Does ET Text or Talk?

This week's poll is designed to get you thinking about how ET could communicate with us. It is often argued that SETI is so difficult that the aliens will send us the simplest conceivable signal -- a sine wave -- which tells us no more than the simple text ("I'm at the mall this location in the galaxy.") A growing number of scientists believe that ET may send us a signal containing unlimited information, like a prolonged monologue. The tradeoff is that while a sine wave is easily discovered, informative signals are more easily proven to be artificial and are more interesting. On Friday, Apr. 22, 2011 (11 AM, Pacific Time; GMT 6 PM), Gerry Harp will be giving an online colloquium (IRC chat) that describes recent research toward sensitive ways to discover informative signals.

Shameless plug: Harp's recent publication on this topic is chapter 4 in this book (


Text or Talk?

Why not both?  The artifact for easy detection and the complex message associated with it.  Remember Sagan's palimpsest? 

But it is important to remember that even a cosmic dial-tone tells us that, on average, technologies survive for a long time and therefore we can as well --- no successful detection of anything would be possible without that underlying longevity.


why not both?

Yes indeed. Very good arguments can be made for sending two signals, one that is easy to detect with little information (beacon), and another that is harder to detect but carries lots of information (message). Since power is money (so to speak), ET may choose to divide their available power for transmission between those two signals (50-50, 10-90, or similar).

It is often argued that once we discover the ET beacon, there will be sufficent political will on Earth to build a larger (10x to 100x) telescope to hunt for the information. On this point I'm not convinced. Suppose we went to the Natoinal Science Foundation and told them that we'd discovered a narrowband signal coming from a certain star. We'd say it looks like ET, and I don't doubt that some other scientists would offer different explanations that may undermine our argument to build a super telescope. Would NSF give us support for a telescope 10x or 100x larger than the GBT or EVLA (large radio telescopes in US? I don't thnk so, even in good times. In the current climate, the proposal would be dead on arrival.

This suggests that ET may decide to send a single signal that is both beacon and message. If ET has politics like ours, then they might consider that by adding more information to the signal they would increase the evidence that the signal has an artificial origin, thereby improving the chances of making a connection with the receiving civilization. 

Maybe this is not a concern since in 100 years we will probably build a super telescope anyway. I don't plan to be alive (^_^) but perhaps humanity will still be around. If, as many suspect, our civilization is hanging by a thread, then time is of the essence. ET may be know that 100 years could make a significant difference in the probability of survival of a young civilization.

speaking of talking...

The IRC chat/presentation will be held Friday April 22, 2011 at 11am PDT. Please synchronize your watches.

The presentation will be short, with lots of time for discussion in between. For the presenter, this kind of presentation is an experiment and I hope we'll have fun. The discussion will surround the Book, Communication with Extra Terrestrial Intelligence edited by Douglas Vakoch, and highlight chapter 4.

I'll be updating a pdf file later today which will support the conversation. Later today, please check

Presently, there is a placeholder file with only 7 slides.

Talk to you tomorrow!


#setiquest IRC connection info

For information and links on how to connect to the #setiquest IRC channel see:

