Unique Information Security Approach using Carnatic Music

Ideas and innovations in Indian classical music
Post Reply
deeps92
Posts: 8
Joined: 02 Mar 2016, 20:46

Unique Information Security Approach using Carnatic Music

Post by deeps92 »

Dear All,

I am a new member to this forum. I am working on an information security related project where I have come up with an approach that makes use of carnatic music.
The basic idea is to convert a text message (secret between sender and receiver) into a music clip at the sender's side which will be sent to receiver. The receiver then deciphers the music clip back to the text message.
I have come up with a logic that does this conversion from English characters to a sequence of 'swaras'. The music clips generated as now are just sounds of specific frequencies (pure notes without 'gamakas' and other emotional aspects of music).
I wanted to make a survey which can give me an idea if I am going in the right direction with respect to my approach.

Please listen to the 5 music clips that I have shared in the below link (1) and kindly record your opinion at link (2).
link (1) : https://drive.google.com/folderview?id= ... sp=sharing
link (2) : https://docs.google.com/forms/d/179-S18 ... _form_link

Each of the five music clips belong to one of the ragas in the following list :
Mayamalavagowla , Kalyani, Kharaharapriya, Shankarabharana, Harikambhoji.

Kindly give your opinion on the raga of each clip.

Please do help me with this!

vasanthakokilam
Posts: 10956
Joined: 03 Feb 2010, 00:01

Re: Unique Information Security Approach using Carnatic Music

Post by vasanthakokilam »

deeps92, I gave it a try. But I can not make out which raga each one of the clips belong to. May be someone with better swara gyana can do better.

btw, is the same text encoded in the various ragas or these are different texts for each of the clips? They all seemed to be similar in terms of the overall pattern.

Rsachi
Posts: 5039
Joined: 31 Aug 2009, 13:54

Re: Unique Information Security Approach using Carnatic Music

Post by Rsachi »

Dear Deeps92,
As one listening in to the mating calls of dolphins in the deep to the other who has taught them these cryptic sounds, Howdee!
Your goal, methodolology, and means are all obscure to me, but what the hell, I went ahead, played the clips in VLC at 0.4x speed, and answered your questions too. If you can find my answers in your drive, give me a shout.

Just in jest, if someone were to be listening to this kind of sounds and decoding them for a career, he would ardently wish for the anaahata Naada Bindu!

Now I very much wish to throw YOU a puzzle. Identify the raga based on a chime! Let me look for the video file... :D

deeps92
Posts: 8
Joined: 02 Mar 2016, 20:46

Re: Unique Information Security Approach using Carnatic Music

Post by deeps92 »

vasanthakokilam wrote:deeps92, I gave it a try. But I can not make out which raga each one of the clips belong to. May be someone with better swara gyana can do better.

btw, is the same text encoded in the various ragas or these are different texts for each of the clips? They all seemed to be similar in terms of the overall pattern.
I understand it is not very easy to identify the ragas from a sequence of pure notes with a high tempo.
Thank you for giving a try :)

The same text input has been encoded by using different ragas so as to make the comparison more accurate.

Rsachi
Posts: 5039
Joined: 31 Aug 2009, 13:54

Re: Unique Information Security Approach using Carnatic Music

Post by Rsachi »

Deeps92, and others,
Identify this raga!
http://j.mp/Chimeraga

deeps92
Posts: 8
Joined: 02 Mar 2016, 20:46

Re: Unique Information Security Approach using Carnatic Music

Post by deeps92 »

Rsachi wrote:Dear Deeps92,
As one listening in to the mating calls of dolphins in the deep to the other who has taught them these cryptic sounds, Howdee!
Your goal, methodolology, and means are all obscure to me, but what the hell, I went ahead, played the clips in VLC at 0.4x speed, and answered your questions too. If you can find my answers in your drive, give me a shout.

Just in jest, if someone were to be listening to this kind of sounds and decoding them for a career, he would ardently wish for the anaahata Naada Bindu!

Now I very much wish to throw YOU a puzzle. Identify the raga based on a chime! Let me look for the video file... :D
Thank you so much for taking out some time and filling out the form. It's a great help for me :)

I am glad that the methodology is obscure as that's the whole purpose of the security scheme I am designing :D

As far as the puzzle is concerned, I am more of a technical person with very basic knowledge of carnatic classical music so I can't guarantee an answer ;)

Thanks again!

