Impossibly Stupid • April 2, 2018 9:41 AM
Not of fan of how they mangle the definitions of words. It mostly talks about steganography, not actual cryptography. It gives some rudimentary ways to encode words, not substantial ciphers.
A truly musical algorithm for encoding/encrypting information would be a really fun thing to develop. You could publish the paper in a journal, and then “drop” the album that is the paper played as music. If that already exists, someone please point me to it!
wumpus • April 2, 2018 10:22 AM
@Impossibly stupid: “A truly musical algorithm for encoding/encrypting information would be a really fun thing to develop.”
There are two ways to do this. To me “a truly musical” system would be one that generates “music” from a seed (presumably a large seed, or one that is continually accessed sufficiently many times to send a useful amount of data). This has the advantage as that it essentially undetectable without the generating algorithm (of course, it is axiomatic the attacker has this). Of course, the “music” has to be sufficiently good to avoid suspicion.
The other is of course to simply hide the data bits in the least significant bits of music/audio. This presumably means directly meddling with MP3 (and similar) coefficients, but presumably will be very hard to hear that they have been altered (for high bit rate audio). This is wildly easier, has a vastly higher bit rate, but sufficiently fancy analysis (with various fft/wavelet/fractal analysis) will presumably notice the excessive noise.
The obvious question becomes: how much purely random noise is in most high-rate audio and could an attacker have any chance of determining which is noise and which is encryption (presumably such encryption “salts” the “real data” with increasing randomness to hide the break between audio and encryption).
Snarki, child of Loki • April 2, 2018 11:00 AM
Unfortunately, I have fallen far behind in my project to encrypt messages into Vogon Poetry.
It would both conceal the messages, and cause the intestines of those who intercept the messages to leap out of their own body and strangle them to prevent further exposure to the poetry.
Computerized encryption/decryption is required, of course, but the damn AIs keep killing themselves.
David Rudling • April 2, 2018 11:18 AM
True this is steganography rather than cryptography but it is an interesting, if rather limited, alternative to the more usual steganography of hiding your (already securely enciphered) message in e.g. a jpg file. There seem to be a depressing number of tools to locate such “hidden” messages if using one of the widely available tools to do the hiding in an image. In an age of reasonably secure cryptography but universal metadata, the massively increased use of steganography would be a highly desirable objective to aid privacy. Musical steganography sadly isn’t the answer but I hope someone somewhere is working on a breakthrough approach to meet this urgent privacy need.
Douglas Coulter • April 2, 2018 12:29 PM
C U Anon • April 2, 2018 1:16 PM
@Snarki, child of Loki
Paula Nancy Millstone jennings of Sussex on the other line for you…
Something about purifying Swans, that were once white but are now black. She’s blaiming Earth II and reincarnation apparently she’s not got a heart of gold…
She also says Retsina and olive oil are not what they once were, and something about a party not getting bailed out and it not being Kricket or some such.
Bauke Jan Douma • April 2, 2018 1:57 PM
Instead of spending time on this, I decided to re-read Gödel, Escher, Bach.
Jon (fD) • April 2, 2018 2:40 PM
OT, but fun:
There’s a brief little mystery tale about a singing couple, one with perfect pitch, who conspire to knock off one of their spouses. Unfortunately, their alibi is compromised by the awkward death of the guy who’s supposed to corroborate it. While being interrogated by the police (who know of the death), one of the couple (who also knows, but is being kept from speaking) idly taps on a piano.
Spelling out “A-B-E D-E-A-D”. They then magnificently cook up an instant replacement alibi…
It’s in a collection I have here somewhere. Will look up the citation if anyone cares. J. (fD)
Impossibly Stupid • April 2, 2018 5:49 PM
There are two ways to do this.
I was thinking more of a third way: an algorithm that works to add “order” to a system rather than disorder. Encryption is about scrambling data until it looks random; the output is essentially unrecognizable chaos. I want the opposite of that: the output should essentially be too recognizable as something else. In this case, music.
Steganography is the method for doing that sort of thing with existing cover data. I’m just wondering how much more could be hidden if there were algorithms that went the extra measure and generated that cover data, too. Ideally, the secret data could also be analyzed in a way that picked from among many kinds of algorithms (e.g., “The data you’re trying to encode is a best fit for a dub step AIFF (80%) or a PNG of a beach scene (72%)”).