chat transcript

(11:00:26 AM) gerryharp: Hey everyone, its time to begin the presentation -- or conversation
(11:00:43 AM) Jill: where are the slides to be found?
(11:00:47 AM) gerryharp: I have uploaded a refreshed version of the slides at -- these contain page numbers
(11:00:54 AM) sigblips: Aloha Gerry.
(11:01:16 AM) gerryharp: Ahayoo Gozaimas!
(11:01:50 AM) gerryharp: I'll be looking at the same slides as everyone. The format here is to be informal
(11:02:13 AM) gerryharp: Feel free to bust in and ask questions, or tell me I'm full of it. I hope for a lively discussion!
(11:02:35 AM) gerryharp: When a few of you have responded that you have the slides open to look at we
(11:02:38 AM) gerryharp: will begin
(11:03:31 AM) gerryharp: FWIW, some of this material was presented in a talk at the SETI institute in 2009, I think
(11:03:44 AM) Jill: looking at slides
(11:03:48 AM) gerryharp: If you want to see more information that is a good place to look.
(11:03:55 AM) sigblips: The .pdf loaded up fine for me and I'm reading.
(11:04:34 AM) Jill: since i can read faster than i can type, this is going to be a challenging format
(11:04:55 AM) gerryharp: Right!
(11:05:19 AM) gerryharp: So starting from the title slide, I'm not going to to read every slide or go over every point.
(11:05:35 AM) gerryharp: Some slides are just FYI in case you want to come back
(11:06:02 AM) gerryharp: Today I'm talking about some work Rob Ackermann and I have been spearheading at the SETI institute to look for new kinds of signals.
(11:06:38 AM) gerryharp: Another part of the project is to capture raw data from the telescope to allow off-line processing with new algorithms (not necessarily real time)
(11:07:07 AM) gerryharp: SetiQuest now has ~10 TB of data available, of which maybe 1/2 is uploaded?
(11:07:56 AM) gerryharp: Some of the data is just plain bad -- we made a mistake in the experiment. SO we won't post that. We have a slow link to amazon, so we'll continue to post data for a few more months until it is all uploaded.
(11:08:10 AM) sigblips: Wow, that is a lot of data. I think I've only downloaded about 500 GB of it so far.
(11:09:14 AM) gerryharp: Right? We've had a chance, at the institute to look at maybe 50% of the data in one way or another. Rob Ackermann has been generating waterfall plots in tradiational seti style, and other very interesting plots (later)
(11:09:20 AM) Jill: that's the reason we try to process in real time - datarate of 100TB/day is huge.
(11:09:25 AM) c0unt_zer0: The amount of data really impressive
(11:09:52 AM) gerryharp: With two interns, we've also been analyzing the data for autocorrelation signal types, which is the main focus today.
(11:10:12 AM) gerryharp: SLIDE 2, please advance the film strip...
(11:11:10 AM) gerryharp: Morrison and Cocconi suggested contact with extra terrestrial intelligence (ETI) in 1959, and the first realization of this experiment was by our famous scientist, Franke Drake in 1960
(11:11:52 AM) gerryharp: Very shortly thereafter, Schwartz and Towns suggested using optical radiation to communicate across interstellar distances
(11:12:23 AM) gerryharp: Unlike radio, the optical seti search didn't start right away -- came in 2000's
(11:13:07 AM) Jill: fast photon counters weren't available until then
(11:13:12 AM) gerryharp: BTW, we are referencing a book that just came out from the SETI Institute, "Communication with ETI" by Doug Vakoch, available on amazon. Take a look!
(11:13:32 AM) Jill: actually the military had them, but they weren't affordable for SETI
(11:13:55 AM) gerryharp: SLIDE 3
(11:14:26 AM) gerryharp: In the referenced book, Jill writes a fine introduction that motivates the SETI search in terms of today's science
(11:14:55 AM) gerryharp: To young people it may seem like forever, but its been only a decade since extrasolar planets were found.
(11:15:12 AM) gerryharp: Prior to that, no one was sure if _any_ other stars had planets.
(11:15:34 AM) gerryharp: This was a big boost to SETI -- now we were more confident that life would be found
