(Click on thumbnail to download full picture)

The basis of SETI@home is as follows: participants download a screensaver-like program from the SETI@home website, as well as a block of raw Project SERENDIP data collected at the Arecibo Radio Observatory in Puerto Rico. That data is be analyzed in the background, harnessing your unused computer power. Your computer sifts through the cosmic noise for patterns of artificial extra-terrestrial origin. None are seen here; just a combination of receiver noise and the natural background noise of the cosmos. N6TX Image

The SERENDIP receiver at Arecibo has a 2.5 MHz instantaneous bandwidth. For SETI@home processing, its output is parsed out into 256 sub-bands each 9765 Hz wide. Notice how the amplitude of the noise rolls off at both the top and the bottom frequency ends of this analysis spectrogram. At first the curve appeared to follow the frequency response of an audio bandpass filter optimized to the desired sub-channel bandwidth. But member David Wolley observes, "The roll off at the edges is almost certainly because it is from sub-band 0, i.e. the centre of the passband. This is notched out because the receiver is quadrature detecting and this region corresponds to DC. The fact that is at both edges is either because of the chirp compensation, or because the work units are actually slightly rotated in frequency." N6TX Image

Strong, coherent signals such as this one quicken the pulse of many a SETI@home project participant. Unfortunately, all so far have been generated not by ETI, but rather by terrestrial inteference, or by the Wizards of Arecibo as they inject test signals to verify the proper operation of their equipment. N6TX Image

Here's a signal with a weak Gaussian fit, which is not evident from viewing just the 3D spectrogram (bottom window).The SETI@home client divides the 9765 Hz wide data block into thousands of very narrow bins. The amplitude of the signal in each individual bin is analyzed over time, and the bin with the best fit to the antenna's expected drift-scan time series is displayed as a red trace in the Data Analysis (upper left hand) window. The white trace represents an ideal Gaussian curve (normal distribution) corresponding to the pattern of the Arecibo antenna. The two curves are statistically compared. The closer the fit, the more credence is given a candidate signal. This one will not be reported to Berkeley; it is out of bounds on both of the pre-qualification criteria: it is too weak and the fit is worse than 10. Of course, the Gaussian test is only one of many hurdles a signal must pass before it is considered to be of extra-terrestrial origin. N6TX Image

Here's another anomalous SETI@home signal. We know it is terrestrial interference because it persists at a nearly constant amplitude for the entire 107.4 seconds of the data sample. A true extra-terrestrial signal would rise out of the noise, and fall back into it, in just a few seconds, as the rotation of the Earth passes it across the narrow beamwidth of the Arecibo dish. Also, this example is at chirp zero, which is a reject condition, even if the time domain aspects were right; the most probable chirp is approximately, -0.15 Hz/s, assuming the transmitter is chirp compensated. Zero chirp is assumed to be terrestial. Follow this link to learn more about radio frequency interference. N6TX Image

SETI@home's 2 million users continue to see occasional anomalies such as this one, observed by our Executive Director in December 1999. Members sometimes call or email The SETI League, requesting that we check out such signals (most of which turn out to be terrestrial interference). Unfortunately, there's nothing we can do from here to analyze these detections, since all verification is performed by the SETI@home team at Berkeley CA. Be sure to uploaded your analysis files to them, and rest assured that they will indeed follow-up on all interesting candidate signals, and inform you if yours is The One. N6TX image

The typical computer takes tens of hours to fully analyze a single SETI@home data block. Occasionally, strong, wideband terrestrial interference obliterates any useful information. When that happens, the SETI@home client determines that no further analysis of that data block is possible, quickly terminates analysis of that particular file, and requests another one for analysis. SETI@home gave up on this file after just five minutes of attempted analysis. It most likely represents an example of receiver overload. N6TX image

Data dropouts in the middle of a SETI@home block, as seen here, are most likely caused by a piece of equipment being turned off briefly during an observing run. N6TX image

We're not quite sure what causes this anomaly, seem from time to time, in which all the adjacent bins in the display run together. We suspect that an extremely strong interfering signal is overloading the analog to digital converter. N6TX image

Current versions of the SETI@home client analyze data blocks for regular pulses, as well as CW Gaussians. Seen in red in the Data Analysis window (upper left hand corner) is a moderately strong, highly periodic pulsed pattern. N6TX image

Another feature which the newer SETI@home client software seeks out is pulse doublets and triplets. Here we see a regular pulse triplet (that is, three similar pulses, all evenly spaced in time), highlighted by the white index marks in red in the Data Analysis window (upper left hand corner). N6TX image

This unusual SETI@home detection shows a repeating pulse of immense amplitude and consistent regularity. Such signals are generally due to strong terrestrial interference, and are reminiscent of the high-altitude military aircraft "discovered" by Frank Drake during Project Ozma in 1960. N6TX image

Here's an odd error that can occur if you are running the SETI@home client on a very slow older computer. Note at the bottom of the Data Analysis box that the elapsed CPU time for analyzing this data packet (just below the red horizontal bar) is shown as a negative number. In fact, the client ran so slowly on a 486DX/33 computer that the number fields for hours, minutes and seconds overflowed, giving this bogus result. N6TX image

Clarence Vranish, N7DT, detected this strange phenomenon. He writes, "I discovered a very distinctive signal trough, which I would interpret as frequency specific, doppler shift, signal attenuation, in one of my work units. It goes from the upper left corner to the middle third of the front axis. Calculating the frequency shift, it would appear to be traveling at 1/3C, a very possible occurrence in our galaxy. What the meaning of a signal trough would be except for the presence of a gravitational lens or some other phenomena, I'm not sure. " N7DT image

A new open-source SETI@home client, operating under the Berkeley Open Infrastructure for Networked Computing (BOINC), went on line in 2005. On 15 December of that year, support for the classic SETI@home client was discontinued. N6TX image

This BOINC screen capture shows a strong periodic signal, most likely from a terrestrial pulsed radar source. The BOINC version of SETI@home excels at analyzing non-Gaussian signal patterns. N6TX image

This is not actually a SETI@home detection. Rather, it is a simulation of how the legendary Ohio State University "Wow!" signal of 15 August 1977 would have displayed in the SETI@home software, if it had been received at Arecibo. Clearly, the system is capable of detecting the next "Wow!" that comes along. Bruno Moretti Turri image

Click here for lots more pictures.

email the Webmaster

| Home | General | Memb Svcs | Publications | Press | Technical | Internet | Index | entire website copyright © The SETI League, Inc.; Maintained by Microcomm this page last updated 17 September 2016

Top of Page