Linear Noise Removal (LNR)
Linear Noise Removal (LNR) removes linear energy from records. It uses a windowed, time-domain, high-resolution tau-p transform to model the energy, which is then subtracted from the records.
- In the Control Panel, open the Process tab.
- At the tab header, click the Add icon and select New Process.
- Scroll down and double-click on Linear Noise Removal.
- Type a name for the process and click OK.
Configuring Linear Noise Removal
- Volume: Select the input volume.
Optionally applies an adaptive gain to the input data and removes it before returning the final result. This may improve the results depending on the input data and should be tested. If used, the only parameter is the AGC window length in milliseconds.
Low cut settings
Optionally limits the frequency range of the noise model. In most situations it is beneficial to limit the (very low) frequencies of the modelled noise. If the linear noise has significant low frequency energy remaining then a second pass of LNR with the low cut filter disabled can be run (if required). It is recommended to run with the low cut filter enabled in the first instance.
By default, this option is enabled using a low cut range of 3-7 Hz.
- Volume: The records are based on the sort order of the input data (i.e. input data organised into CMPs will be processed as CMP records).
- Offset: The relative location of the traces within a record is determined either from the OFFSET header, or by calculating offset from the SX, SY, GX, GY headers (which honour the SCALCO header).
- Start Time: The time at which to begin modelling. If blank, the entire record will be modelled.
- Start Ramp: The ramp to be applied, if processing doesn't begin at the start of the record. The ramp, centred on the start time, is used when blending the processed and unprocessed data back together.
- # Traces in Window: The size of each processing window, each of which covers the full time range of the record. It must be an odd number.
- # P Traces: The number of basis functions to be used per window.
- Min/max dip to model: The full range of dips (specified in velocity units) to be modelled for the input gather. The range of dips should span those present in the input record (both signal and noise).
Min/max dip to remove: The range of dips to be removed (specified in velocity units). The specified mute ramp (also in velocity units) is added to this range. This should be some smaller subset of the min/max dip to model.
Note: As velocity decreases dips become steeper. 1,500 m/s is a much steeper dip than 10,000 m/s. To ensure you model flat events, your min/max dip to model should span both negative and positive velocities, as shown in the figure below. So, for example modeling a range of dips from -5,000 m/s to 1,000 m/s would include all of the dips shown in the figure (including flat events). Note that a vertical dip corresponds to a velocity of 0 m/s.