(11:16:01 AM) gerryharp: Over roughtly the same period, the science of extremophiles came into being
(11:16:41 AM) gerryharp: Humans have now discovered life in many places once thought impossible. This bodes well for life on other planets -- even if the planet isn't absolutely perfect. Pause...
(11:17:22 AM) gerryharp: SLIDE 4
(11:17:48 AM) gerryharp: The two favored methos for SETI are optical and radio searches. My involvement so far has been with radio.
(11:18:01 AM) Pranav_rcmas left the room (quit: Quit: Leaving).
(11:18:33 AM) gerryharp: Radio SETI is challenging because space isn't empty. Although the vacuum obtained in interstellar space is better than any vacuum ever obtained on Earth, there is still material in there.
(11:19:07 AM) gerryharp: Star light causes ionizatoin of this rare gas between stars, releasing free electrons.
(11:20:02 AM) gerryharp: These free electrons impede the propagation (travel) through space in a complex way. The strongest affect is that high frequency light (radio waves) travels faster than low frequency light (radio waves)
(11:20:20 AM) gerryharp: This leads many to ask -- how can light travel faster than the speed of light?!
(11:20:32 AM) gerryharp: Anyone bothered by this?
(11:20:46 AM) sigblips: Not me.
(11:21:20 AM) gerryharp: Well, the answer to this conundrum is that light in a thin plasma (in thin gas of electrons) is not just light
(11:22:09 AM) gerryharp: The light field is coupled to the electron field (if you will) creating a field that is somewhere between light and matter. The light picks up some (effective) rest mass, and acts more like a very light particle of matter.
(11:22:27 AM) gerryharp: I think this is pretty cool.
(11:22:54 AM) gerryharp: But it is bad news for sending signals via, say, pulses from distant stars to us.
(11:23:16 AM) gerryharp: The graphs at the bottom show a pulse emitted from 4 LY away and then the same pulse as received on earth
(11:24:01 AM) gerryharp: As you can see, even for this very short distance (in terms of distance to other stars in galaxy), the pulse loses 30% of its peak power, and spreads over 2 microseconds.
(11:24:06 AM) gerryharp: Now it is much harder to detect.
(11:24:20 AM) gerryharp: Pause for comments.
(11:24:52 AM) gerryharp: SLIDE 5
(11:25:16 AM) sigblips: How do we know that dispersion is uniform?
(11:25:40 AM) gerryharp: Good question! It is known to be not uniform.
(11:26:05 AM) gerryharp: Very not uniform. We really can't predict the dispersion for a random star
(11:26:32 AM) Jill: circularly polarized sine waves were favored in Cyclops Report specifically because they are unmodified by ISM
(11:26:59 AM) gerryharp: There is an entire industry of mapping the dispersion in the galaxy using pulsars. However, there aren't enough pulsars to make anything like a predictive map
(11:27:13 AM) Jill: and some information can be embedded by polarization modulation of sine wave
(11:27:14 AM) sigblips: What about mild non-uniformity over short timescales?
(11:27:37 AM) gerryharp: Sigblips: yes, that is another problem.
(11:28:10 AM) gerryharp: I've been acting like the ISM is uniform in time, but it isn't. So sources scintillate, some fast, some slowly
(11:28:48 AM) gerryharp: Jill raises a good point that a lot of this science was worked out decades ago.
(11:29:18 AM) Jill: there's a whole group of radioastronomers who study intraday variable sources to set limits on sizes of radiators
(11:29:30 AM) gerryharp: On slide 5, we mention that the one kind of signal that gets through the ISM with minimal distortion is a pure sine wave
(11:30:21 AM) gerryharp: For this and other reasons (below), radio SETI has mostly focused on narrowband (nearly perfect sine wave) searches
(11:30:41 AM) gerryharp: SLIDE 6
(11:31:02 AM) gerryharp: Here I make some opinionated comments about how we do the SETI we can
(11:31:31 AM) gerryharp: Different kinds of SETI searches compete with one another for telescope time
(11:31:54 AM) gerryharp: If you have a search algorithm that is _faster_ than others, you'll get the telescope time.
