Fall weather has finally set in Berkeley CA and the SETI team met this week at the Space Sciences Lab despite the tempestuous rainstorm outside.

Bob reported that last week we accidentally stopped replicating beta, so we have to restart it this week. Other than that everything is going fine.

Eric has been sent an Astropulse application that uses ATI graphics card and would like to release it to beta. Dan asked whether it uses the Streaming SDK or OpenCL. It uses the Streaming SDK. Eric opinedthat scientific libraries for the streaming SDK are not as readily vailable as those for CUDA. Future OpenCL libraries may remedy that.

In order to start blanking some radar, Matt is testing out his new radar blanking software on some data from 2006, and that will be some of the first data to be blanked. He may even start writing the blanking into current data. It might be a good idea to run the software on some Astropulse data. If we pick a current file that was recently finished by Astropulse, run the new blanking on it, and then compare what was thrown away in each case, it would be a good way to test out our RFI algorithm.

Josh has made a big step in the completion of the RFI detection algorithm. He currently has about five or six of them running. Bob needs to create separate DB space, but in the same database, to store this data.

The best algorithm Josh has observed so far is the one that checks two polarizations at the same time. Dan originally suggested that this was similar to just raising the threshold. But on second thought, it is actually quite different. If we’re trying to get rid of signals that come from noise, are they are not linearly polarized, and do not statistically have the same polarization. So this algorithm might actually be effective.

To investigate this further, Dan suggested that Josh measure the effectiveness of the algorithm by taking the ratio of signals that the algorithm threw away as opposed to what we know is just white noise. And so, the search for the perfect algorithm continues.

And finally, the most important question of the day: Halloween is coming up soon, so when will we be buying candy?