The following are screenshots of data from the live receive session we did with our Ettus Research Software Defined Radio unit attached to the Arecibo antenna today (19 May). "Waterfalls" were generated by post-processing the recorded data. There are four recordings of various lengths as we were testing the setup, and this is the very, very initial result.
The first and second screenshot images were taken during the live capture. You can see the faint diagonal line of the carrier. This is a simple flowgraph Balint Seeber put together in GNU Radio that receives samples from the USRP (which is in turn connected to the IF output from their existing S-band receiver), records the samples to disk and also displays the FFT and waterfall plots of a selectable narrow-band portion of the spectrum.
Figures 1 and 2 (images 3 and 4) are the output of a Python script Balint wrote quickly after the data capture to perform a large FFT on the data (256K points) and produce the waterfalls with numpy & matplotlib (this one capture is ~14.5 min). Figure 1 is the entire baseband spectrum captured from the USRP (250 kHz BW). You can see the carrier left of center (log color scale). Figure 2 is zoomed into that area and you can see the Doppler shift (~977 Hz BW, linear color scale). Scales on both are just the FFT bin indices.
Unfortunately the signal is a little weaker than we expected, and it's also odd that it fades out toward the end of this capture (it returns and fades in subsequent ones too). Again, this is all very preliminary data done tonight on a rush basis. Much more detail to follow.
Click on Images to enlarge
Image 1: Screenshot taken during live capture.