Nick H
Posts: 9379
Joined: 03 Feb 2010, 02:03

Re: Unique Information Security Approach using Carnatic Music

Post by Nick H »

I'm no cryptologist, in fact, basic arithmetic is largely beyond me, but, so far as my slight understanding of cryptology and keys is concerned, they depend on very large numbers and very very-difficult-to-solve mathematical problems. Stuff that, even if you throw a supercomputer at it (Quantum computers, whatever they are, excepted, at least theoretically) it can't give you answer in a period of time that is practical.

Aren't there too few ragas?

deeps92
Posts: 8
Joined: 02 Mar 2016, 20:46

Re: Unique Information Security Approach using Carnatic Music

Post by deeps92 »

Nick H wrote:I'm no cryptologist, in fact, basic arithmetic is largely beyond me, but, so far as my slight understanding of cryptology and keys is concerned, they depend on very large numbers and very very-difficult-to-solve mathematical problems. Stuff that, even if you throw a supercomputer at it (Quantum computers, whatever they are, excepted, at least theoretically) it can't give you answer in a period of time that is practical.

Aren't there too few ragas?
Cryptography basically deals with transforming a meaningful message into something illegible so that only the intended person/entity is able to make any sense out of it. This can be achieved in multiple ways, one of which is what you just mentioned ( that deals with working with very large numbers).

Music cryptography is a branch of cryptography where a message in text/any_other_format is converted into a musical clip using various techniques and sent to the receiving end who then performs the reverse operations that were performed at the sender's side and gets back the original text message.

The method that I have designed can be generalized to any 'melakarta'/'sampurna' raga that has all 7 swaras. I wanted to confirm if the musical clips so generated are indeed valid in terms of conformance to a particular raga. For testing purpose I have just asked to verify these five ragas.

Quantum computers are currently not existent and once they do come into use, most of the present cryptographic schemes including the famous RSA scheme are likely to become stale.

deeps92
Posts: 8
Joined: 02 Mar 2016, 20:46

Re: Unique Information Security Approach using Carnatic Music

Post by deeps92 »

Currently the music cryptographic schemes that are in use produce just a sequence of notes. I am trying to propose a new methodology which produces a musical clip which is also a valid musical composition. Validity here corresponds to conformance to any particular raga. Basically, I want to test if the musical clip so generated does not contain any swara sequence which is considered as illegal in carnatic music.

I am not even sure if by making use of just the swaras present in a particular raga ( ex: Sa Re1 Ga3 Ma1 Pa Dha1 Ni3 in mayamalavagowla) , any sequence of notes produced will be a valid musical composition in that raga !!!! Is there something like "illegal" sequence of swaras with respect to each raga??

Rsachi
Posts: 5039
Joined: 31 Aug 2009, 13:54

Re: Unique Information Security Approach using Carnatic Music

Post by Rsachi »

Deeps,
1.If you want valid raga scales, please use an advanced instrument with the timbre and the correct pitch values. You will be best served by using the 22 shruti harmonium or the Gaayaka software. If you Google these terms, you will get the sites.
2. Don't call them ragas. They are merely scales. A raga has gamaka and specific sequences which seem irrelevant here.
3. If encryption requires using seven pitch positions, I wonder why you want any valid scale. Why not just seven monotonically increasing pitch values? Why not use very high (inaudible) pitch values? At this stage, with your five samples, I can say listening to these encrypted sounds is NO fun.

MaheshS
Posts: 1186
Joined: 02 Feb 2010, 22:36

Re: Unique Information Security Approach using Carnatic Music

Post by MaheshS »

deeps92 wrote: The basic idea is to convert a text message (secret between sender and receiver) into a music clip at the sender's side which will be sent to receiver. The receiver then deciphers the music clip back to the text message.
I have come up with a logic that does this conversion from English characters to a sequence of 'swaras'. The music clips generated as now are just sounds of specific frequencies (pure notes without 'gamakas' and other emotional aspects of music).
In this method, there is no need for people to know Carnatic Music (or for that matter ANY music) to decrypt. They can easily fix the note frequency and frequency of the notes and the job is done. Not pouring scorn on your idea, just mentioning this as your primary aim seems to be encryption / decryption, so I am not sure how valid this methodology is.

Nick H
Posts: 9379
Joined: 03 Feb 2010, 02:03

Re: Unique Information Security Approach using Carnatic Music

Post by Nick H »

They can easily fix the note frequency and frequency of the notes and the job is done.
Like the basic schoolboy code of just making letters match other letters, so, for instance, finding the most common letters in the code gives e, t, s, [assuming English] and so on. I suspect that deeps92 is a long way past that in his knowledge and thinking.

It occurs to me that I have met someone who was a cryptographic expert at, I think, highest levels of government. A rasika of course! How else would I ever get to meet eminent physicists and maths genii! No, I don't remember the name, but the carnatic world is not so huge: I bet someone is thinking, "I know who you mean!" as they read this.

