Each data file contains 33 million probe samples in a raw, binary format, where each sample is represented by a big-endian signed short (2-byte) integer. There is no meta data in the files, but the sample rate is known to be 10^8 samples per second.

The python script snippet load.py, included in the same tar file as the data files, demonstrates how to load a data file, perform an FFT over a one-millisecond window, and plot the result to a PNG file.

Figs. 1-6 were generated from 130608_07-10-08_1115_F-G_mid.bin

Fig. 7 was generated from 130621_07-10-08_1357_F-G_mid.bin

Fig. 8 was generated from 130597_07-10-08_0922_F-G_mid.bin

Fig. 9 was generated from 130595_07-10-08_0902_F-G_mid.bin, 130601_07-10-08_1002_F-G_mid.bin and 130610_07-10-08_1140_F-G_mid.bin