We can use this topic as a clearning house for new data uploads.
There have been requests for data from a known emitter that looks as similar as possible to what we expect from ET.
Rosetta is a spacecraft that emits a medium-weak tone near 8422 MHz. It drifts around depending on the earth's position and rotation angle w.r.t. the satellite. The tone is not ALWAYS observed at ATA since we are pushing the usable frequency limit and compared to the system noise, the tone is weak. But the real problem is calibrating our beamformer at this very high frequency. So sometimes we see it, sometimes we don't.
This is the one and only observatoin of Rosetta under the setiQuest rubric (where we save the data). We have not analyzed this data for the tone. It would be a great exercise if someone could help us and find the (drifting) tone. Feel free to start a new topic if you do.
I might mention that under the "setiCode" project (Jon Richards is the heavy), the SETI Institute is making code available now or very soon to allow you to pipe setiData into our own observing software. Presumably, someone could use this engine to find the signal. Or tell us if it isn't there!
I might add, however, that a determined analyzer person could look at longer FFT-lengths than are supported by setiCode, and see signals setiCode misses. This is perhaps the only advantage of non-real-time analysis.
When "calibrating" (that is, focusing) a radio telescope, we usually look at a strong point source of radio emission and then tweak our antenna delays and phases so that we produce the best image of a point as possible.
Fortunately, very distant radio galaxies are sometimes in a process of continuous explosion. The black holes at the centers of these galaxies generate huge energy and they are so far away that they cannot be resolved with our telescope (very long baseline interferometry can resolve some, etc.). Such galaxies are also called quasars.
We chose to look directly at some of our favorite calibrators (they're all galaxies) 3c119, 3c123, 3c138, 3c147, and 3c84, at 4462 MHz. This is a known "clean" band so it should be mostly noiselike. Further, the signal from the galaxy is ~1x10^4 smaller than the noise in the telescope system.
You can see the quasar in an image, but with a single beamformer pointed directly at the source, I don't know of a way to extract the galaxy signal from the system noise. The galactic radiation for these quasars looks unmeasurably different from Gaussian White Noise except for a broad (GHz scale) dependence of amplitude on frequency. (Almost every quasar goes down in amplitude with increasing frequency.) And when you mix noise with noise, all you get is noise. (Unless you don't. There are some very rare galaxies that have variable output on scales of years to hours.)
This doesn't mean there isn't a SETI signal in there. The gaxaxy in the background is still relatively weak and would hardly influence a conventional SETI observation.
Finally, these galaxies are a good testbed to analyze our system. We expect that these data will be very close to a measurement of GWN.
Last May we followed up the original exoplanet observations of Kepler-04 with multiple re-observations. This was a day where we did much testing so not all the data were good. Out of 4 observations, we trust only the last two. These two have just been uploaded to the "Getting Data" page of the setiData website.
The "interesting feature" (drifing tone), independently discovered by several people, does not appear in these captures. This disappointment lead me to not post the data immediately and this was not a good choice. Now the data are there. We have not gone back to this source again (yet).
Looking at the metadata, I notice that the first observation was at 1418 MHz and the second observation at 1420 MHz. The orignal observation in January was at 1420 MHz. So be careful when lining up the spectra!
The "interesting feature" does appear in these captures. It is drifting, it is modulated, and it is roughly the same distance left of Hydrogen.
The next dataset posted is taken while the telescope is pointed at a geostationary satellite. The frequency was chosen t obe "in the range" of where Galaxy-19 transmits, but the band pass might no overlap with the transmitter. Our goal in this measurement was to measure in a so-called "bad" band in the middle of the 4 GHz frequency allocation for satellite TV. We call this a bad band because the regular SETI detector can be overwhelmed by the large number of signals that appaer. We have not analyzed this data yet. Although it would be difficulty to prove a SETI signal is present in this band since there are so many signals.
On the positive side, this band may contain a large number of signals with different properties. We can use the signals in this band and other observations to generate a taxonomy of signal types. Sound like fun?
ps. Notice that like the most recent Kepler data posted, these data were taken in May when we still had hydrogen sidelobes (due to periodic corruption of time series). We believe that _every_ signal will have similar sidelobes.
I've just uploaded 7 new datasets taken with the telescope pointed in the direction of pulsars. The frequency is 4462 MHz, which is higher than before. Therefore the pulsars are expected to be weaker. It will be interesting to see if it is possible to observe the pulsars at this frequency. Anyone want to fold this data?