MaheshS
Posts: 1186
Joined: 02 Feb 2010, 22:36

Re: Unique Information Security Approach using Carnatic Music

Post by MaheshS »

Nick H wrote:
They can easily fix the note frequency and frequency of the notes and the job is done.
Like the basic schoolboy code of just making letters match other letters, so, for instance, finding the most common letters in the code gives e, t, s, [assuming English] and so on.
Pretty much. If you do encrypt in the way deeps92 suggests, [letter] = [note] AND (to make it worse imo) for the resulting sequence be within a *defined* scope, of an already defined melakartha / raga / scale / gamaka / whatever, then the scope itself in computing terms is miniscule and can be cracked with an average PC. If it's just a straight forward substitution method then the security bit of the "Information Security" goes out of the window, regardless of the usage of musical notes / frequency. The higher / lower the frequency the better to go undetected is not true either.
Nick H wrote:I suspect that deeps92 is a long way past that in his knowledge and thinking.
I do not know, will see what deeps92 says :D

More stuff here.

VK RAMAN
Posts: 5009
Joined: 03 Feb 2010, 00:29

Re: Unique Information Security Approach using Carnatic Music

Post by VK RAMAN »

Giving ISIL, and terrorists ideas how to communicate through sounds and beats

deeps92
Posts: 8
Joined: 02 Mar 2016, 20:46

Re: Unique Information Security Approach using Carnatic Music

Post by deeps92 »

MaheshS wrote:
deeps92 wrote: The basic idea is to convert a text message (secret between sender and receiver) into a music clip at the sender's side which will be sent to receiver. The receiver then deciphers the music clip back to the text message.
I have come up with a logic that does this conversion from English characters to a sequence of 'swaras'. The music clips generated as now are just sounds of specific frequencies (pure notes without 'gamakas' and other emotional aspects of music).
In this method, there is no need for people to know Carnatic Music (or for that matter ANY music) to decrypt. They can easily fix the note frequency and frequency of the notes and the job is done. Not pouring scorn on your idea, just mentioning this as your primary aim seems to be encryption / decryption, so I am not sure how valid this methodology is.
I am currently doing my M.tech. in Information Security field at N.I.T.K.. From the little knowledge that I have gained working in the area of cryptographic field, frequency analysis (the type of cryptanalytic attack that you mentioned) requires that there has to be a "pattern" in which this mapping of plaintext to ciphertext (in this case sequence of swaras) happens. I have taken care of that aspect already. Frequency mapping requires one alphabet character to be mapped to one frequency value which is not the case in my method.

I could have disclosed my concept but since this is a public forum and the idea that I have come up with, is currently not existent I am afraid I cannot do so as of now.

deeps92
Posts: 8
Joined: 02 Mar 2016, 20:46

Re: Unique Information Security Approach using Carnatic Music

Post by deeps92 »

Rsachi wrote:Deeps,
1.If you want valid raga scales, please use an advanced instrument with the timbre and the correct pitch values. You will be best served by using the 22 shruti harmonium or the Gaayaka software. If you Google these terms, you will get the sites.
2. Don't call them ragas. They are merely scales. A raga has gamaka and specific sequences which seem irrelevant here.
3. If encryption requires using seven pitch positions, I wonder why you want any valid scale. Why not just seven monotonically increasing pitch values? Why not use very high (inaudible) pitch values? At this stage, with your five samples, I can say listening to these encrypted sounds is NO fun.
Currently my sole purpose is to just know if the sequence of musical notes that I have generated as the encryption output can infact be used in that order in a valid musical composition. Software generated music and usage of MIDI files for this purpose will cause the encryption output to be too large when compared to the size of the actual text message being transmitted. In order to reduce that overhead, I have used the pure notes generated using a software tool.

A gamaka as far as I know is an ornamentation to any musical composition and depends totally on the singer's manodharma. In the future work of this particula approach that I am using, I plan to include the gamaka features as well but before I do that I need to know that sequence of swaras that are currently generated conform to the framework of any raga.

There are already many approaches that have come up with ideas where in they are making use of just pitch positions and the final musical clip generated does not seem to follow any music rules.

And as far as the FUN in listening is concerned, I have no intention of producing a musical composition that is ready to be used in a concert. My only purpose is to encrypt the message !!

deeps92
Posts: 8
Joined: 02 Mar 2016, 20:46

Re: Unique Information Security Approach using Carnatic Music

Post by deeps92 »

Everybody seems to be curious to know the methodology that I have employed in achieving this task. I would be happy to disclose it once I get a permit from my guide in college. For that I require that this work gets approval at the earliest. That means I will have to give a report of the survey of conformance of the musical clips to the ragas that I have mentioned which in turn requires all of your support in filling out the form !! :D :D

Please kindly fill those forms so that I can complete my task. Once that is done, I will be able to publish my work !!!


