Trackline
    Comparison plots of PostSurvey processing and Sioseis processing and the description.
    This large (9MB) plot is after some parameter "tweaking" and shows
the high spatial resolution of multi-ping SBP (sub-bottom profiler or chirp 
profiler).  There are 2-3 pings per second (~160 in 5 minutes).  The 
annotation along the top of the plot is time (day-of-year and gmt time).
The vertical is two-way travel time.  The plot is early in analyzing the
post-processing of .sgy recording.  It may help to enlarge the plot 
(Firefox=left click).
    Note some thin ponds (1538z to 1545z and 1550z to 1602z).  Note the 
water column at 1655z.
Initial post-processing steps
Step 1 was the usual mkenv script which makes an envelope file from a
correlate file.  (See mkKnudsen.html and other examples for the 
fundamentals of SBP processing).

Step 2 was slightly different from the "usual" mkplt.html script 
which makes a png plot file from an envelope file.  The "usual" "hangs" an 
exponential gain from the seafloor and performs a bandpass filter.
Unfortunately, Knudsen's automatic water depth picker picked the
strong, constant, event at .01 seconds on every ping.  Thus, the
plot was created using a .02s AGC window.

The plot parameters used are slightly different too.  The plot was
quite low (too gray, the water bottom was not black).  PLOT parameters:
colors .001 gray1 .002 gray2 .003 gray3 .004 gray4 .005 gray5 .006 gray6 .007 gray7 
and  def .005   were used.

The horizontal stripping is noise that was brought up to the same
level as the signal (water bottom). 

The plot at this stage is excellent for QC (quality control) purposes
and determining if and what should be processed further.  In this case,
it was decided to compare the Knudsen PostSurvey processing of .keb
files and Sioseis processing of .sgy files.  PostSurvey is the only
known program to read .keb files.

Step 3 was to eliminate the very strong instrument noise in the
first .025s by muting ( ttp 1 .025 ).  It took a bit of analysis
to realize that the early event was so strong.  All the automatic
pickers in WBT (PEAK POS  and   THRES) all picked it.  It was only
after dumping the amplitudes on several pings (PROUT SETS 0 .5) did
it become apparent that ALL horizontal events are instrument noise
and the early ones were larger than the water bottom reflection.

Step 4 is a combination of trial and error processing tries, resulting in
plot  produced by this script.