filttest.m

基于matlab的反演程序,用于地球物理勘探中射线追踪及偏移成像程序.

function [fltr,f]=filttest(trin,t,fmin,fmax,phase,max_atten)

% [fltr,f]=filttest(trin,t,fmin,fmax,phase,max_atten)
% [fltr,f]=filttest(trin,t,fmin,fmax,phase)
% [fltr,f]=filttest(trin,t,fmin,fmax)
%
% FILTTEST designs and returns the frequency spectrum of the 
% filter which FILTF applies. It allows the direct examination of 
% the spectrum of this filter. 
% FILTF filters the input trace in the frequency domain.
% Trin is automatically padded to the next larger power of
% two and the pad is removed when passing trout to output. 
% Filter slopes are formed from Gaussian functions.
%
% trin= input trace
% t= input trace time coordinate vector
% fmin = a two element vector specifying:
%        fmin(1) : 3db down point of filter on low end (Hz)
%        fmin(2) : gaussian width on low end
%   note: if only one element is given, then fmin(2) defaults
%         to 5 Hz. Set to [0 0] for a low pass filter  
% fmax = a two element vector specifying:
%        fmax(1) : 3db down point of filter on high end (Hz)
%        fmax(2) : gaussian width on high end
%   note: if only one element is given, then fmax(2) defaults
%         to 10% of Fnyquist. Set to [0 0] for a high pass filter. 
% phase= 0 ... zero phase filter
%       1 ... minimum phase filter
%  ****** default = 0 ********
% note: Minimum phase filters are approximate in the sense that
%  the output from FILTF is truncated to be the same length as the
%  input. This works fine as long as the trace being filtered is
%  long compared to the impulse response of your filter. Be wary
%  of narrow band minimum phase filters on short time series. The
%  result may not be minimum phase.
% 
% max_atten= maximum attenuation in decibels
%   ******* default= 80db *********
%
% fltr= filter spectrum. Will be complex if phase=1
% f= frequency coordinate vector for fltr
%
% by G.F. Margrave, November 1991
% 
% NOTE: It is illegal for you to use this software for a purpose other
% than non-profit education or research UNLESS you are employed by a CREWES
% Project sponsor. By using this software, you are agreeing to the terms
% detailed in this software's Matlab source file.
 
% BEGIN TERMS OF USE LICENSE
%
% This SOFTWARE is maintained by the CREWES Project at the Department
% of Geology and Geophysics of the University of Calgary, Calgary,
% Alberta, Canada.  The copyright and ownership is jointly held by 
% its author (identified above) and the CREWES Project.  The CREWES 
% project may be contacted via email at:  [e-mail]
% 
% The term 'SOFTWARE' refers to the Matlab source code, translations to
% any other computer language, or object code
%
% Terms of use of this SOFTWARE
%
% 1) Use of this SOFTWARE by any for-profit commercial organization is
%    expressly forbidden unless said organization is a CREWES Project
%    Sponsor.
%
% 2) A CREWES Project sponsor may use this SOFTWARE under the terms of the 
%    CREWES Project Sponsorship agreement.
%
% 3) A student or employee of a non-profit educational institution may 
%    use this SOFTWARE subject to the following terms and conditions:
%    - this SOFTWARE is for teaching or research purposes only.
%    - this SOFTWARE may be distributed to other students or researchers 
%      provided that these license terms are included.
%    - reselling the SOFTWARE, or including it or any portion of it, in any
%      software that will be resold is expressly forbidden.
%    - transfering the SOFTWARE in any form to a commercial firm or any 
%      other for-profit organization is expressly forbidden.
%
% END TERMS OF USE LICENSE
 
% set defaults
 if nargin < 6
   max_atten=80.;
 end
 if nargin < 5
   phase=0;
 end
 if length(fmax)==1
   fmax(2)=.1/(2.*(t(2)-t(1)));
 end
 if length(fmin)==1
   fmin(2)=5;
 end
 
 [rr,cc]=size(trin);
 trflag=0;
 nt=length(t);
 if(cc==nt)
		trin=trin';
		trflag=1;
 elseif(rr~=nt)
   warning('time vector length not found in input matrix dimensions, filtering columns'); 
 end
 
 dbd=3.0; % this controls the dbdown values of fmin and fmax
% forward transform the trace
  ntrout=size(trin,1);
  trin=padpow2(trin);
  t=xcoord(t(1),t(2)-t(1),length(trin));
  [Trin,f]=fftrl(trin,t);
  nf=length(f);
  df=f(2)-f(1);
% design low end gaussian
  if fmin(1)>0
   fnotl=fmin(1)+sqrt(log(10)*dbd/20.)*fmin(2);
%HDG commented out, otherwise fmin(1) is not 3 dbd point
%  fnotl= round(fnotl/df)*df;
   gnot=10^(-max_atten/20.);
   glow=gnot+gauss(f,fnotl,fmin(2));
%HDG added to force mean to zero
   glow(1)=0;
else
   glow=0;
   fnotl=0;
  end
% design high end gaussian
 if fmax(1)>0
  fnoth=fmax(1)-sqrt(log(10)*dbd/20.)*fmax(2);
%HDG commented out, otherwise fmax(1) is not 3dbd point
%  fnoth= round(fnoth/df)*df;
  gnot=10^(-max_atten/20.);
  ghigh=gnot+gauss(f,fnoth,fmax(2));
 else
  ghigh=0;
  fnoth=0;
 end
% make filter
  fltr=ones(nf,1);
%HDG change to floor and ceil from round
  nl=floor(fnotl/df);
  nh=ceil(fnoth/df);
  if nl==0
    fltr=[fltr(1:nh) ghigh(nh+1:length(f))];
  elseif nh==0
%HDG change from fltr=[glow(1:nl-1);fltr(nl:length(f))];
    fltr=[glow(1:nl+1);fltr(nl+2:length(f))];
  else
%HDG change from fltr=[glow(1:nl-1);fltr(nl:nh);ghigh(nh+1:length(f))];
   fltr=[glow(1:nl+1);fltr(nl+2:nf)].*[fltr(1:nh);ghigh(nh+1:length(f))];
   fltr=fltr/max(abs(fltr));
  end
% make min phase if required
  if phase==1
    L1=1:length(fltr);L2=length(fltr)-1:-1:2;
    symspec=[fltr(L1) conj(fltr(L2))];
    cmpxspec=log(symspec)+i*zeros(size(symspec));
    fltr=exp(conj(hilbm(cmpxspec)));
  end