Receptions
MPL
The MPL data was recorded into
files in SIOPs format (MPL has a very nice software library built around
a consistent, self-describing data file format). This format consists of
a header, whose length is the same as all the "records" that follow it.
Each "record" contains a block of samples, all from a single channel. The
"records" cycle through all the channels, then start over at the first
channel. During this experiment, all the time series data was recorded in
16 bit fixed integer (two's complement) format.
The MPL data is organized into files containing 10 minutes of time series
sampled at 50000 Hz on 16 channel. There is no gap between each 10 minute
file, although the timestamp in the name might suggest it. The clock used
to sample the data was undoubtedly more accurate than the the clock used
to timestamp the data (please correct me, MPL,
if this is incorrect).
Testbeds
The low band was recorded on
two channels at a sample rate of 48000 Hz. The data is sample-multiplexed
(i.e. the order is: samp 1 from chan 1, samp 1 from chan 2, samp 2 from
chan 1, etc). The low band data files have an extension of *.2CH.
The mid band was recorded on a single channel at a sample rate of 96000
Hz. The mid band files have an extension of *.1CH.
The high band was recorded on a single channel at a sample rate of 144000
Hz. The high band data files have an extension of *.5CH.
The data files are either in 5 minute intervals during the Communications
schedules, or1 minute intervals during the Probes intervals.
The testbed data can be opened directly in CoolEdit as a PCM file (this
format has no header, just straight time series data in 16 bit two's complement).
UDel
This data was recorded in roughly
four minute intervals. There may be gaps between the four minute files.
UW