That may be a bit sci-fi right now, but we should expect things like that to start happening at some point.
albert • April 2, 2018 6:16 PM
It’s kinda disappointing that the link of the article cited has about 20 scripts associated with it. I didn’t bother to see how many are required to listen to the examples.
Modern music is based on the 12 tone scale (see “Arnold Schoenberg” for details). A decent code would have to map out the 12 tones to include letters beyond the standard A-G mappings. The question is: do you want a simple code, like the classical composers cited, or do want ‘secret code’?
Any system based on notes only lends itself to computer analysis. Even though there are 7 1/3 octaves on the standard piano, that’s only 88 choices for encoding. Adding rhythm, note duration and key signature helps a lot. It looks to me like your melodies are not going to sound diatonic, that is, not related to familiar tunes. So it’s going to sound like some of those ‘modern’ classical compositions, i.e., musical garbage.
For this reason, I suggest ‘modes’. Listen to Miles Davis, “So What” https://www.youtube.com/watch?v=ylXk1LBvIqU It’s C Dorian, the notes of the C scale, from D to D’. Seven-note modes only allow for about 49 choices, but you can use modes from any of the 12 major keys in a single piece. In “So What”, C Dorian starts the piece, but shifts up to C# Dorian in the middle, and back down to C Dorian at the end of each chorus.
I’d definitely -not- use computer-generated music. It should be played by actual musicians (preferably jazz musicians), who can alter tempos and note durations slightly. Software for music analysis has come a long way. The idea is to make it -necessary- for a -skilled human- to analyze the piece.
I would love to see some examples, in notation if possible.
Sorry for the technical stuff. I made it as simple as I could.
. .. . .. — ….
Luzugaz Fenyev-Baixar • April 2, 2018 9:38 PM
Of course, this was all done long ago, viz. the Rodgers-Lawrence-Hammerstein encrypted identity musical whistling code, used by a 19th Century English woman intelligence operator whose cover was governess to the children of the ruler of an East Asian country.
justinacolmena • April 2, 2018 9:42 PM
I can’t even read sheet music.
I understand that very few people, even if they are capable of reading sheet music and playing instruments, are able to actually listen to music and transcribe the notes into some kind of notation.
How is this supposed to work? Someone please explain!
Hmm • April 2, 2018 10:19 PM
Like all things it takes doing but it’s not that difficult to learn to read sheet music.
milkshaken • April 2, 2018 10:24 PM
If you read The Piano Tuner by Daniel Mason, in a sudden plot twist one of the main characters in the novel is accused by the military of secretly communicating with the Russian enemy through the coded messages disguised as sheet music. (But you understand it is most likely just a canard used to destroy him)
echo • April 2, 2018 11:04 PM
Computer generated German music is the pits. Germany had a thing for this buried in its architecture and art waves. It’s very clever and good for what it is. I can only take about five seconds before screaming. Jazz is definately easier on the ears.
Roboter • April 2, 2018 11:13 PM
“Computer generated German music is the pits.”
Bits… Is the bits.
Luzugas Fenyev-Baixar • April 2, 2018 11:39 PM
Kraftwerk’s music decodes the ciphertext of our everyday presumption of confident self-understanding to reveal the plaintext of our somnambulant transformation into a diminished souless machine state of being. They are in a way sacer.
Jon (fD) • April 3, 2018 1:15 AM
Or they, like you, are just a load of meaningless noise. 😉 J.
Cassandra • April 3, 2018 4:47 AM
It should be of no great difficulty to the musically inclined to devise a code represented by the choice of variation applied to a particular piece. There is more than sufficient scope to apply a Baconian encoding, and once an encoding process has been agreed between sender and recipient, a cipher can additionally be used. The skill would be in producing a musical variation that was musical, rather than displaying mechanical variations, as pointed out in the text linked to by Bruce. I suspect Jazz musicians would have the greatest facility in doing this.
VinnyG • April 3, 2018 7:38 AM
@wumpus re: “Of course, the ‘music’ has to be sufficiently good to avoid suspicion.” Ha! Possibly the scope of your listening isn’t terribly wide – I find some of what passes for music and is popular these days mind-bogglingly awful (and I have pretty eglatarian genre preferences.)
In general, this is a very intereting concept. There is a plethora of analog information present or implied in a musical performance (much more so than in the score.) Just to pick one example, different examples of dissonance produce beat frequencies that vary. It’s been a long time since I studied music theory, but I seem to recall that in some cases there secondary beat frequencies in addition to the primary (certainly true when discussing chords, rather than single notes.) There is also an entire sub-discipline in piano tuning concerning the magnitude of beat frequencies and where in the 88-key range they are located. There seem to me to be potential elements of cryptographic, steganographic, and one-time pad techniques here. One question is if a message were encoded using such techniques, how much would survive the digitizing process.
