Preprocessing: Noise Removal


    T-F Trim and FF3D

    Noise removal is a critical step in data pre-processing. TGS has several specific routines that are effective tools for noise removal:

    • T-F Trim is effective for noise patterns that are high amplitude but restricted to a limited frequency range within the seismic bandwidth.
    • FF3D is used for linear noise train removal in three dimensions. This can be used in 2D or in 3D using offset as one of the dimensions.

    All of these tools can dramatically improve data signal of the final product by removing noise and avoiding associated artifacts.

    T-F Trim

    T-F Trim is designed for cable jerk noise and other high amplitude and/or band-limited types of noise. The data are transformed to the frequency domain and divided into sub-bands. Within each variant sub-band, the median amplitude value is calculated and amplitudes that exceed the median are replaced by the median value.

    T-F Trim Examples