(11:32:07 AM) gerryharp: This isn't a hard and fast rule, but it is an important factor
(11:32:36 AM) gerryharp: Since sine-wave searches are the fastest by some measure, almost all searches are for sine waves
(11:33:42 AM) gerryharp: However, nowadays we realized that there are better methods to communicate over long distances that include error correction via redundancy built in to the signal (like satelltie comms)
(11:33:58 AM) gerryharp: Why not use a detection method that discovers these kind of signals?
(11:34:10 AM) gerryharp: Pause...
(11:34:39 AM) gerryharp: SLIDE 7
(11:35:06 AM) gerryharp: The sine-wave search is extremely efficient thanks to hot computers and a special algorithm called teh FFT
(11:35:44 AM) gerryharp: The Fast Fourier Transform is a unique algorithm that tells you which sine waves are strongly present in your signal.
(11:36:21 AM) gerryharp: Sine waves are just one of an infinitude of "bases" in which to express our measured signals
(11:36:41 AM) gerryharp: So why not use another basis? Because we don't have a super-fast algorithm to do it!
(11:37:15 AM) gerryharp: IN the past couple of years, we're proposing to look for signals similar to those used in satellite communication
(11:37:42 AM) gerryharp: We can perform these searches almost as fast as an FFT search (factor of 2x)
(11:38:19 AM) gerryharp: Admittedly, these searches are not as sensitive as the sine-wave search, but this can be made up for in other ways
(11:38:41 AM) gerryharp: One important factor is that satellite commmunication transmits nonzero information (more below)
(11:38:45 AM) gerryharp: Pause...
(11:39:28 AM) gerryharp: This idea that the reasons behind the way we do SETI are more about technology than about science, its kind of a new idea. Anyone object?
(11:40:56 AM) gerryharp: SLIDES 7-8: We go over the simple communicatino format for GPS and how a sine wave carrier is phase modulated to generate teh GPS siganl received by everyones GPS receiver in their car.
(11:41:15 AM) Jill: the way we do SETI is and will continue to be limited by the technologies we know, and we reserve the right to get smarter
(11:41:39 AM) Avinash [~aagrawal@] entered the room.
(11:42:01 AM) gerryharp: Yes, I agree. Even the methods proposed here are somewhat driven by technology.
(11:42:14 AM) gerryharp: SLIDE 10
(11:42:58 AM) gerryharp: The GPS satellite sends a highly redundant signal. For every 1 bit of information in the signal, 20,000 bits of information are sent.
(11:43:40 AM) gerryharp: This redundancy is especially obvious in that the Gold Codesa are sent 20 times per info bit
(11:43:57 AM) gerryharp: We can design a detector that locks in on the Gold Codes
(11:44:28 AM) gerryharp: Since we won't know the GOld Codes used by ETI, we need to build an algorithm that will detect them even when we don't know what they are.
(11:44:37 AM) gerryharp: All we know (assume) is that they repeat.
(11:45:01 AM) gerryharp: For this kind of signal, a simple Autocorrelation method can pick up on the repeating Gold Codes
(11:45:52 AM) gerryharp: Before proceeding, we mention that there are other ideas of how ETI might make a signal. They could pulse a natural maser with a very strong pump maser. Causing the
(11:46:22 AM) gerryharp: natural maser to get brigher / dimmer with time. In this way they can send bits (bright = 1, dim = 0)
(11:46:44 AM) gerryharp: In this case, the maser is amplitude modulated. We'll come back to detecting this signal later.
(11:46:48 AM) gerryharp: Pause....
(11:47:25 AM) sigblips: Have you looked at the GPS 1.2276 GHz L2 signal which uses a much longer gold code?
(11:48:06 AM) gerryharp: No, we haven't tried that. The repeat time is very long as you say.
(11:48:12 AM) gerryharp: SLIDES 11 goes over a simple tutorial of how AC works. For more info I've made a recent posting on the setiQuest Algorithms forum.
(11:48:49 AM) gerryharp: SLIDE 12 shows the result of applying AC to a signal received from GPS with our telescope. Clearly, the technique works.
(11:49:24 AM) gerryharp: The GPS repeat time (spacing between AC peaks) is very close to 1 ms.
