WOW Message

[coreybv]

I'm sure anyone with even a passing interest in SETI is familiar with the WOW message. This post isn't specifically about the message itself, but I was reading a bit about it again and it brought up some thoughts.

In particular, IF our radio SETI programs detected a genuine signal from an ET civilization, would our standards of proof permit us to acknowledge it as such?

We would want independent confirmation of the signal, from geographically distant observatories, and probably over some span of time.

Yet, the most famous of the CETI signals that we have intentionally broadcast, the Arecibo message, was less than 3 minutes in duration. One wonders, if it were to be received by a technological civilization, would they even acknowledge such a message as having come from an intelligent species? If we were to receive a similar message from an alien civilization, would we acknowledge it as such, or just write it off as a computer glitch (or whatever mundane reason you choose)?

Our standards of proof would be quite high, and rightfully so. Yet, our own efforts at transmitting messages would quite likely not meet those standards.

This seems to be a problem.

[IsaacKuo]

This problem highlights how expensive CETI would be if done right. I've thought about what sort of signal I'd send out so that it was unambiguous and informative, and the sort of signals I come up with would require a huge amount of bandwidth.

The Arecibo message is insanely terse. The scientists who invented it tried to pack in as much as they could into every little bit of data, and the result is a dot matrix cartoon. To me, it looks like they basically imitated ancient Egyptian hieroglyphics. I can imagine the unconscious thought process being--"Hey, the ancients didn't make it easy for us to comprehend their writing so why should we make it easy on the aliens?"

Same with the Pioneer and Voyager plaques--cartoon depictions.

It makes no sense to me. If you're going to send a picture, how about sending a realistic photograph or painting? Send them a greyscale photo so they can tell what we REALLY look like.

If we received an Arecibo style message and decoded it into a small matrix of on/off dots...well that'd be neat but then we'd have to puzzle over what the heck it meant, or even whether it was really artificial.

But what if we received a 256x256 greyscale picture of the Andromeda galaxy, followed by a 256x256 greyscale image of a crab-like creature? The former would be something which we both can see, to confirm the way the image is encoded. The latter is an image of the creature who sent the signal. I think there would be no way we could mistake this for anything else. But this requires a lot of bits of data--a lot more than an Arecibo style message.

[iquestor]

This problem highlights how expensive CETI would be if done right. I've thought about what sort of signal I'd send out so that it was unambiguous and informative, and the sort of signals I come up with would require a huge amount of bandwidth.

The Arecibo message is insanely terse. The scientists who invented it tried to pack in as much as they could into every little bit of data, and the result is a dot matrix cartoon. To me, it looks like they basically imitated ancient Egyptian hieroglyphics. I can imagine the unconscious thought process being--"Hey, the ancients didn't make it easy for us to comprehend their writing so why should we make it easy on the aliens?"

IIRC they used prime numbers so that the correct width and height of the data depiction could be discovered, and then hopefully some data that could give them a quick idea of who we are and where we are. Id hope its easier to decode than say Linear B or Heiroglypics, which we still wouldnt know if not for the rosetta stone.

Same with the Pioneer and Voyager plaques--cartoon depictions.

I dont think that is a serious message. As you said elsewhere, its undetectable unless you are looking for it and know what it is already. Of course, alien tech might spot it, but to me, its doubtful.

It makes no sense to me. If you're going to send a picture, how about sending a realistic photograph or painting? Send them a greyscale photo so they can tell what we REALLY look like.

lots of bandwidth, and there would need to be intructions on how to decode it as the Arecibo message had, with the added difficulty of encoding the values for greyscale coloration. I'd imagine though there is a way to do it.

If we received an Arecibo style message and decoded it into a small matrix of on/off dots...well that'd be neat but then we'd have to puzzle over what the heck it meant, or even whether it was really artificial.

The use of prime numbers and patterns would likely make it instantly recognizeable. IMHO.

But what if we received a 256x256 greyscale picture of the Andromeda galaxy, followed by a 256x256 greyscale image of a crab-like creature? The former would be something which we both can see, to confirm the way the image is encoded. The latter is an image of the creature who sent the signal. I think there would be no way we could mistake this for anything else. But this requires a lot of bits of data--a lot more than an Arecibo style message.

It would have to be decoded and rendered first though. They can't transmit a bitmap and expect our software to handle it, unless they saw Independance Day, that is.