Thanks

Nick H
Posts: 9379
Joined: 03 Feb 2010, 02:03

Re: Unique Information Security Approach using Carnatic Music

Post by Nick H »

VK RAMAN wrote:Giving ISIL, and terrorists ideas how to communicate through sounds and beats
We all use encryption every day, so it is not giving anything more to anyone who wants to use it for nefarious purposes of any sort.

Research in these matters is important. I guess it is useful exercise for the brains of the students, whether they come up with anything useful or not, but, as the seeking to break through our privacy and security are also working hard, the cryptography researchers must stay ahead.

So, it matters that the students are having ideas and working on them. It matters less whether or not the ideas turn out to be good ones or not.

sureshvv
Posts: 5523
Joined: 05 Jul 2007, 18:17

Re: Unique Information Security Approach using Carnatic Music

Post by sureshvv »

I recommend that you employ the full power of CM - use the RTP. With a pallavi and the choice of tala which can be the key to decrypt the message you may simultaneously be able to encrypt & compress your original message for safe & fast transmission.

VK RAMAN
Posts: 5009
Joined: 03 Feb 2010, 00:29

Re: Unique Information Security Approach using Carnatic Music

Post by VK RAMAN »

Most creative and inventive are children and they will figure a way to keep their parents blind how they communicate with outsiders and criminals.

Nick H
Posts: 9379
Joined: 03 Feb 2010, 02:03

Re: Unique Information Security Approach using Carnatic Music

Post by Nick H »

Oh well, that's the end of the world then! It was nice while it lasted. Well, bits of it were nice. Sometimes.

MaheshS
Posts: 1186
Joined: 02 Feb 2010, 22:36

Re: Unique Information Security Approach using Carnatic Music

Post by MaheshS »

The method that I have designed can be generalized to any 'melakarta'/'sampurna' raga that has all 7 swaras. I wanted to confirm if the musical clips so generated are indeed valid in terms of conformance to a particular raga.
From the little knowledge that I have gained working in the area of cryptographic field, frequency analysis (the type of cryptanalytic attack that you mentioned) requires that there has to be a "pattern" in which this mapping of plaintext to ciphertext (in this case sequence of swaras) happens. I have taken care of that aspect already. Frequency mapping requires one alphabet character to be mapped to one frequency value which is not the case in my method.
I am trying to propose a new methodology which produces a musical clip which is also a valid musical composition. Validity here corresponds to conformance to any particular raga. Basically, I want to test if the musical clip so generated does not contain any swara sequence which is considered as illegal in carnatic music.
Currently my sole purpose is to just know if the sequence of musical notes that I have generated as the encryption output can infact be used in that order in a valid musical composition.
And as far as the FUN in listening is concerned, I have no intention of producing a musical composition that is ready to be used in a concert. My only purpose is to encrypt the message !!
I am glad that the methodology is obscure as that's the whole purpose of the security scheme I am designing
So, from what I can understand, you are trying to encrypt something and the result being a set of notes that is recognisable as a valid scale [raga / melakartha / etc] in carnatic music. I guess that's the "key" you are using. As soon as the "evesdropper" listens to the encrypted MP3 or whatever, it'll only take a little time to figure out its just a bunch of notes without any musical value and therefore it has another purpose [ie, hidden message]. So security by obscurity is all nice and good but of little protection in this day and age.

Once they've recognised that there is a hidden message, it only takes about the same time as you'd drink a cup of coffee to break that.

Also, I don't understand how you say there is no pattern when you also say that it must confirm to a scale :) Example, using say Mayamalagowla as a key, generate 100K encrypted messages and see for yourself if there is a pattern or not. The hard part is finding out the key and you've restricted yourself to carnatic music scales. I re-iterate that it's a *defined* scope, therefore music knowledge for breaking this is just optional.

Sidenote : I am speaking as a guy who works in "Info Sec" with the weird and wonderful people that are cryptos / whitehats, I've just run this past them as well. University thesis Yes, in a product or for professional use, No.

Nick H
Posts: 9379
Joined: 03 Feb 2010, 02:03

Re: Unique Information Security Approach using Carnatic Music

Post by Nick H »

its just a bunch of notes without any musical value
I've heard lots of stuff like that; often, people claim that it is music! :lol:

But "random" notes at high speed. Hmmm, reminds me of the old modems... yes, I think people would guess that it was a data stream

Post Reply