Given a Syqwest sub-bottom profiler file: 20120815170627R-LF.seg
No other information was given.

For general information about processing Sub-Bottom Profiler (chirp) data see:
October 2006 Sea Technology article pg1 pg2 pg3 - "Chirp Sub-Bottom Profiler Processing - A Review"
Also see:  
Notes - Paul's review of sub-bottom profilers.

Analysis steps:
1) Just plot the file - looks like crap, so I suspected it was just correlates.
2) Plotted just a few traces magnified.  Noticed that there were negative numbers,
   so that means it was a) correlates or b) was envelopes that had been filtered.
3) Used sioseis program dutil to dump the SEGY file header.  The header is EBCDIC.
   C 1Bathy2010 - V2. 0. 1. 20  August 15, 2012  17:06:27                          
   C 2SyQwest Inc, 222 Metro Center Blvd, Warwick, RI 02886, USA                   
   C 3Reel ID:  1,      Record Mode: Raw   
4) Confirmed, the data are "raw" or "correlates".
5) get more information by using lsd:
    lsd 20120815170627R-LF.seg 1 1 1
        SHOT   TR       RP    TR ID  RANGE DELAY NSAMPS    SI   YR DAY HR MIN SEC
           1     1   10001     0  0      0   282   2000    50 2012 228 17   6  30
    shot           1  tr           1  src           0  rp       10001  tr           0  id      0  eog           0
    fold      0  range           0  relev   0.0000000      selev   0.0000000      sdep   0.0000000      wdep   219.00000    
    time switch or shot time millisecond       0
    Coordinates in seconds of arc and had scalar  -1000
    source long          -72          48   38.287445     (   262118.28     arcsecs)
    source lat.          -51           3   36.879440     (   183816.88     arcsecs)
    Correlation switch (1=no, 2=yes):      0
    Sweep frequency at start (Hz).   2750
    Sweep frequency at end (Hz).  14500
    Sweep length:      1
6)  Note to self - 50us sample interval - also high frequency chirp (2750-14500Hz)
7)  Make an envelope file with the mkenv script.
    Note the stdout of mkenv says:     data times: 0.230 to  0.439 secs.
8)  Using a slight variation of script mkplt with just procs diskin plot:
    a)  Change the vertical scale to vscale 20
    b)  Use stime .2 secs .2 to eliminate the changing deep water delay.
    c)  Use anninc 1 to annotate the GMT every second.
    d)  Add ann2 shotno to annotate the shot number at the bottom.
9)  Filter the data to eliminate the very low frequencies.  Use pass 10 2000.
    See the chirp review article.  Making an envelope is also called "basebanding".
10) Good guess at pass.  try pass 20 2000

Final plot

Other notes:
Using SIOSEIS ver 2013.2.3 (9 Apr. 2013) in mkplt caused the deep water delay to be
bad on subsequent reads due to a bug that is fixed in SIOSEIS ver 2014.1.7 (6 Feb. 2014).
Details of bug:  Since 2003(?) sioseis has set SEGY to Rev 1 whenever the trace length 
of every trace is the same so that subsequently the file can be accessed by random access.
In ver 2013.2.3 (9 Apr. 2013) diskin was changed to honor Rev 1 deep water delay scalar.
This particular file has a non-zero entry in trace header bytes 215-216, so it was used
as the scalar.  ver 2014.1.7 (6 Feb. 2014) zeroes bytes 215-216 when converting from
Rev 0 to Rev 1 in process disko.