[coreybv]

I like your proposal for the type of message to send.

Myself, I prefer Sagan's method from Contact. Just send the first dozen or so prime numbers as a series of beeps. This would be much harder to mistake for background noise than a greyscale image, and impossible to dismiss as being of natural origin.

Perhaps use a combination of the two methods. The prime number "beacon" to catch their attention and get them to point their telescopes at us. Once they're convinced that there is an intelligence here, they may be much more likely to check other frequencies and put forth the effort to decode your digital imagery.

It seems to me that so far all of our CETI efforts have been nothing but publicity stunts. And given the effort it takes to find information on them, or that they even happened, they were miserable failures even at gaining publicity.

[iquestor]

I like your proposal for the type of message to send.

Myself, I prefer Sagan's method from Contact. Just send the first dozen or so prime numbers as a series of beeps. This would be much harder to mistake for background noise than a greyscale image, and impossible to dismiss as being of natural origin.

Perhaps use a combination of the two methods. The prime number "beacon" to catch their attention and get them to point their telescopes at us. Once they're convinced that there is an intelligence here, they may be much more likely to check other frequencies and put forth the effort to decode your digital imagery.

It seems to me that so far all of our CETI efforts have been nothing but publicity stunts. And given the effort it takes to find information on them, or that they even happened, they were miserable failures even at gaining publicity.

I'd say alternate the messages: Send #1 with the Primes, then #2 with pictures. then # 3 with a frequency we are listening on. Then back to #1.

[formulaterp]

And given the effort it takes to find information on them, or that they even happened, they were miserable failures even at gaining publicity.

If you enter "ceti" into Google, the first result is a link to a rather informative Wikipedia entry, with plenty of links to followup articles.

Really not that much effort.

[Quadrazar]

An interesting problem using REAL pictures is that we can't know for sure that the ET's are using the same 'senses' as we do. If they use vision, their visual range might be limited, or might expand into the x-ray or infra-red spectrum? Sending real pictures could induce problems. The use of abstract pictures, in my opinion is a better way start communicating. (after a matthematical handshake)

For those who are interested in the search for ET, watch this lecture by Prof Ian Morison

[IsaacKuo]

An interesting problem using REAL pictures is that we can't know for sure that the ET's are using the same 'senses' as we do.

That's right, but we can make some good guesses at what sort of senses they would use to detect the Andromeda galaxy. Sonar won't work. Touch won't work. Smell won't work. The only thing that will work is something like a telescope, and it will give some sort of 2d picture rather than, say, the three dimensional data of sonar or touch.

We can't know what sort of senses the creatures use from day to day. But we can make a reasonable guess that the only way for them to point a communications dish at our star system is if they have something like telescopes. Their astronomical observation machines must have "vision" even if they themselves don't.

If they use vision, their visual range might be limited, or might expand into the x-ray or infra-red spectrum? Sending real pictures could induce problems. The use of abstract pictures, in my opinion is a better way start communicating. (after a matthematical handshake)

I favor greyscale pictures because that's got to be something that they can comprehend and manipulate. If not, then they can't have astronomy and they can't point a communications dish at us.

An abstract picture is something open to interpretation, and aliens might not be able to comprehend it.

Still, the decision of what filter to use is interesting. Should we send a picture of the Andromeda galaxy as seen in visible light? Infrared? X-ray? Radio frequencies? I say we should send a picture of the Andromeda galaxy as seen in visible light, centered around the peak frequency of our Sun. Any aliens which can detect our Sun can determine its spectrum, and guess that we would be using sensors optimized for the light it generates.

Basically, I reject the basic philosophy of the Arecibo message, which is to try and figure out how to communicate when you can't just point at something and say, "That!" Instead, I consider what sort of things are out there which we can indeed point to. Instead of starting from scratch, let's start with what we must have in common.

[iquestor]

That's right, but we can make some good guesses at what sort of senses they would use to detect the Andromeda galaxy. Sonar won't work. Touch won't work. Smell won't work. The only thing that will work is something like a telescope, and it will give some sort of 2d picture rather than, say, the three dimensional data of sonar or touch.

We can't know what sort of senses the creatures use from day to day. But we can make a reasonable guess that the only way for them to point a communications dish at our star system is if they have something like telescopes. Their astronomical observation machines must have "vision" even if they themselves don't.

