Recording turned off after seg 3, but we kept the radar running until we were over ocean.

This segment is two pairs of rolls to +/- 15 deg... very slow.  These should be used to calibrate
the MVDR algorithm so we have optimum performance for nadir at least across this whole set of
angles.  Regular radar settings... hopefully the data is not saturated.  We were at 20000 ft.

This data should be processed like normal still (low altitude settings).