1D Filtering Operations

A large part of what is called seismic processing, may be thought of as being ``filtering.'' There are filtering operations for 1D applications in the SU package that span simple filtering tasks, to more sophisticated tasks including deconvolution and wavelet shaping operations. These operations are 1D, in that they are applied trace by trace.

Several types of filtering operations that arise in seismic processing are

• zero bandpass, bandreject, lowpass, and highpass , and notch filtering,
• minimum or zero phase Butterworth filtering,
• Wiener prediction error (deconvolution),
• Wiener wavelets shaping,
• convolution,
• crosscorrelation,
• autocorrelation,
• data resampling with sinc interpolation,
• fractional derivatives/integrals,
• median filtering,
• time varying filtering.

The programs that are provided to meet these needs are:

• SUFILTER - applies a zero-phase, sine-squared tapered filter
• SUBFILT - apply Butterworth bandpass filter
• SUACOR - auto-correlation
• SUCONV, SUXCOR - convolution, correlation with a user-supplied filter
• SUPEF - Wiener predictive error filtering
• SUSHAPE - Wiener shaping filter
• SURESAMP - Resample in time,
• SUFRAC - take general (fractional) time derivative or integral of data, plus a phase shift. Input is TIME DOMAIN data.
• SUMEDIAN - MEDIAN filter about a user-defined polygonal curve with the distance along the curve specified by key header word
• SUTVBAND - time-variant bandpass filter (sine-squared taper)