Tapping Morse code on a wall | Hacker News Reader

Tapping Morse code on a wall

mra-raycom.com

67 pointslightlyused59 comments

I am fairly sure that I have read (probably in a 1950:s edition of the Swedish navy soldier’s handbook "Örlogsboken"), that "tap morse", for example for communication with persons trapped inside a damaged ship, is done by tapping once for a dit, and twice in rapid succession for a "dah".

(Page 106 in the 2003 edition, which can be found in various ad-infested document-showing web sites.)

This is correct, see my other comments.

It is in fact the way that a standard telegraph sounder encodes Morse.

> Recently there was real-life incident, the Titan submersible accident which started speculation about a banging noise being Morse Code. When the submersible imploded over 3,800 meters below the surface, the resulting sound was heard as a loud bang by the crew of a passing ship. Conspiracy theories quickly emerged suggesting that the bang was Morse code, a desperate attempt for help from the doomed submarine. However, experts dismissed this idea, explaining that the sound was unrelated.

What kind of game of telephone led to this? The periodic banging noises were reported on Wednesday[0]. Rescue ships heard these noises. The sub had imploded on the previous Sunday. To my knowledge, no passing ship heard the implosion; it was detected by US Navy sensors. The #1 thing reported about the banging noises was that they were coming at 30 minute intervals. Obviously the submarine imploded only once.

Nobody mistook the implosion of the submarine for Morse code - the banging noises were some other third thing. This paragraph must have been the product of substantial confusion.

[0] For example: https://nypost.com/2023/06/21/glimmer-of-hope-30-minute-inte...

> To my knowledge, no passing ship heard the implosion; it was detected by US Navy sensors.

One plausible reason why immediate but not slightly delayed release of the fact that it was caught by Navy sensors was considered sensitive os that those sensors were on (and thus the location would reveal thr location of with more specificity than was desired) a ballistic missile sub.

During (what we Americans call) the Vietnam War, American prisoners of war, often held in solitary confinement in adjacent cells in the Hanoi Hilton, would use the tap code extensively to talk to each other, sometimes carrying on long conversations, for example about ways to resist their captors. The article alludes to that but doesn't go into a lot of detail about that bit of history.

(This is discussed in https://en.wikipedia.org/wiki/Tap_code cited downthread by 'dmckeon.)

If they're in adjacent cells, using a tap for dit and a scrape for dah should work.

In the film adaptation of "Let the right one in" they use this method.

I was introduced to this tap code just after being incarcerated and used it often. A friend and I also modified it to be able to communicate in/out of solitary confinement as they had kiosks in each cell block allowing you to order commissary food. The kiosks had many bugs I found in the code, but they especially lacked any sane range checking, so you could enter orders for up to 9999 noodles for instance.

We broke the knock code into two "byte" pairs. So HELL would be an order of 2315 Beef noodles and 3131 Chicken noodles.

To access the orders between cellblocks it was necessary for the person in solitary to write their PIN (their username was their jail ID#) on a piece of toilet paper and pass it to someone outside solitary on a day they were in court. Then both sender and recipient could log into the same account from different cellblocks and read the order. Once read the order would be deleted and another order placed to reply to the message.

At first I was out of solitary, but then I got sent for extortion (charge dropped after 10 days, story of my life) and while I was in solitary the staff went through my stuff and found some pieces of paper where I had sat encoding and decoding messages. The staff didn't have the table though and through the FOIA I managed to get access to their working out, but they never successfully decoded any of the undecoded messages they obtained from the kiosk. After that we rotated the encoding table 90 degrees in our heads to act as a very simple form of encryption in case they tried again. It took the authorities another 6 months to persuade the kiosk developers to add a piece of code which checked the jail ID# that was logging in against the cellblock they were in to make sure you couldn't log into an account if you weren't in the right cellblock.

That's an amazing story!

This sounds like something out of a Cory Doctorow book, I love it.

The game her story features some tap code. I hacked together a trainer https://projects.sys3.xyz/knockcode/index.html

This is article is a pedantic waste of time.

Tapping out SOS is a thing… https://www.smithsonianmag.com/history/the-civilian-worker-w...

Tapping out SOS with a wrench makes sense. SOS has an unmistakable sound to it that is instantly recognizable to even the most novice Morse operator (even when I've practiced it, it has chilling effect on me).

Even though each tap (each clunk of the wrench) has the same duration the altered mark/space ratio makes it unmistakable. (Try it by tapping a rule on the desk or clapping hands.)

If a skilled Morse operator were ever in that situation I'm pretty certain at least simple messages could be communicated that way—almost anyone will recognize SOS, skilled operators will be able to do better.

Anecdotally, I clicked the article, read nothing, and immediately listened to the first sound clip where I was able to make-out that it represented "CQ", but to be fair I'm a ham and "CQ" is by-far the most common morse code message.

