|
<< Click to Display Table of Contents >> Navigation: General > Signal Noise Attributes |
Signal Noise Attributes is a gather-browsing and signal-to-noise ratio (SNR) analysis module. It loops through a sorted seismic dataset gather by gather, reading each gather from a SEG-Y file and making it available for display and downstream processing. At each gather location, the module computes SNR attributes using one of three selectable algorithms: STA/LTA (short-term average over long-term average), Velocity Mute separation, or Cross Correlation.
The module is typically used for pre-processing quality control: it allows a geophysicist to step through individual gathers interactively, or to run a full batch pass over a specified gather range to build spatially varying SNR attribute maps. Results can be passed to child modules in the processing sequence, or their computed attribute values can be spatially interpolated across the survey area using the built-in interpolation framework. Clicking directly on the survey location map loads the nearest gather instantly.
The seismic dataset to analyse. Connect this to the output of an upstream data source or geometry-application module. All gathers in this dataset will be available for iteration.
A handle to the SEG-Y file from which raw trace data is read. This must point to the same underlying file as the Input DataItem. The module uses this handle to read batches of gathers efficiently into memory.
A sorted gather-index vector that defines how the dataset is divided into individual gathers (for example, sorted by CMP, shot, or receiver). The iteration counter is applied to this index, so the sorting key determines which header values appear in the First and Second fields and in the survey location map.
Displays the primary header key value (for example, CMP number, shot number, or X-coordinate) of the currently loaded gather. The label of this field updates automatically to match the sorting key used in the Input sorted headers, for example Gather value <CMP>. You can type a specific value here to jump directly to a gather with that primary key.
Displays the secondary header key value of the currently loaded gather when the data is sorted by two keys (for example, inline and crossline numbers). Together with First, these two fields uniquely identify the current gather position in the dataset. This field is labeled N/A and becomes non-editable if the sort index uses only one key.
The 1-based sequential index of the currently loaded gather within the full sorted gather list. For example, a value of 1 corresponds to the first gather in the sort order. You can set this value directly to jump to any gather by its position number rather than by its header key. The valid range is automatically set to match the total number of gathers in the connected Input sorted headers.
The sequential gather number at which batch execution begins. Set this to restrict processing to a specific subset of the dataset. Default is 1 (start from the very first gather). Must be less than or equal to Last seq. gather.
The sequential gather number at which batch execution stops. When the Input sorted headers are connected, this value is automatically set to the total number of gathers. Reduce it to process only an initial portion of the survey, which is useful during setup and testing.
The step size between gathers during batch execution. A value of 1 (default) processes every gather. Set to a larger integer, for example 10, to process every 10th gather for a rapid preview pass. This parameter works with the sequential index, not the header key values.
When enabled (default: on), pressing Execute advances the Sequential gather number to the next gather automatically after each run. This is convenient when stepping through gathers one by one with repeated single executions. Disable this option to keep returning to the same gather.
Controls batch execution behavior when a gather cannot be read or processed. When enabled (default: on), the batch run stops immediately at the first error and reports it. When disabled, the module skips the problematic gather and continues to the next one. Disabling this is useful when a few bad or missing gathers are expected and the rest of the survey should still be processed.
The pause (in seconds) inserted between gathers when running in animation (movie) mode. Set to 0 (default) for no delay, which is appropriate for batch processing. Increase this value, for example to 0.5 seconds, when using the module as a data browser to give time to visually inspect each gather before the display updates to the next one.
When enabled (default: on), clicking a point on the survey location map view automatically loads the nearest gather and triggers execution of any child procedures in the sequence. When disabled, clicking still loads the gather for preview but does not start child-procedure execution, so you can inspect the gather before deciding to run processing.
The half-aperture (in gather units) used to build a super gather by merging neighboring gathers around the current position. For example, a value of 2 combines the current gather with the two preceding and two following gathers into a single super gather. A value of 0 (default) disables super gather formation. Super gathers are useful for improving fold when testing velocity analysis or noise estimation on sparse datasets.