HRA • April 3, 2018 8:07 AM
While not coding to the note, the BBC was involved in coded playlists during WW2: https://www.telegraph.co.uk/culture/4707901/Playing-for-victory.html
i’ll drop this here:
she invented frequency hopping spread spectrum.
by changing frequency of the transmission according to a secret song played on the piano, each piano key changed the transmission frequency.
the sad part is that military waited for the copyright to expire, she didn’t renew it because noone cared, and than military started to use it…
today is found in many “user” (=not military) transmitters
Anders • April 3, 2018 3:11 PM
Getting local root via beep sound is close enough to the topic 😉
Clive Robinson • April 3, 2018 4:58 PM
The skill would be in producing a musical variation that was musical…
I think it was Mozart who came up with a “compose your own minuet system that he sold. In essence he wrote a list of musical phrases that you selected by rolling dice. The results though stylised were something you could play and dance to if you wished.
He probably got the idea from having been taught to “sketch” out phrases and then combine in to works very rapidly.
Oh and what appears to be a requirment for some anonymous deliberartly pedantic types that appear to have taken over the blog…
albert • April 3, 2018 5:26 PM
“…very few people, … are able to actually listen to music and transcribe the notes into some kind of notation….”
It’s called ‘transcribing’. Writing music notation by listening to it. Did you ever try to write or type what someone is speaking? It’s a lot like that. Some folks are talented transcribers and some are not. It helps to have ‘perfect pitch’ or ‘perfect relative pitch’, but you don’t even need that. You can match the notes you hear by finding them on an instrument, then writing them down as letters (A, B, C…) And you’ll need to know sharps(#) and Flats(b) and the octave you’re in, but that’s it. Anyone can then read your transcription and play it. What’s difficult is when you have multiple notes playing at the same time, like in chords.
The Masters of classical music could write orchestral scores from memory. The great jazz pianist Errol Garner (who couldn’t read music) once listened to a concert by a classical pianist, then went home and played pretty much the same concert, from memory. Such musical memory is rare. He wrote his famous ballad, “Misty” while on a plane, then later found a piano to ‘fix’ it forever in his mind.
If you can sing a song ‘on key’, then your ear is good enough. If you can listen to a melody, then sing it, so much the better.
I encourage everyone to learn notation. Most all of my guitar students can read. It’s the universal language of music.
And you might as well learn an instrument (piano and guitar are best) along with the notation:)
. .. . .. — ….
GregW • April 3, 2018 6:59 PM
There was a Scooby Doo (childrens mystery cartoon) episode in th 70s featuring musical encoding as a key plot element if I recall correctly. Dont remeber if it explicitly mentioned Bach encodings or not.
echo • April 4, 2018 6:32 AM
I played string and brass and wind instruments when I was a child and have a string instrument I bought later but my interests and abilities have moved so I should dispose of this. For learning (or re-learning) musical notation a recorder would help and is relatively cheap. I also have a tin whistle I bought for fun which will do too.
VinnyG • April 4, 2018 11:57 AM
@echo re learning musical notation – In my limited experience, many inexpensive recorders (of the wooden variety) have demanding embrochure requirements and can be difficult to blow correctly (I suppose this could be a symptom of my many years spent playing single reed woodwinds.) I’d be more inclined to suggest a rudimentary electronic keyboard for the purpose. In the US, those can typically be found on Craigslist in good used condition for $20 or so. Not as portable as a recorder, but some of the smaller examples are battery-powered and not too heavy or awkward to transport. A small chromatic set of “bells” (xylophone-ish) might be another candidate instrument.
justinacolmena • April 4, 2018 1:23 PM
For learning (or re-learning) musical notation a recorder would help and is relatively cheap.
We played those in kindergarten. “Hot cross buns, hot cross buns, one-a-penny, two-a-penny, ….” I don’t remember all the notes, there were some holes you covered with your fingers and some you left uncovered for the different notes, and there was the timing, too. Whole note, half notes, quarter notes, and rests.
