BXCR

The command line arguments are:
 bxcr infile1 infile2 t1 t2 tlagmx 
  
 infile1: name of input file #1 
 infile2: name of input file #2 
 t1:     =start time of cross correlation gate (sec.)
 t2:     =end time of cross correlation gate (sec.) 
 tlagmx: =maximum cross correlation lag time (sec.)
If the second input file is the same as the first, the result will be an auto correlation. If the input files are different, then the result is a cross-correlation between the two, and the order of the file names is important when looking at relative time shifts.

Example: Auto correlation is computed for data shown in Figure 75.
bxcr c008.seg c008.seg 0 1.2 .25
Both the auto correlation and the stack of the auto correlation are shown in Figure 85. The stack presents an average of the auto correlations at each offset. In (A) of Figure 85 we see that the near offset data (on left of the figure) present a broader bandwidth than at the further offsets. The spectral computation of the stack will provide an average spectrum, while spectral computations of the simple auto correlation in (A) will show the change in bandwidth with offset.

The zero lag sample is at the middle time. In this case, sample time of 125 msec. corresponds to zero lag (125 msec is 1/2 of 250 msec). The command above took all 1.2 seconds of data and computed the auto correlation out to $\pm 125 $ msec.

To compute an all pole spectrum, see OCTAVE YULEWALKER in section 6.0.7.

Figure 85: BXCR: (A) Auto correlation of data shown in Figure 75 (B) Stack of the auto correlation (all traces replicas of the stack result).
\includegraphics[scale=.9]{Figurebxcr.pdf}