I favor greyscale pictures because that's got to be something that they can comprehend and manipulate. If not, then they can't have astronomy and they can't point a communications dish at us.

An abstract picture is something open to interpretation, and aliens might not be able to comprehend it.

Still, the decision of what filter to use is interesting. Should we send a picture of the Andromeda galaxy as seen in visible light? Infrared? X-ray? Radio frequencies? I say we should send a picture of the Andromeda galaxy as seen in visible light, centered around the peak frequency of our Sun. Any aliens which can detect our Sun can determine its spectrum, and guess that we would be using sensors optimized for the light it generates.

Basically, I reject the basic philosophy of the Arecibo message, which is to try and figure out how to communicate when you can't just point at something and say, "That!" Instead, I consider what sort of things are out there which we can indeed point to. Instead of starting from scratch, let's start with what we must have in common.

Isaac, how do you propose to sending a "picture" without encoding it as a binary signal? The Arecibo message was a primitive form of this, using Prime numbers to establish the width and height, then the message bits drew a rudmentary picture in black and white.

If we wanted to send a greyscale picture of the Andromeda Galaxy, then we would have to digitize it into binary and then encode it in some format in which to send it as a radio signal. And we would have to enclose some legend in the header on how to change the stream of 1's and 0's into a coordinate system and then apply different levels of color to each point in order to arrive at a picture. We cant "just send them a picture".

I absolutely agree it would be great to show them something in the sky they can point at as a starting point, but its more complicated than that, unless they have a copy of photoshop, or can provide us an interstellar fax number.

[IsaacKuo]

I propose encoding it as a binary signal--a simple raster scan of 256x256x16. 16 bit words results in greyscale levels from 0 to 65535.

I'd start with 65536 16 bit words counting from 0 to 65535. Then, pictures would follow. Every 16th "picture" would be the 0 to 65535 count instead of a greyscale picture.

I favor using powers of two instead of prime numbers because it's almost certain that anyone capable of receiving the signal is using binary computing technology. The aliens may or may not care about the mathematics of prime numbers, but if they've got computers they almost certainly care about binary numbers and powers of two.

If they can't recognize binary numbers counting from 0 to 65535, then there's no hope of us sending any sort of message to them.

Once they recognize this, then it should be obvious that 65536 is a perfect square. Assuming they have astronomy and digital electronics, it's highly likely that they have telescopes with square sensors.

There's no technical reason to develop telescope sensors with a prime number of rows/columns, but there are good technical reasons to develop sensors with a power of two for the rows/columns. Binary digital computing technology "likes" powers of two.

[iquestor]

I propose encoding it as a binary signal--a simple raster scan of 256x256x16. 16 bit words results in greyscale levels from 0 to 65535.

I'd start with 65536 16 bit words counting from 0 to 65535. Then, pictures would follow. Every 16th "picture" would be the 0 to 65535 count instead of a greyscale picture.

I favor using powers of two instead of prime numbers because it's almost certain that anyone capable of receiving the signal is using binary computing technology. The aliens may or may not care about the mathematics of prime numbers, but if they've got computers they almost certainly care about binary numbers and powers of two.

If they can't recognize binary numbers counting from 0 to 65535, then there's no hope of us sending any sort of message to them.

Once they recognize this, then it should be obvious that 65536 is a perfect square. Assuming they have astronomy and digital electronics, it's highly likely that they have telescopes with square sensors.

There's no technical reason to develop telescope sensors with a prime number of rows/columns, but there are good technical reasons to develop sensors with a power of two for the rows/columns. Binary digital computing technology "likes" powers of two.

so, send a color chart, then a picture using values from the color chart.

I like it!

8)

[coreybv]

It also occurs to me that perhaps it's a good idea to design different messages for different targets.

If there's intelligent life orbiting Alpha Centauri, we could carry on something loosely resembling a conversation with them. We might want the message to include some sort of a language primer to assist them in composing an understandable response.

Something aimed at a star 500,000 light years away would be nothing more than a "Hey, you're not alone", with no chance for a response.

[clint]

It also occurs to me that perhaps it's a good idea to design different messages for different targets.

If there's intelligent life orbiting Alpha Centauri, we could carry on something loosely resembling a conversation with them. We might want the message to include some sort of a language primer to assist them in composing an understandable response.