All you need is a ternary system where there is dit, dah, and silence. Even just tapping on different parts of the "wall" can serve this purpose, so long as it is able to be differentiated.

The whole point of the article is that there is no need for differentiation. eg just click/click is sufficient. There are two clicks for each element, and what changes is the space between the clicks. eg it's the rhythm which defines the character.

It operates in exactly the same way as the armature in a telegraph sounder.

I disagree, the stated point the author is making is that I shouldn't be able to interpret the first clip.

Edit: Okay, I see your point, reading your other comment. Yes, it's the case that you do not even need dit/dah/silence, you can get away with just 'dit' and silence, provided you're consistent with your timing.

CQ is instantly recognizable to us, so is SOS even though most of us have never heard over the air. SOS of course would be recognizable to almost anyone whereas CQ requires at least some Morse training.

The argument is pedantic. It may not be Morse, but it's definitely possible to communicate by tapping on the wall. Imagine tapping out all of Anna Karenina to save a fellow inmate's life.

https://www.npr.org/2017/09/11/550058353/rough-translation-h...

https://www.bbc.co.uk/programmes/p03ttfks

One of my college classes was network communications and for a group project, we had to devise how to transmit a binary message visually standing across from each other by 100 yards or so. Four people to a group, split two and two.

The idea was to get us to strike a balance between speed of encoding/decoding, ability to distinguish symbols, and how many bits to encode in each symbol.

I think my group came up with a system to transmit 3 bits at a time which failed spectacularly on the windy day we had for showing it off. My recollection is we had a big box with neon colored flaps that we could fold in/out.

I wish we'd researched prior works of art. To this day, I have no idea why we didn't just learn flag semaphores. Maybe there was a rule against using existing systems though?

I'm confused. Semaphore is not a binary system.

But you could use it to encode 4 (or with the full list of signs and something additional) even 5 bits at once (as he said, they had a system of 3 already).

The message we had to transmit was given to us as 1's and 0's. But we could transmit as many bits per symbol as we wanted. I think we were scored based on speed and error rate, so it did you no good to finish first if the message was garbled.

This was almost three decades ago and I'm working from memory so I may very well have some bits wrong.

Ah, that makes sense. Semaphore would have been ideal, then.

https://en.wikipedia.org/wiki/Semaphore#/media/File:Semaphor...

>The argument is pedantic. It may not be Morse, but it's definitely possible to communicate by tapping on the wall.

IMO the article made a useful distinction. The whole second half covers "tap code", as is used in prisons, etc. I've known Morse for decades but somehow hadn't bumped into tap code before.

You should re-read the article. It didnt reject communications via tapping is NOT POSSIBLE. It is saying it is unrealible and can confuse the messages as some tappings are hard to distinguish. Next time, try to use a successful example to back your argument. The article is stressing the fact that Morse code cannot be use to transmit message reliably over a pipe via tapping. It is not refuting doing this communication is IMPOSSIBLE. Explain to me how you are going to do flags over pipes or walls. Stay within the context of discussion.

If "Morse code cannot be use to transmit message reliably over a pipe via tapping", then morse via the tapping of a sounder would also be impossible.

The original morse sounders used a large relay which went "Click" when operated and another "Click" when released.

So a Dit sounds like Click/Click, while a Dah sounds like Click ..... Click.

Which is how only one sound (plus spaces) can encode morse.

In other words it's all in the rhythm of the sounds.

Morse SOS:

bang-bang-bang---bang---bang---bang---bang-bang-bang

While Morse in general has problems with just banging, with SOS it is not a problem.

This is nonsense.

The original morse sounders used a large relay which went "Click" when operated and another "Click" when released.

There is no problem in sending morse via clicks. If there was, then the original telegraph would not have been possible.

Exactly, as I mentioned above. CQ requires training (but not much—I have no trouble recognizing it when the signal is only modulated noise), but SOS is another matter altogether, even those not trained in Morse recognize the sound especially if repeated (that 'ring' is unmistakable).

Compare to: https://en.wikipedia.org/wiki/Tap_code

Disagree.

I'm a ham; my kids aren't... but they grew up with me using Morse (my computers talk to me when they boot up, and when a 'kitchentimer' goes off, and...) so they were naturally exposed to the paradigm. Maybe that's a limiting factor, but...

With CW you're limited to two states: key and unkey. Knocking on a wall, you're not: you can go loud and soft as well. A dah has the same pacing as it always does (3 dits in length), but if its knock is perceptibly louder, at least twice as hard as a dit, in practice that seems to make it hold together as readable Morse, at least for us. Certainly our family signal, questionmark (..__..), is usable that way, so is each son's "call-letter".