(11:49:59 AM) gerryharp: If this were a signal from ET, we'd probabloy say that a signal like this is suspiciously human-like, and probably RFI.
(11:50:35 AM) gerryharp: Although it would be important to look at many stars to see if the signal is local to Earth or really coming from just one point on the celestial sphere
(11:50:49 AM) gerryharp: Pause...
(11:51:40 AM) gerryharp: So far, I've seen responses from sigBlips, Jill and Avinash is here. I think saksam is alive. Anyone else?
(11:51:58 AM) khrm: Yes
(11:52:11 AM) gerryharp: Oh great! Probably Rob, too.
(11:52:20 AM) robackrman: rob, this is a good talk Gerry, I am enjoying it
(11:52:44 AM) c0unt_zer0: yes, i'm here
(11:52:52 AM) bobbens: also alive
(11:53:01 AM) gerryharp: (Thanks everyone!) On slide 13, we take the GPS example and try to generalize it and maybe improve it.
(11:53:53 AM) gerryharp: The GPS has a redundancy of 20,000 : 1. THat is a lot! Maybe that is what you need, but then again, the AC of GPS has many many peaks. In a sense, this is more than we need.
(11:54:07 AM) Jill: you always need to be able to sample fast enough to capture the whole bandwidth
(11:54:18 AM) gerryharp: Yes and no.
(11:54:38 AM) gerryharp: Even if you don't sample the whole bandwidth, you will see a feature in the AC spectrum.
(11:54:51 AM) Jill: but you don't get the embedded info
(11:54:54 AM) gerryharp: But if you don't sample the whole bandwidth, it will be difficult to recover the information
(11:55:05 AM) gerryharp: Right. Also, you lose some sensitivity.
(11:55:28 AM) gerryharp: So in Slide 13 we propose to send only one repeat of the signal
(11:55:46 AM) gerryharp: In this case, there is no limit on how long the signal can be.
(11:55:52 AM) gerryharp: Our redundancy is only 2:1
(11:56:23 AM) gerryharp: And the longer we integrate, the stronger our recieved signal appears.
(11:56:46 AM) gerryharp: SLIDE 14
(11:57:08 AM) gerryharp: Here we show a mock-up of a very complex information containing signal (modelled with Gaussian white noise)
(11:57:25 AM) gerryharp: We arrange to send the signal twice (using same antenna if we like)
(11:57:42 AM) gerryharp: The second copy is sent with a delay relative to the first. We randomly chose a delay of 16 samples
(11:58:01 AM) gerryharp: The receiveing telescope sees a superposition of the two signals
(11:58:08 AM) Jill: time variation of the ISM can/will impact AC detector
(11:58:30 AM) gerryharp: Ah yes. This limits the maximum delay time between signals.
(11:59:11 AM) gerryharp: However, if the signals are sent with short-ish delay (like 1 microsec to 1 s), then changes in ISM are tolerably small to be negligible (unless transmitter is VERY far away)
(11:59:44 AM) gerryharp: SLIDE 15 shows how the AC algorithm picks up such a signal.
(11:59:55 AM) gerryharp: We used only 1024 samples in this graph.
(12:00:06 PM) gerryharp: In terms of setiQuest data, that is only 0.1 ms.
(12:00:25 PM) gerryharp: If we integrated for much longer, like 10 minutes, lets see....
(12:00:59 PM) gerryharp: we'd have a sensitivity more than 2000x larger.
(12:01:18 PM) gerryharp: By sensitivity I mean, the signal to noise ratio in the graph.
(12:01:34 PM) Avinash is now known as setiquest
(12:01:47 PM) setiquest is now known as SetiInstitute
(12:02:02 PM) gerryharp: Hmmm? Changing your nickname?
(12:02:13 PM) gerryharp: SLIDE 16
(12:02:18 PM) sigblips: An important thing to point out is that autocorrelation of the GPS signal is picking out the periodicity of the gold code BPSK modulated with the header bytes and the slowly changing time stamps. Randomize either the gold code (L2 case) or the data and the detection difficultly increases greatly.
(12:02:55 PM) gerryharp: Well, yes.
(12:03:30 PM) gerryharp: If the data being sent in GPS case is very close to random, we don't get the same increase in sensitivity with longer integrations.
(12:04:01 PM) gerryharp: This is a problem. Ian Morrison has addressed this problem with a new technique called symbol-wise AC.
(12:04:08 PM) gerryharp: Beyond the scope of this talk.
