is a new SEG-Y file containing the smoothed volume, suitable for use as a migration velocity model or for velocity quality control.
The module reads a full 3D volume from the connected SEG-Y data handle, sorts it by inline, and applies the time-variant filter inline by inline. The spatial skip parameters allow the computation grid to be coarsened for efficiency at the cost of spatial resolution.
The collection of trace headers for all traces in the 3D velocity volume. These headers provide the inline, crossline, and coordinate information used to sort and spatially organise the volume before smoothing.
A handle to the SEG-Y file containing the 3D velocity or attribute volume to be smoothed. Connect this to the read module that opens the input velocity volume.
The full path and file name of the SEG-Y output file (.sgy or .segy) where the smoothed volume will be written. The output file will have the same number of traces, sample rate, and record length as the input volume. Use the file browser to select the destination directory.
The number of inline lines skipped between smoothing computations in the inline direction. A value of 1 processes every inline; a value of 10 processes every 10th inline and interpolates the rest. Increasing this value reduces computation time but may reduce the lateral resolution of the smoothed volume in the inline direction.
Default: 10.
The number of crossline lines skipped between smoothing computations in the crossline direction. Works analogously to Inline skip. Use the same value as Inline skip to maintain isotropic spatial decimation of the computation grid.
Default: 10.
The smoothing method applied within each time window:
Median — replaces each sample with the median value within the window. Robust to outliers and spikes; recommended when the velocity field contains erroneous picks or strong local anomalies.
Average — replaces each sample with the arithmetic mean within the window. Produces a smooth result but is sensitive to extreme values.
Default: Median.
The initial smoothing window length in milliseconds at the start time of the volume (defined by Input volume start time). The window length increases linearly with time from this base value, growing by a percentage of elapsed time set by the % of time parameter. Increase this value to apply stronger smoothing at shallow depths.
Default: 24 ms.
The rate at which the smoothing window length increases with time, expressed as a percentage of the elapsed time since the start time. For example, a value of 10 means the window grows by 10 ms for every 100 ms of elapsed time. Higher values apply progressively stronger smoothing at greater depths, which is appropriate when velocity uncertainty increases with depth.
Default: 10%.
The two-way time at the first sample of the input volume, in milliseconds. This value defines the reference time from which the time-variant window growth is computed. Set this to match the start time of your velocity volume (typically 0 ms for a surface-referenced dataset). If the volume starts at a non-zero time (e.g. a cropped dataset), set this accordingly so that the window length is computed correctly at each depth.