I also remember the BASIC programming language, which could be instructed to play music on a simple internal speaker driven by a square wave straight from a logic chip on the motherboard. (A square wave, of course, has serious harmonics at 3, 5, 7, etc. times the pitch of the intended musical note.) There was a standard clock tick (some type of internal metronome) that generated a hardware interrupt at 18.2 times per second, and called an interrupt service routine, which required pushing the contents of all the registers on the stack and then restoring them; those routines were not even re-entrant. The sound played for 7/8 of the duration of the note by default, but this could be lengthened or shortened by using some Italian words.
echo • April 4, 2018 4:57 PM
Plastic recorders work and a decent quality brand is available for £5-10 from the usual trans-national tax dodger.
Wesley Parish • April 5, 2018 4:58 AM
FWVLIW, in Dune, and Dune Messiah Frank Herbert has the distrans:
DISTRANS: a device for producing a temporary neural imprint on the nervous system of Chiroptera or birds. The creature’s normal cry then carries the message imprint which can be sorted from that carrier wave by another distrans
or for that matter, on the normal speech patterns of a human individual chosen as a distrans.
And then there’s the drum speech of Western Africa, where drum speakers use their language’s tonal patterns and multiple redundancy to “talk”. See Jeremy Montague, Timpani and Percussion.
Of course, what everybody’s not telling you is that to fully immerse yourself in the musical experience, you have to take up a very, very simple instrument like the French Horn. Or the Timpani – the older, non-pedal sort – ideal for handling fast tuning changes … I myself can confess to having taken up that very simple and easy instrument the Pedal Steel Guitar when a teenager, for some relief from difficult instruments like the recorder and the tambourine. (I was never able to handle the ukulele – far too many strings!!! )
VinnyG • April 5, 2018 9:55 AM
@Wesley Parish – I’m not sure Justina Colmena has enough musical background to fully appreciate your sarcasm… @Justinacolmena – Mr. Parish seems to have omitted the adverb “deceptively” which should occur just before every instance of the adjective “simple” in his paragraph… @Wesley Parish – If you devoted significant time to learning pedal steel, do you find the name J Winston recognizable?
vas pup • April 6, 2018 8:53 AM
That is related to acoustic and security:
Computer system transcribes words users ‘speak silently’
Electrodes on the face and jaw pick up otherwise undetectable neuromuscular signals triggered by internal verbalizations
The system consists of a wearable device and an associated computing system. Electrodes in the device pick up neuromuscular signals in the jaw and face that are triggered by internal verbalizations — saying words “in your head” — but are undetectable to the human eye. The signals are fed to a machine-learning system that has been trained to correlate particular signals with particular words.
The device is thus part of a complete silent-computing system that lets the user undetectably pose and receive answers to difficult computational problems. In one of the researchers’ experiments, for instance, subjects used the system to silently report opponents’ moves in a chess game and just as silently receive computer-recommended responses.
The basic configuration of the researchers’ system includes a neural network trained to identify subvocalized words from neuromuscular signals, but it can be customized to a particular user through a process that retrains just the last two layers.
“The other thing where this is extremely useful is special ops,” Starner adds. “There’s a lot of places where it’s not a noisy environment but a silent environment. A lot of time, special-ops folks have hand gestures, but you can’t always see those. Wouldn’t it be great to have silent-speech for communication between these folks?”
As I recall there was special device in tanks (armor machines) called laryngophone which was placed on the neck of tank’s crew member and caught signals directly from vocal cord (noise is high in the tank – kind of ‘microphone’ on the neck). Then it was developed Neurophone technology by Patrick Flanagan which could silently transmit audio input to the brain through the skin (not bone conducting). Do you think that combination of those two devices could do the same function of silent-speech without computer involvement?
You professional opinion is as usually highly appreciated
Wesley Parish • April 7, 2018 3:04 AM
Thanks! Yes, I recognize the name J Winston 🙂 And Bill Keith, if you should mention that. I did worry that I was being too hard on Justina Colmena, but I figured since I wasn’t making a personal attack, she should be able to laugh it off.
Ben Dover • April 8, 2018 2:16 PM
Has anyone thought of whether this could somehow be involved in numbers stations? Sometimes a strange little melody will be played before the eerie recitation of numbers and letters. It always seemed strange to me, I assumed it was just a way of getting the listener’s attention or signaling the beginning of the broadcast, but perhaps there was more to it… The Conet Project has many recordings. It seems pretty unlikely, but I’d love to hear informed speculation one way or the other…
Subscribe to comments on this entry
Sidebar photo of Bruce Schneier by Joe MacInnis.
Leave a comment