Normal-speed low-light-level video recordings began at 02:24:24 UT. The sferic data acquisition system started recording at 02:25:04 UT. The 3-channel sferic data was comprised of 8-bit magnetic field, 12-bit photometer, and 12-bit fast antenna electric field data sampled at 500 kHz. A bipolar-amplitude threshold on the fast antenna data was used to trigger the system. The pre- and post-trigger lengths were set at 3 and 9 ms respectively. The data were also recorded continuously at 10 kHz.
Unfortunately, the magnetic loop antenna was not oriented to the proper azimuth for an optimal signal-to-noise level for the sprite-producing +CG activity. Rather, the loop remained in an orientation used on the previous day for +CGs to the south-southwest (see Section 5.3). Since the +CGs on this day were situated to the east-southeast, they were close to the null in the loop's reception pattern. The adverse effect of the improper orientation was alleviated by the fact that the vertical electric field is proportional to the magnetic field tangential to a circle around the source at long distances (within the ``radiation zone'') and thus provides the same information. The magnetic field instrument had a better signal-to-noise ratio at low frequencies when it was properly oriented (due, perhaps, to the lack of static field contamination from local space charge which effects only the electric field). Charge moment change measurements derived from ELF magnetic field data are shown in Section 5.3 for sprite events on another day.
The photometer monitored a wide region of sky above the MCS. The
photometer was high-pass filtered above 
600 nm in order to
improve the signal-to-noise ratio for the sprites relative to the
parent discharge by capturing the dominant sprite emissions from N
first positive bands between 600-760 nm (Hampton et al., 1996; Mende et al., 1995). The signal-to-noise ratio was further improved by
positioning a horizontal slit in front of the photometer so that the
sprites would be observed through the slit while light escaping from
the thundercloud would be rejected. In spite of these precautions,
light scattered through the atmosphere from the discharge still
reached the photometer, though the scattering was greatly reduced from
what it would have been.
The high-speed video system was armed shortly after the sferic system began recording. The high-speed video system was manually triggered after sprite events were observed on normal-speed video. A total of 23 triggers were obtained with the first being at approximately 02:39:56 UT and the last at 04:50:57 UT. Of these 23 high-speed video records, just over half (12) were obtained at 2000 frames/second while the remainder were obtained at 1000 frames/second. Seven of the 23 video records did not contain sprites due to improper camera orientation, a late manual trigger, or perhaps unusually dim sprites. Of the remainder, some contained multiple sprite clusters. The total number of sprite clusters observed on October 7 was 24, which is over half of the 42 total sprite clusters observed by the high-speed camera in this work.