A parameter group that controls the spatial decimation of gathers independently along each sorting key dimension. These step values are applied on top of the sequential gather increment, allowing you to skip every N-th gather along the first key and every M-th gather along the second key separately.
Step size along the primary sort-key dimension (for example, along CMP inline). A value of 1 (default) processes every gather position along this axis. Set to 2 to process every other inline, and so on. Useful for building coarser attribute grids quickly.
Step size along the secondary sort-key dimension (for example, along CMP crossline). A value of 1 (default) processes every gather position along this axis. Works identically to Iteration step first but for the second sorting key. Only relevant for datasets sorted by two keys (for example, 3D inline and crossline).
Selects the algorithm used to estimate signal-to-noise ratio attributes on each gather. Three options are available:
STA / LTA Algorithm (default) — computes the ratio of a short-term amplitude average to a long-term amplitude average, sliding along each trace. This classic approach is widely used for first-break detection and noise-burst identification. Configure the window lengths using the STA/LTA Algorithm parameters group.
Veleocity Mute — separates signal from noise by applying a hyperbolic mute defined by a velocity V0 and a time origin T0. Samples inside the mute window are treated as signal; samples outside are treated as noise. This is suitable when the signal and noise zones are well separated kinematically, for example in refraction surveys.
CrossCorrelation — estimates SNR by cross-correlating adjacent traces within a time window. High correlation between neighboring traces indicates coherent signal; low correlation indicates noise. Use this option when signal coherency across traces is the primary SNR discriminator.
Parameter group for the STA/LTA algorithm. These settings are active only when SNR Detection Type is set to STA / LTA Algorithm.
Length of the short-term averaging window in seconds. Default is 0.04 s (40 ms). The STA window must be shorter than the LTA window. A shorter STA responds faster to abrupt amplitude changes such as first breaks or noise bursts, but may be more sensitive to individual noisy samples. Typical values range from 0.005 s to 0.1 s.
Length of the long-term averaging window in seconds. Default is 0.6 s (600 ms). The LTA window provides the background noise reference level. It should be significantly longer than the STA window — typically 5 to 20 times longer. A longer LTA gives a more stable noise estimate but adapts more slowly to gradual amplitude changes along the trace.
Parameter group for the velocity mute SNR algorithm. These settings are active only when SNR Detection Type is set to Veleocity Mute.
The apparent velocity (in m/s) used to define the mute boundary between signal and noise zones. This value is used in a hyperbolic equation: the mute time at a given offset is computed as the two-way traveltime corresponding to that offset divided by V0. Default is 700 m/s, which separates direct-wave signal from ground-roll and refraction in a typical near-surface scenario. Adjust to match the actual apparent velocity of the boundary you wish to track.
The zero-offset intercept time (in seconds) of the mute function. Default is 0 s. This shifts the entire mute curve later in time when there is a static delay or when the signal of interest does not start at time zero. Set T0 to the approximate two-way time of the reflector or onset that defines the signal window.
Parameter group for the cross-correlation SNR algorithm. These settings are active only when SNR Detection Type is set to CrossCorrelation.
The time window length (in seconds) over which each pair of adjacent traces is cross-correlated. Default is 0.04 s (40 ms). A shorter window provides higher temporal resolution of the SNR estimate but is more sensitive to trace-to-trace variability. A longer window gives a more stable SNR estimate over a broader time interval. Choose a window length comparable to the dominant period of the signal you expect to detect.
Expand this group to access performance tuning options that control how gathers are read from disk.
The number of gathers read from the SEG-Y file in a single disk-read operation during batch (animation) mode. Default is 100 gathers. A larger value reduces the number of disk seeks and can improve throughput on sequential storage, but increases the memory footprint per batch. Reduce this value if memory is limited or if gathers are very large (many long traces). Increasing it above the default is beneficial on high-speed SSDs or network storage.
A collection of parameter references that enables spatial interpolation of computed attribute values across the survey. Add references to output parameters from child modules (for example, an estimated noise level or SNR value) into this list. After the full batch pass, the module interpolates those values at any survey point using the observed gather locations as control points. This is used, for example, to build a spatially smooth SNR map that can then be applied as a spatially varying gain or mute function in downstream processing.