Also make sure to include:

[clint]

Something aimed at a star 500,000 light years away would be nothing more than a "Hey, you're not alone", with no chance for a response.

In that case, I would pick one of these:

[01101001]

I propose encoding it as a binary signal--a simple raster scan of 256x256x16. 16 bit words results in greyscale levels from 0 to 65535.

And this is data, not necessarily a picture. It is a surface in 3D, a function of two variables in tabular format. Beings worth talking with can figure it out.

It might be displayed as a picture in human-visble light, but also in other ranges of the EM spectrum. It can be rendered by the discovering scientists as a physical surface for consumption by the inhabitants of Planet Touch. Or as smell intensities for the giant nosesloths of Odorworld.

Data and display are different things.

[mugaliens]

Also make sure to include:

[IsaacKuo]

This thread is a bit old, but I had a couple extra thoughts...

And this is data, not necessarily a picture. It is a surface in 3D, a function of two variables in tabular format. Beings worth talking with can figure it out.

It might be displayed as a picture in human-visble light, but also in other ranges of the EM spectrum. It can be rendered by the discovering scientists as a physical surface for consumption by the inhabitants of Planet Touch. Or as smell intensities for the giant nosesloths of Odorworld.

Data and display are different things.

Right. I'm essentially sending data which emulates the sort of data which a digital sensor telescope would provide. I don't know what sort of display the aliens would use to "see" this data. They might not even have any sort of usable display for it--they might be limited purely to theoretical analysis of the data.

I mentally debated whether it makes more sense to send 16 bit greyscale data or 8 bit greyscale data. The overwhelming majority of images we humans use are limited to 8 bit greyscale data. Most of the images we'll be sending would be limited to 8 bit luminance resolution; the other 8 bits per pixel would be wasted.

However, the reason we use 8 bit luminance is because it's a good match for human vision. It is NOT a good match for telescope observations. The superior dynamic range and luminance resolution of 16 bit luminance is a better match for astronomical observations.

So, my guiding principle is to emulate what makes sense for a digital sensor telescope, not human vision. Therefore, I chose a 16 bit greyscale resolution rather than an 8 bit greyscale resolution.

[publiusr]

Well, if any extrasolar planets are found from the area of the WOW signal, it might be good to keep some assets pointed there in perpetuity http://en.wikipedia.org/wiki/File:Wo...l_location.jpg http://en.wikipedia.org/wiki/Wow_signal

[Ammonia]

I once heard an edited recording of the wow signal and it sounded like an old earth transmission or radio.

[clint]

I once heard an edited recording of the wow signal and it sounded like an old earth transmission or radio.

Do you have a link by any chance?

[Bynaus]

I sometimes wonder: Say, it is a genuine ETI signal. Say, the ETI detected the Earth as it transited our sun from their point of view (the position of the signal is very close to the ecliptic), analyzed the atmosphere with their CoRot/Kepler counterpart, found free oxygen and finally decided to send a message in our direction, "just in case" there would be someone to listen. Is there a certain distance X to the source of the signal for which the line of the ecliptic matches the position of the signal (at X years before present)? It's a little bit complicated to explain, but can you see what I mean?

[coreybv]

So, my guiding principle is to...

That's a good principle, but I think we need a simpler, more basic principle ahead of it. I would suggest:

Any CETI signal we send should be at least good enough that it would convince ourselves.

[iquestor]

That's a good principle, but I think we need a simpler, more basic principle ahead of it. I would suggest:

Any CETI signal we send should be at least good enough that it would convince ourselves.

Carl Sagan once wrote about such an attempt. Two teams of scientists were assembled. The first team was to compose a message to ET and convey basic information about who and where we are. the second team was supposed to play the part of the ET reciever and decode it.

On the given day, team 1 hands over their message team 2, which by prior agreement, was to be coded into a text file and handed over on a 3.5 floppy.

It was quickly discovered that Team 2 wouldnt be able to decode the message. Team 1 had used an MS DOS PC, while Team 2 only had an Apple, which could not read the file.

[coreybv]

It was quickly discovered that Team 2 wouldnt be able to decode the message. Team 1 had used an MS DOS PC, while Team 2 only had an Apple, which could not read the file.