This is off topic but I’ve always wanted to know what hams talk about on their radios (when not using it for emergency purposes)

Ham radios. The weather. Space weather, because that affects ham radios. Growing old sucks; it's a hobby dominated by the aged. Antennas. Kids these days. We avoid religion and politics, mostly, though there's an increasing segment of MAGA wing nuts who no longer think of their hateful rejection of the majority of their fellow humans as politics. Fortunately, one can spin the dial.

It's the ARRL field day for a few more hours; check out http://websdr.ewi.utwente.nl:8901 near 14100 kHz, looking for 2.7 kHz wide voice transmissions (USB), or near 7200 kHz LSB.

Anything but politics and religion, mostly.

Casual small talk, antennas and other ham gear, perhaps trying to speak a foreign language (ham radio worked wonders for both my Russian and my Portuguese!), family, other interests...

Some communications between hams are no more than call signs, signal reports, and 73s (“best regards”, ending the QSO (conversation)). This is useful for things like satellite communications, which are very time-limited opportunities for making contacts.

This last weekend was ARRL Field day, a huge operating event designed to be "practice" for when the "stuff hits the fan" but also just gets a ton of people on the air using portable stations set up in public and austere environments (like in a park pavilion, not in a cozy basement or office).

I made a playlist of as many FD videos i could find on youtube [0]. There's a lot of examples of stations, setups, and the actual contacts - which are very short.

Field day is a sort of contest. In a nutshell, the more contacts you make, the more points you get. So, the contacts are short and have a bare minimum of information (callsign, operating class (#radios + power level) and US state/CAN province).

"Normal" voice and CW conversations are much different, and just sound like two people chatting about life, the weather, family, health, kids these days, or technical radio-related topics like descriptions of their station, propagation studies, or reports of the last hamfest they went to.

There's also "nets" which usually have their "check-ins" give a short report, or none at all (aka short-time). Some nets relay information about severe weather (skywarn / storm spotting), others relay official national traffic in accordance with the Incident Command System (ICS) to practice for wide-spread disasters like hurricanes and earthquakes. The radiogram system still runs via ham radio too, and you can hear national traffic system (NTS) nets relaying messages the old-fashioned way.

Right now, the majority of amateur radio traffic on HF is currently FT8 [1]. IT's a low-speed weak-signal FSK mode, in which operators exchange only each other's callsign, signal report, location (4-char maidenhead gridsquare), and a few extra characters to say hello (CQ), roger (RR), or goodbye (73) in a 77-bit message.

Other stations might not even be hams talking, like APRS, WSPR, or signal propagation beacons. There are quite a few automated stations that perform a variety of tasks.

There's a TON of low-earth orbit cubesats that use amateur radio for telemetry and command/control, and a few operate as a repeater that hams use to relay a signal across wide areas. There are even hams aboard the ISS, as well as a repeater. So sometimes we say hi to astronauts. They don't say much since a lot of people are in line to get that highly-desired contact. Sometimes schools coordinate with radio clubs to have students ask questions directly to astronauts via ham radio. It's old-school, but it provides an excellent learning opportunity about the utility of radio.

Bottom line, hams talk about anything. And some don't talk about anything at all. And some aren't even people. Some are astronauts. It's a very broad, multi-faceted hobby so there's lots to talk about.

--N0SSC

[0] https://www.youtube.com/playlist?list=PLG8UQxewXrtbT9cBWL-3J...

> With CW you're limited to two states: key and unkey

This might be true for radio morse, but in the original Telegraph, each element was encoded by two clicks. One click when the morse sounder operated, and another click when the sounder released. In other words, each element is encoded via two clicks, but with different spacing.

see my other posts.

So sad to read the misleading stuff here.

The original morse sounders used a large relay which went "Click" when operated and another "Click" when released. (some will argue that it's "Click-Clack", but that makes no difference)

So a Dit sounds like Click/Click, while a Dah sounds like Click ..... Click.

Which is how only one sound (plus spaces) can encode morse.

In other words it's all in the rhythm of the sounds.

Years ago, divers were taught to send morse by holding a spanner in their fist, and rocking it back and forth against the hull (eg click/click) in exactly the same way as the armature in an original morse sounder operated.

And yes, there is also a completely different system called the "Tap Code".

The only Morse code message I know is SOS which is ...---... and I imagine it's reasonably easy to understand when being tapped in an emergency situation. It would tell rescuers that someone is still alive and to continue their search.

I'd be screwed. Probably would be alot of irritated hams that day wondering why some asshole keeps tapping out OSO.

Fun fact: the default message ringtone on old Nokias was ...--..., which is Morse for "SMS".

https://youtu.be/-uuG4ZCJmyU?t=57

It's 'inverted morse', the space length becomes the dot and dash. With a double/triple length space for a word break

Yeah, this is how I always thought it worked, three different length pauses for dot, dash, and space/break.