(12:04:46 PM) gerryharp: But Morrison's SWAC changes the algorithm to allow gold codes to add up even if the data they represent is random
(12:05:16 PM) gerryharp: By comarison, Morrison's method only works for a subset of signals as compare to regular AC
(12:05:32 PM) SetiInstitute is now known as SetiInstitute|Av
(12:05:47 PM) gerryharp: I'm actually very excited about SWAC. The biggest challenge is that SWAC is not a "fast" algorithm.
(12:06:17 PM) gerryharp: So far he (we) have not found a way to give it N log(N) scaling
(12:06:17 PM) sigblips: Has anything about SWAC been posted online yet?
(12:06:22 PM) gerryharp: Doesn't mean it cant be done
(12:06:27 PM) gerryharp: Not online.
(12:06:52 PM) gerryharp: I asked Ian (Morrison) if I could post a preprint, but he'd rather wait until it is published. He's close to publication now.
(12:07:21 PM) Jill: Ian wants to keep some of it from publication until he has presented it as part of his PhD defense
(12:07:23 PM) gerryharp: One day I'll try to post more about SWAC on the algorithms forum
(12:07:43 PM) gerryharp: yeah, its for Ian to decide -- its his work
(12:08:26 PM) gerryharp: Getting back to AC, it has some advantages over conventional SETI and some disadvantages. It is a fast algorithm. So why don't we give it a try?
(12:08:46 PM) Jill: why don't we?
(12:09:10 PM) gerryharp: Slide 17, no number visible, shows the first signal we discovered at ATA with regular AC
(12:09:34 PM) gerryharp: This very interesting signal comes up at a non-round repetition rate of 1267.5 microsec
(12:10:03 PM) gerryharp: The plot is compressed (note ms scale) so you can see many many peaks related to this signal.
(12:10:35 PM) gerryharp: We were initially pretty excited, and spent substantial time trying to figure out where this signal is arriving from
(12:11:01 PM) gerryharp: SLIDE 18: This shows the experiment that put the stake through its heart. (!)
(12:11:51 PM) gerryharp: We had already noticed that the signal was earth-bound (hence RFI) since it appeared strongest at a particular Azimuth and Elevation no matter what time of day.
(12:12:18 PM) gerryharp: We broke our interferometer (ATA) into 3 sub arrays, and observed with each (north block, mid block, south block)
(12:12:43 PM) gerryharp: We notice that the signal appears very strong in teh north block, and only weakly in others
(12:13:07 PM) gerryharp: This means that not only is the source Earth-bound, but it is close to the north end of our array
(12:13:36 PM) gerryharp: We haven't precisely identified the source. These experiments show how you can tell if a signal is ETI or not.
(12:13:42 PM) gerryharp: Pause...
(12:13:56 PM) gerryharp: Anyoen care that we are over time?
(12:14:11 PM) gerryharp: If you gotta go, feel free!
(12:14:55 PM) gerryharp: For regular AC, we've discovered ~10 very interesting signals so far. They've all proven to be RFI, so far.
(12:15:09 PM) saksham: maybe some of the things discussed here could be added to the slides
(12:15:26 PM) gerryharp: However, for the amount of data that we have looked through, this is a very low RFI existence rate
(12:15:48 PM) gerryharp: Saksham: YOu mean, like update the slides for later? I'll see what I can do.
(12:16:34 PM) gerryharp: In SLIDE 19, we go back to the idea of an interstellar Amplitude Modulated signal
(12:17:06 PM) gerryharp: Suppose ETI uses a pump maser to brighten/dim a gigantic natural maser
(12:17:27 PM) gerryharp: We're showing a spectrum of a giant natural maswer in 19, taken at ATA
(12:17:30 PM) Jill: idea of using maser as free amplifier in space for ET signal goes back to Tommy Gold
(12:17:57 PM) gerryharp: OK, I guess I knew that, but it fell off the slide.
(12:18:37 PM) gerryharp: Now, regular AC is not sensitive to pure amplitude modulation, so we need to modify the technique
(12:19:03 PM) gerryharp: Instead of computing AC of the raw voltage signal, we square the voltage to get electric field power
(12:19:24 PM) gerryharp: That means, for every input sample, we square the sample, and make a new time stream of power samples versus time
