Melville 2002 seismics with OYO recorder and AMG streamer.

    The OYO tape output is SEG-D, not SEG-Y.  SEG-D is not for
the faint of heart;  dd'ing a SEG-D is not a good idea since
every shot is a file.
    Convert the SEG-D tape to a SEG-Y disk file.  e.g.
sioseis << eof
procs segdin diskoa prout end
segdin
   device /dev/rmt/0bn end
end
diskoa
   opath oyo_data.segy end
end
prout
   fno 0 lno 999999 noinc 100 ftr 1 ltr 1 end
end
end
eof


     Following steps similar to the AVON04 example:
>lsd oyo_data.segy 1 10
      SHOT   TR       RP    TR ID  RANGE DELAY NSAMPS    SI   YR DAY HR MIN SEC
     10273     1       0     0  1      0     0   7000  1000 2002 256  3  44  13
     10273     2       0     0  1      0     0   7000  1000 2002 256  3  44  13
     10273     3       0     0  1      0     0   7000  1000 2002 256  3  44  13
     10273     4       0     0  1      0     0   7000  1000 2002 256  3  44  13
     10274     1       0     0  1      0     0   7000  1000 2002 256  3  44  23
     10274     2       0     0  1      0     0   7000  1000 2002 256  3  44  23
     10274     3       0     0  1      0     0   7000  1000 2002 256  3  44  23
     10274     4       0     0  1      0     0   7000  1000 2002 256  3  44  23
     10275     1       0     0  1      0     0   7000  1000 2002 256  3  44  33
     10275     2       0     0  1      0     0   7000  1000 2002 256  3  44  33


    So it's 4 trace data.  Plot a few shots for QC:
sioseis << eof
procs diskin plot end
diskin
   ipath oyo_data.segy ntodo 20 end
end
plot
   vscale 1.25 srpath sunfil.ras nsecs 7
   nibs 2859 recsp yes ann shotno
   end
end
end
eof
xloadimage -r 90 sunfil.ras &



   All four traces look okay, so "vertically stack" them by using
ntrgat 4 in diskin and adding stack in the procs.  Remember that
filter should be done after stack for speed.  Finding the right plot
"def" was hard because of the noise at the beginning of the data.
plot
sioseis << eof
procs diskin stack filter plot end
diskin
   ntrgat 4
   ipath oyo_data.segy ntodo 2000 end
end
filter
   ftype 0 pass 20 150 dbdrop 48 end
end
agc
   winlen .5 end
end
plot
   vscale 1.25 srpath sunfil.ras nsecs 7
   nibs 2859 ann gmtint anninc 5
   trpin 200 def .02 wiggle 0
   end
end
end
eof
xloadimage -r 90 sunfil.ras &


    A 3 trace mix does wonders for reducing the noise in the
water column.  (plot)
sioseis << eof
procs diskin stack mix filter plot end
diskin
   ntrgat 4
   ipath oyo_data.segy ntodo 2000 end
end
filter
   ftype 0 pass 20 150 dbdrop 48 end
end
mix
   weight 1 2 1 end
end
agc
   winlen .5 end
end
plot
   vscale 1.25 srpath sunfil.ras nsecs 7
   nibs 2859 ann gmtint anninc 5
   trpin 200 def .02 wiggle 0
   end
end
end
eof
xloadimage -r 90 sunfil.ras &


    Compare the stack to just using a single trace (stack plot) (trace 4 plot)
sioseis << eof
procs diskin prout mix filter plot end
diskin
   ftr 4 ltr 4 fno 11500 lno 12000 allno no
   ipath oyo_data.segy end
end
prout
   noinc 100 fno 0 lno 999999 ftr 0 ltr 999 end
end
filter
   ftype 0 pass 20 150 dbdrop 48 end
end
mix
   weight 1 2 1 end
end
agc
   winlen .5 end
end
plot
   scalar 40
   vscale 1.25 srpath tr4.ras nsecs 7
   nibs 2859 ann gmtint anninc 5
   trpin 200 def .04 wiggle 0
   end
end
end
eof
xloadimage -r 90 tr4.ras &


    Does decon get rid of the 100mil "bubble" pulse? (plot without)  (plot with decon)
sioseis << eof
procs diskin stack mix decon filter plot end
diskin
   ntrgat 4
   fno 11500 lno 12000 allno no
   ipath oyo_data.segy end
end
decon
   sedts 4.5 5.5 end
end
filter
   ftype 0 pass 20 150 dbdrop 48 end
end
mix
   weight 1 2 1 end
end
agc
   winlen .5 end
end
plot
   vscale 1.25 srpath sunfil.ras nsecs 7
   nibs 2859 ann gmtint anninc 5
   trpin 200 def .02 wiggle 0
   end
end
end
eof
xloadimage -r 90 sunfil.ras &