There's a lot of commentary here about Morse code being possible by tapping on a wall, which it is if and only if both sender and receiver are already trained in Morse. Problem is, your average prisoner isn't, the encoding is not at all obvious (because it's designed for speed of use, not ease of learning), and they can't just look up it up on Wikipedia when in solitary confinement,

This is why simpler tap codes exist: A = 1 tap, B = 2 taps etc is self-evident, and from there it's only a small leap to the Polybius square 5x5 tap code covered in Wikipedia.

For those who still don't understand how the Telegraph uses simple clicks to encode the alphabet....

Download "Morse Code Tools" from https://morse-code-tools.software.informer.com/2.1/

Then run "Morse Keyer" and select "Straight Key" and "Telegraph Sounder"

You can hear exactly how a Morse Telegraph sounded back in the day

Two clicks per element, with different spacings between them.

Likewise run the "Morse news" program to hear some text.

As anyone with even a smattering of Morse knows counting dots and dashes essentially gets you nowhere. Morse is about the rhythm of the sound, each character has a unique rhythmic and ultimately recognizable sound. (One only makes progress when one stops hearing individual dots and dashes of a character.)

It's worth noting that during the American civil war many telegraph operators used something called dot code, which consisted of clusters of 1-3 taps that corresponded to a contemporary flag signal code. That could certainly be transmitted through a wall. Likewise, you could probably slow down morse and transmit long pulses as clusters of 2 quick taps if you wanted to reuse that code on a wall.

You don't have to "slow down Morse". That is exactly how Morse is sent on a telegraph system.

That's not entirely correct. As the article describes, landline Morse with a sounder has two distinct sounds for every code element, a different sounding click at the beginning and the end. A quick dit ("dot") is likely to resolve as a single sound to the listener, but it's actually two sounds.

What I'm suggesting is to replace a dah ("dash") with two short dits when tapping out a message. But in order to do that, you need to increase the inter-element spacing so you can tell the difference between a dah and a pair of dits. That's what I meant by "slow down". For the same dit timing, the time required to send a message is longer, as you can see here:

    original  --- . --- .   --- --- . ---
    2 dots    . .   .   . .   .       . .   . .   .   . .

Note: The top line is based on the standard timing:

- Each dit is one unit

- Each dah is 3 units

- Inter-element space is one unit

- Inter-letter space is 3 units

I have seen Shawshank probably more than 100 times and I cannot for the life of me remember a Morse code scene that is mentioned in this article.

Thank you. I thought it was just me having this same experience.

Ok, so not only is this article a pedantic waste of time, but also evidently it hasn't undergone routine fact checking.

One the most interesting uses of morse code was a Vietnam POW blinking the word "torture" during an interview.

See here for the interview: https://www.youtube.com/watch?v=rufnWLVQcKg

I don’t know if I agree with the author’s conclusion that you can’t differentiate Morse dit and dahs tapped on a pipe.

That’s assuming that your listener has no brain and can’t figure out what the longer pauses between your taps mean.

I think the author has stricter interpretation of the morse code tappings shown in movie which largely is just tappings. In actual fact there are a large variety of customized Morse code for wall/pipe tappings. One such is using loud tap for a dash and soft one for a dot. Same goes with another variant where a tap for a dot and a quick double tap for dash. The problem with all these is the tappers need to be very well train to produce a consistent tapping rythmn which is very difficult if you're not doing it daily let alone those shown in the movie doing it for the first time or after years of not doing it. Then on listener side need to have very sharp ear with probably perfect pitch and rythmn sense which we assume had some kind of strong musical background also equally well trained in Morse code tappings on pipe/wall and in those Morse code variant being used. In short, this will turn into "perfect shot" kind of youtube series...extremely unlikely but "common" in Hollywood movie depiction.

This is nonsense.

Any experienced morse operator can copy the code from a morse sounder.

The secret is that a Dit is "click/click", while a Dah is "Click....Click".

It's all in the rhythm of the sounds.

Is Morse code encoded in the lengths of the beeps or the pauses between the beeps?

Both :) the standard unit Of time is a dit. A dah is 3 dits long. 1 dit between each element (so s is ... But 5 dits long because its on, space, on, space, on). 3 dits between letters. 7 dits between words.

When people make the dits too long or the dahs too short (c vs y), or run characters together (ee or i?), it gets hard to copy correctly.

From what I remember reading, when they sent code over telegraph wires, dit and dah were distinguished by the time between clicks. Each one required two clicks to send, but the clicks were closer together for a dit.

You could probably Google it to be sure.

This is exactly correct. It amazes me that so many here seem to thing that the telegraph couldn't work because it only uses clicks.

Reply on news.ycombinator.com