(12:19:34 PM) gerryharp: The we apply the same AC rountes as before.
(12:19:52 PM) c0unt_zer0: Sorry, in Siberia already 2 hours after midnight. Good by
(12:19:54 PM) gerryharp: We've applied this approach to several masers (more to be done in setiQuest data)
(12:20:00 PM) gerryharp: Bye!
(12:20:20 PM) c0unt_zer0 left the room (quit: Quit: Ухожу я от вас).
(12:20:43 PM) gerryharp: We didn't see interesting signals on the masers yet, but we're applying the same approach to all our data
(12:21:37 PM) gerryharp: SLIDE 20 (sorry for poor reproduction -- Acrobat or Powerpoint failure?) shows a "power" AC spectrum for data taken in the direction of the Galactic North POle
(12:22:22 PM) gerryharp: We see lots of stuff in this spectrum. The most noticable item is a comb of peaks separated by almost exactly 10 ms.
(12:22:47 PM) gerryharp: This is just the sort of thing we were hoping for, but you can tell right away that this is human-generated
(12:23:18 PM) sigblips: Are there any satellites up there at the galactic north pole?
(12:23:27 PM) gerryharp: Because ETI won't use the same definition of "second" as we do, it is extremely unlikely that they would send a "round number" like this by accident
(12:23:34 PM) gerryharp: No satellites
(12:23:46 PM) gerryharp: Remember, all satellites must orbit to stay up.
(12:24:09 PM) gerryharp: Satellites follow (almost) great circle paths around Earth
(12:24:41 PM) gerryharp: The only great circle paths that appear stationary on the ground are so-called Geostationary orbits, which are always on the equator
(12:24:52 PM) gerryharp: Thank you, Arthur C. Clarke
(12:25:42 PM) gerryharp: However, you make a good point. It is possible that a satellite could appear in our beam by accident, no matter where we look
(12:26:16 PM) gerryharp: Since we don't see this 10 ms signal in most of our data, it probably is a satellite.
(12:26:26 PM) Jill: along with the galactic center, imaginary line extending up and down towards galactic poles have been considered 'magic places' but a lot of station keeping is required
(12:27:14 PM) gerryharp: Rob -- I'm remembering that you showed this signal appeared only in part of the observation on Gal N Pole?
(12:27:59 PM) gerryharp: Now for some fun. Slides 21-23 show samples of other signals found in setiQuest data.
(12:28:52 PM) gerryharp: SLIDE 21 - in recent observation of a star found to have a planet by Kepler project (Kepler object of interest)
(12:29:19 PM) robackrman: yes, it was transient -- occupied approx. 1/9 of observation time
(12:29:34 PM) gerryharp: Rob pointed out that conventional SETI (sine wave search) shows up a whole lot of signals. Lots and lots.
(12:30:23 PM) SetiInstitute|Av is now known as Avinash
(12:30:31 PM) gerryharp: The graphs in SLIDE 21 are at a frequency where we believe we'll see RFI, based on previous observations
(12:30:40 PM) gerryharp: That certainly appears to be true.
(12:31:13 PM) gerryharp: Slide 22 is a snippet of a conversation about what sorts of funky signals we see in regular SETI searches
(12:31:29 PM) gerryharp: At top we see some so called "Squiggles"
(12:31:36 PM) Jill: these are the parts of the spectrum that we want citizen scientists to help us work through
(12:32:12 PM) saksham: what sort of objects will emit pulse signals?
(12:32:18 PM) gerryharp: There is a sketch of what pulsed signals look like, we see them sometimes
(12:32:24 PM) gerryharp: Good question.
(12:32:35 PM) Jill: note that slide 21 has conventional power spectra not AC plots or waterfalls and covers 8 MHz
(12:32:46 PM) robackrman: pulsars :-)
(12:32:50 PM) gerryharp: The signals have to be narrow bandwidth to be seen
(12:33:01 PM) Jill: radars give pulses of different types
(12:33:04 PM) gerryharp: (to Rob) at least in the plots shown here
(12:33:28 PM) robackrman: right
(12:33:31 PM) Jill: an ET signal that is polarization modulated could look pulsed in one sense of polarization
(12:33:37 PM) gerryharp: We're talking about, say, a radar that sends a continuous tone and chops it on/off regularly (pulses)
