Originally Posted by 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?"

Originally Posted by 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.

Originally Posted by coreybv

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

Originally Posted by 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.

Originally Posted by IsaacKuo

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.

Originally Posted by 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.

Originally Posted by 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.

Originally Posted by coreybv

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.

Originally Posted by 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.

Originally Posted by clint

Also make sure to include:

Originally Posted by 01101001

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.

Originally Posted by Ammonia

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

Originally Posted by IsaacKuo

So, my guiding principle is to...

Originally Posted by coreybv

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.

Originally Posted by iquestor

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.

Originally Posted by coreybv

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.

Originally Posted by iquestor

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.

Originally Posted by coreybv

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.

Originally Posted by 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 teh actual message.)

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

Originally Posted by Ammonia

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