Observing at Hi * 2 frequency (2840 MHz) here are two more quasars from July observing. We haven't analyzed this data, yet.
Lots of new data, including spanking new data on GL581:
|2010-10-01||Exoplanet 211 (4462 MHz)|
|2010-10-01||Exoplanet 219 (4462 MHz)|
|2010-10-01||Exoplanet 221 (4462 MHz)|
|2010-10-01||Exoplanet 222 (4462 MHz)|
|2010-10-01||Exoplanet 225 (4462 MHz)|
|2010-10-01||Exoplanet 226 (4462 MHz)|
|2010-10-01||Exoplanet 227 (4462 MHz)|
|2010-10-01||Exoplanet 228 (4462 MHz)|
|2010-10-01||Gliese 581 (1420 MHz)|
|2010-10-01||Gliese 581 (4462 MHz)|
|2010-10-01||Gliese 581 "Off Source" (4462 MHz)|
If you believe Willie Sutton ("Thats where the money is.") then exoplanets are the place to look.
Some of the extreme fervor over Gl581 has faded but the results from Hawaii stand on their own. It will be a while before we are sure if there is an earth-like planet over there, but the chances are better than zero which is true for most exoplanetary systems. (Or so we think, ha ha!)
Just for review, there are a few magic frequencies we visit often:
HI line - 1421 MHz
sqrt(2) * HI = 2008 MHz
2 * HI = 2840 MHz
Pi * HI = 4462 MHz
Why not 3 * HI? Because it is in the middle of a "bad band."
8200 MHz -- because it is a high frequency where some distant spacecraft make SETI-like transmissions. ALso, no one has done much SETI work in this frequency range.
Hi Again, and that is not all! Here are some more datasets recently added.
Again: The frequency tells most of the story. 3991 MHz and 4020 MHz are both in the same "bad band." Again, there are so many signals in these frequency ranges that it is beyond our capability to characterize them all. People who are interested in visualizing SETI data should look at these signals and find how many there are of different kinds, e.g. (straight liens in frequency/time waterfalls, autocorrelated signals, chirped signals (like dispersed pulses), non-dispersed pulses, measures of information content, etc. etc.
THe GPS signal is measured at, duh!, the GPS frequency. This is an excellent dataset to test algorithms that search for BPSK encoding (or related encodings). Because the signal is totally dominated by the satellites. This one will be fun to examine in several ways.
Measurements at 1420 (HI line) and 2008 ( sqrt(2) * Hi line) are more regular SETI observations.
I hope you will be as excited as I am about recent observations of the Lagrange-4 point, mentioned in the previous note.
The Lagrange point observations are particularly interesting as noted by the famous astronomer Seth Shostak. A while back there was a paper that suggested that it would be cheaper and more effective to not send EM waves but to send solid objects (artifacts) that could be read by the recipients. The point was that there is greater potential to send more information with a LOT less energy. But even if you don't believe this, no one can forget 2001 a space odessy, and how aliens left an artifact on the moon.
Well, one of the stable Lagrange points is a great place to leave an artifact. If you place an artifact near the center of the Lagrange point with low enough velocity, it will stay there. It just orbits slowly around the Lagrange point. Even though irregularities in the earth moon and planets cause the Lagrange point to change over time, the orbits are stable enough that the artifacts may stay near a Lagrange point for a very very long time.
Now suppose you put a cell phone with an appropriate energy source in the Lagrange point. It could be sending us a signal continuously and we'd never know it because we're always looking at stars! See, the L-point moves across the fixed background of the stars .So our SETI algorithms would tend to throw away Lagrange artifacts unless we track them most concienciously. AFAIK, we haven't done this yet in regular SETI observing.
Thanks to magic by Rob Ackermann, an ephemeris for the L4 point was produced and we tracked it with the telescope. No one has looked at this data yet. We will though, when we get a chance. For now, the data are yours.
[The naive among setiQuest users might think that we perform all kinds of testing before we ever release a piece of data. But it simply isn't true -- we don't have time! The weeks delay between data aquisition and data release is the same delay time between when we get a crack at the data and when all users get a crack at the data. In this mode, there is every reason to believe that citizen astronomers will find interesting signals before the in-house analysts will find them.]
Was this the Sun-Earth or the Earth-Moon L4 point? I'm guessing it's the Sun-Earth L4.