(12:34:02 PM) gerryharp: Keep the comments coming!
(12:34:18 PM) saksham: Don't we have a lot of pulsar data?
(12:34:29 PM) gerryharp: At the bottom we see some "relatively wide bandwidth signals"
(12:34:50 PM) gerryharp: A lot? Yeah, we have sevearl observations on pulsars (dozens?)
(12:34:57 PM) gerryharp: Each one is ~20 GB
(12:35:26 PM) gerryharp: These are pulsars known to be some of the brightest visible in the galaxy, so we have hopes of seeing them.
(12:35:49 PM) gerryharp: Rob has "folded" some of these data and seen the pulsar signals.
(12:36:17 PM) gerryharp: In the last SLIDE (YAY! Thanks Johnny)...
(12:36:36 PM) gerryharp: Rob shows a result from an observation of the Crab pulsar
(12:36:57 PM) sigblips: That is some dispersion (DM)!
(12:37:00 PM) gerryharp: The Crab is really unusual. Its a pulsar, but sometimes the pulses are very very bright, other times more ordinary
(12:37:35 PM) gerryharp: Yes, sigblips has already figured it out. This is a plot of one of those giant pulses
(12:38:12 PM) gerryharp: There is just a short pulse containing many frequencies. But thanks to the ISM, the high frequencies arrive before the low frequencies
(12:38:31 PM) gerryharp: This means that in any given short sample, you see a pulse but only in a small frqeuency range
(12:38:46 PM) gerryharp: In the next short sample, the pulse moves slightly to lower frequency range.
(12:39:17 PM) gerryharp: This process gives you a slanted line in the time/frequency plot a.k.a. waterfall plot
(12:39:36 PM) gerryharp: Like sigblips, I was pretty impressed.
(12:40:05 PM) bobbens: is it always a line or can it be more non-linear?
(12:40:05 PM) gerryharp: If there were many giant pulses, separated by a long time between pulses, what would the waterfall look like?
(12:40:28 PM) gerryharp: bobbens: in the physical model it is nonlinear
(12:40:32 PM) saksham: By ordinary pulses, you mean the tiny regions that don't make the line continous?
(12:40:48 PM) gerryharp: the line is curved like (1/frequency)^3
(12:40:57 PM) bobbens: ah ok
(12:41:10 PM) sigblips: It would look the same because of the folding but weaker in amplitude.
(12:41:27 PM) gerryharp: saksham: by ordinary pulses, I mean they are weak and don't show up at all
(12:41:49 PM) gerryharp: But if there were many giant pulses?
(12:42:52 PM) gerryharp: -- Then there would be many parallel lines in the waterfall plot. Correct me if I'm wrong
(12:43:09 PM) gerryharp: OK - how about a quick session of lessons learned?
(12:43:22 PM) gerryharp: Was this style of presentation very boring to watch?
(12:44:03 PM) gerryharp: With only 20 slides, it took ~1.5 hours. In a regular talk, it might have been 30 minutes...
(12:44:27 PM) Jill: i found it impossible not to read ahead - the limitations of typing speed. you did a great job.
(12:44:36 PM) gerryharp: Ah, thanks!
(12:45:04 PM) bobbens: by being more slow it lets you digest stuff a bit more
(12:45:09 PM) gerryharp: I had thought of pre-typing some of my remarks, but didn't. It wouldn't have allowed me to be as responsive to questions.
(12:45:10 PM) bobbens: and read back and ahead :)
(12:45:31 PM) gerryharp: Hmm. that's nice. I want to try being on the receiving end
(12:45:42 PM) robackrman: thank you gerry
(12:45:48 PM) gerryharp: An obvious point about IRC -- multiple conversations at once.
(12:45:50 PM) sigblips: Thanks Gerry. I will post this transcript as a comment to your setiquest blog post.
(12:45:56 PM) saksham: that was very informative
(12:46:08 PM) robackrman left the room (quit: Quit: Leaving).
(12:46:09 PM) gerryharp: Thanks everyone! You were a great audience!
(12:46:24 PM) bobbens: thank you for taking your time to give a presentation to us on irc :)
(12:46:30 PM) bobbens: never had experienced that before :)
(12:46:39 PM) khrm: Thanks....
(12:46:45 PM) gerryharp: :-D
(12:47:36 PM) gerryharp left the room.

IRC integrated with PPT slides


For context, I've taken Sigblips posting of the IRC chat and integrated it into the notes sections of the original power point slides and uploaded the combination to