Automatic Gain Control to Correct for Geometrical Spreading+Anelastic Losses.

The seismic amplitudes are usually strongest nearest the source and at early times. Consequently, the raw records usually show very weak or non-existent reflections at the far-offset traces. To display these important signals we gain the data in some fashion. This gain procedure can be a combination of several methods:

• Backing out the geometric spreading by multiplying each trace with the reciprocal of the geometric spreading factor v(t)*t, where v(t) is velocity at time tand t is the 2-way traveltime. The attenuation can also be backed out by multiplying the trace with $e^{\vert\alpha\vert t}$, where $\alpha$ is the attenuation constant.
• Automatic gain control (AGC) of the data. Here the data within a specified time window is gained to a constant energy level. The window is slid down the trace so the energy is equilibrated. Shorter windows boost everything while longer windows tend to show true relative amplitudes.

%
% MATLAB 5.0 script for applying AGC to CSG traces in a CSG. 
% For each trace, the energy within a moving window of length
% np is computed and the sample in the center part of this window
% is divided by this energy.
%
% data(x,t)   - input - nx x nt CSG matrix 
% np          - input - length of AGC window
% out1        - output- traces with AGC applied
%
[nx,nt]=size(data);
inn=[fliplr(data(:,1:np)) data fliplr(data(:,nt-np:nt))];
inn=inn.^2;
f=ones(np,1)/np;
out=conv2(inn,f');
start=round(np+np/2);endd=start+nt-1;
outt=out(:,start:endd);
out1=data./(sqrt(outt)+.000001);