This simply shows that simpler is better. Thus my preference for a repeating stream of prime numbers, encoded simply as "beeps".

Now, to nitpick a bit. Team 2 gave up too easily. The surface of the floppy could have been examined through other means to determine how the bits were arranged. The file system layout could have been deciphered, and the text file found. They could have then proceeded to decode it.

This is clearly too much effort when the disk came from the guys across the hall, but had it reached them via radio telescope from (verifiably) multiple light years away, I'm sure they wouldn't have thrown in the towel so easily.

As an illustrative tool, though, the story makes a good point. We need to make as few assumptions as possible, regardless of how reasonable they may seem.

[iquestor]

This simply shows that simpler is better. Thus my preference for a repeating stream of prime numbers, encoded simply as "beeps".

Now, to nitpick a bit. Team 2 gave up too easily. The surface of the floppy could have been examined through other means to determine how the bits were arranged. The file system layout could have been deciphered, and the text file found. They could have then proceeded to decode it.

This is clearly too much effort when the disk came from the guys across the hall, but had it reached them via radio telescope from (verifiably) multiple light years away, I'm sure they wouldn't have thrown in the towel so easily.

As an illustrative tool, though, the story makes a good point. We need to make as few assumptions as possible, regardless of how reasonable they may seem.

It was meant to be an illustrative story. Of course, they format problem was corrected I am sure, and team 2 got to work on it, but Carl's point was that there are going to be unforseen problems communicating between two very different societies, and we cant possibly forsee them all.

[coreybv]

It was meant to be an illustrative story. Of course, they format problem was corrected I am sure, and team 2 got to work on it, but Carl's point was that there are going to be unforseen problems communicating between two very different societies, and we cant possibly forsee them all.

Gotcha. I'm not sure what I misinterpreted, but after rereading your post it seems clear that we're making the same point.

Heck, even primes might be more complex than necessary. Just counting from one to ten should do the job. Just need something that can't be mistaken as, say, a pulsar.

Of course, we'll want to convert more information than that, but I do think it's crucial to have a very simple, unmistakable "beacon" to grab attention. From there we could have messages of increasing complexity, each building on the last, and hopefully get to a point where they'd have a fighting chance of deciphering useful information. Perhaps each new level could just appear at a higher frequency than the last, so everything could be sent simultaneously.

The message should also be repeated. Over and over. None of this 3 minute stuff, we should broadcast for 3 years, or decades.

Of course, this is all very expensive to do, and the funding to do it "right" just isn't there.

[iquestor]

Gotcha. I'm not sure what I misinterpreted, but after rereading your post it seems clear that we're making the same point.

Heck, even primes might be more complex than necessary. Just counting from one to ten should do the job. Just need something that can't be mistaken as, say, a pulsar.

Of course, we'll want to convert more information than that, but I do think it's crucial to have a very simple, unmistakable "beacon" to grab attention. From there we could have messages of increasing complexity, each building on the last, and hopefully get to a point where they'd have a fighting chance of deciphering useful information. Perhaps each new level could just appear at a higher frequency than the last, so everything could be sent simultaneously.

The message should also be repeated. Over and over. None of this 3 minute stuff, we should broadcast for 3 years, or decades.

Of course, this is all very expensive to do, and the funding to do it "right" just isn't there.

Have you ever looked at the Aricebo message devised by Sagan and Drake (I think?).. Its pretty straight forward once you figure out its format, which uses primes. wiki...

[coreybv]

The Arecibo message is a perfect example of trying to cram too much information into too small a space. Somewhere, some day, some poor alien cryptologist is going to go stark raving mad trying to figure out what number that red blob of pixels in the middle represents. (Yes, I realize there's no color represented in the actual message.)

The picture of Gilligan's hut at the bottom is nice, though.

[iquestor]

The Arecibo message is a perfect example of trying to cram too much information into too small a space. Somewhere, some day, some poor alien cryptologist is going to go stark raving mad trying to figure out what number that red blob of pixels in the middle represents. (Yes, I realize there's no color represented in teh actual message.)

The picture of Gilligan's hut at the bottom is nice, though.

TOO FUNNY!!! thanks

[publiusr]

I once heard an edited recording of the wow signal and it sounded like an old earth transmission or radio.

Well, if I were an alien and heard an Earth broadcast, I might replay it right back at you--where you might mistake it for your own...