The observation was done on Sun-Earth L4.
Thanks for the Sun-Earth L4 confirmation. Here are two forum links that discuss the Lagrange-4 datasets:
Are you ready?
We've just uploaded the most recent Kepler-04 data taken in September. Sorry it takes so long to get it posted, but that is how long it takes! Everyone should thank Rob Ackermann for his tireless and unexcting work that brings all of our data to the cloud.
We have some way-cool data just posted that should be a lot of fun. We were tracking 2 GOES satellites which are weather satellites that transmit with a relatively simple phase-encoded scheme. We'll all have to look up the details.
The point is that you should be able to produce pictures of the earth with this data. And also explore the bandwidth and other properties of these very interesting signals.
Hey -- If earth were to transmit signals about itself, maybe we should transmit satellite images and spectrographs taken from satellites, huh?
Meanwhile, I notice that the meta-data on two of the files says 6672 while the other are 1672. THe latter is the correct number. I wonder if the frequency in the other files is a typo? You could prove this by showing that the same signal appears in both spectra (or not).
A special prize for the first participant to make a good quality image of the earth! Talk to Gerry to claim your prize.
It turns out that all the GOES data were indeed taken at the correct frequency: 1692 MHz. The meta-data entries of 6692 were a mistake. The meta-data entries have not been fixed yet (I think) but they will be soon.
ps. I'm serious about the special prize! It is an object(s) that might be worth as much as $15!
I took a look at some of the weather satellite data and managed to get a reasonably nice image out of the “low-rate information transmission” LRIT channel at 1691 MHz (I ran out of data before the message repeated and I was able to fill in the missing packets).
The image in the link is from GOES-13, second data set. In addition to this data stream, there are a number of other data streams which have less complicated modulation schemes for those that are interested and exploring these data sets. There is a FSK signal at 1690.725 MHz which is simply RS-232 at 9600 Baud. This signal has text based messages, don’t be frustrated if the output seems garbage at times, many of the messages are in a Japanese file format. If I remember correctly, the first couple messages in “2010-09-24-goes-13_2-8bit-01.dat” has English messages. There is also a very easy telemetry signal to decode at 1694 MHz. This is a 4 Kbps, Manchester encoded signal. I don’t have any information on how to decode the demodulated bit-stream, but it’s a nice signal to test out if you are having questions about anything.
I am going to sent that around in internal email.
What do you think is the cause of the horizontal dropouts (thin black lines)? Is that an issue with our data capture?
The horizontal lines were caused by a bug in my decoding program. The image is split into a number of different files for download. Within each file, the pixel data from each row is compressed and put into a packet, these packets are then concatenated together and split into ~1Kbyte frames for the downlink. My program was dropping packets whose header was spread across two separate frames. I fixed this bug and posted a corrected image with no missing lines.
I should also be able to get images from the GOES-11 second data set, however it appears that the first data sets for both GOES-11 and GOES-13 are just noise. This is too bad since both these satellites transmit images of the whole globe, and with the long collection times, there’s a fair chance you would have captured one.
This image looks really great. Everyone at the institute has exclaimed the same thing the first time they see it, "Awesome!"
Would you please send me your home address via a private message? I would like to personally send you a fantastic gift in congratulations for your decoding success. For our group, this is an important proof of principle, that we can use the array as a satellite downlink!
To simplify things, just send email to gharp at seti dot org.
The other two GOES-11 and GOES-13 datasets may have been recorded at the wrong frequency. There was some confusion about that which was never completely resolved. We should be able to make a long observation of GOES-11 and/or GOES-13 in the future. We also could potentially track and record a POES weather satellite.
Could you please post the 2010-05-14 PSR 0809+74 data set that Robackrman mentioned here:
I'm sorry to say we've had some difficulties with existing postings of setiQuest data. One of the Amazon instances went offline in the cloud, and certaing links to "getting data" are broken on the setiQuest webstite.
While the setiQuest website is very good for some things, we believe that it isn't the greatest method for serving up setiQuest data. So we are moving over to a new system where the data are served on the setiQuest Wiki. I'm really looking forward to this rollover, as we'll be posting A LOT more data and many datasets never before seen by our users.
We should have the new site running in a couple of weeks. Until then, if you would like to preview the new data or get a copy of old data, please send email to gharp at seti dot org, and I'll give you a link to the alpha data site. See you soon in data-land.