Arun, thanks. Neat. Sounds good.
The one thing I noticed with KVN oscillation is: The first up and down is from Sa to Da2 and the second one is from Sa to Ni2. Yours is more uniform from Sa to Da in both. I do not know if that made any net difference.
BTW, you had some extras at the end. That is interesting. The Da oscillations are there for those two as well, not that different from the Begada Ni oscillation but it still sounds different. The last one was quite close to the Begada Ni of KVN in Sonic Visualizer.
Pitch Analysis - A fun tool to investigate Shruti in CM
-
uday_shankar
- Posts: 1475
- Joined: 03 Feb 2010, 08:37
Whoa ?! This universe has changed.
I don't know where I left this or what I promised but it all look like Greek and Latin now. You guys are indefatigable and I certainly can't keep pace. Please etchoos me.
Meanwhile I'm noticing a lot said about Begada N2/N3, etc..
I recommend the following all-you-wanted-to-know-about-Begada album, including discussions (Tamil and English) on shades of M1 and N by two of the greatest stalwarts, SSI and Ravikiran:
http://www.ravikiranmusic.com/Webfiles/ ... ls_vv2.htm
Here's an artificially constrained N3-only version of Begada:
http://www.sendspace.com/file/61lpr6
And here's an N2-only version:
http://www.sendspace.com/file/yqmr6r
As you can see it is very contrived to exclusively use only one of the N's at any time. Actually it is easier to play the N3-only version. Also, my flute-playing could do with more gamakas, but sorry no practice so please also etchoos
.
I don't know where I left this or what I promised but it all look like Greek and Latin now. You guys are indefatigable and I certainly can't keep pace. Please etchoos me.
Meanwhile I'm noticing a lot said about Begada N2/N3, etc..
I recommend the following all-you-wanted-to-know-about-Begada album, including discussions (Tamil and English) on shades of M1 and N by two of the greatest stalwarts, SSI and Ravikiran:
http://www.ravikiranmusic.com/Webfiles/ ... ls_vv2.htm
Here's an artificially constrained N3-only version of Begada:
http://www.sendspace.com/file/61lpr6
And here's an N2-only version:
http://www.sendspace.com/file/yqmr6r
As you can see it is very contrived to exclusively use only one of the N's at any time. Actually it is easier to play the N3-only version. Also, my flute-playing could do with more gamakas, but sorry no practice so please also etchoos
.
Last edited by Guest on 30 Jan 2008, 17:31, edited 1 time in total.
-
arunk
- Posts: 3424
- Joined: 07 Feb 2010, 21:41
Yep. In spite of many attempts couldnt match it exactly - especially the peaks. Now, while I felt the attempted nis were ok and in the realm of begada, I did have the feeling "close, perhaps close enough but still a bit different".vasanthakokilam wrote:Arun, thanks. Neat. Sounds good.
The one thing I noticed with KVN oscillation is: The first up and down is from Sa to Da2 and the second one is from Sa to Ni2. Yours is more uniform from Sa to Da in both. I do not know if that made any net difference.
:rolleyes: Oops! Then this wasnt the sample I meant to post! This was just trying to see how the other das (the few I could relate to) matched up with the ni here, i.e. the ni which is mostly da-sa combo. It was supposed to be a "private experiment" - i.e. private between pitchappa and IBTW, you had some extras at the end. That is interesting. The Da oscillations are there for those two as well, not that different from the Begada Ni oscillation but it still sounds different. The last one was quite close to the Begada Ni of KVN in Sonic Visualizer.
Last edited by arunk on 30 Jan 2008, 20:30, edited 1 time in total.
-
uday_shankar
- Posts: 1475
- Joined: 03 Feb 2010, 08:37
-
vasanthakokilam
- Posts: 10958
- Joined: 03 Feb 2010, 00:01
Uday: I also just listened to it and it sounded good.
The begada conversation ventured into the nature of the NI as you know. Our general theory of understanding is that different oscillations produce the perception of the N2 or N3 effect. So when you consciously played N2 or N3, were you engaged in that kind of different Begada specific oscillations to produce Begada N2 and Begada N3 or the regular N2 and N3 gamakas that are used in other ragas? Just trying to gather data points for that 'perception of average pitch' hypothesis/theory.
The begada conversation ventured into the nature of the NI as you know. Our general theory of understanding is that different oscillations produce the perception of the N2 or N3 effect. So when you consciously played N2 or N3, were you engaged in that kind of different Begada specific oscillations to produce Begada N2 and Begada N3 or the regular N2 and N3 gamakas that are used in other ragas? Just trying to gather data points for that 'perception of average pitch' hypothesis/theory.
-
uday_shankar
- Posts: 1475
- Joined: 03 Feb 2010, 08:37
Ok, vk, here's some very cold watervasanthakokilam wrote:'perception of average pitch' hypothesis/theory.
.Much of the "perception of average pitch" has to do with Carnatic training, and has nothing to do with any intrinsic property of the ear or psycho-acoustics. So it is mostly nonsense.
I don't know if Arvindh or anybody else has done experiments with the following kind of subject : "highly perceptive to pitch oscillations but absolutely no background in Carnatic music". Only such a subject can really tell us about the perception of average pitch in Carnatic music.
I say some of this from personal experience. I got "trained" a bit in Carnatic music long after I started playing insturments, etc... and I am somewhat non-standard in my perception of Carnatic music in that I don't really "view" any kampita gamakas as notes but rather as just oscillations. While notating I just play along because that's the norm.
For example, the R2-M1 gamaka in anandha bhairavi or any other raga gives me absolutely no sense of G2. It only falls in my ear as an R2-M1 gamaka. To fit in with the rest of the crowd I notate it as G2
.I was disappointed when a highly learned Carnatic musician was laughing at some Hindustani musician for not being able to understand his lec-dem on kampita gamakam. He apparently demonstrated the G2 in kanada with the R2-M1 oscillation. At the end of the dem, the Hindustani musician walked up to the Carnatic musician and said his G2 demo reminded him of the R2-M1 oscillation in Durga. That apparently was a laughing matter for the Carnatic musician. But really, the Carnatic musician is the fool of the story. The Hindustani musician accurately charatecterized the G2 kampita as being similar to the R2 osicillation Hindustani musicians employ while singing Durga (S R2 M1 P D2 S). Of course, like me, the Hindustani musician really didn't hear ANY G2 in that R2-M1 gamaka. So there goes the "average pitch" theory out the window.
The same thing holds for Begada also. There are S-D2-S type of oscillations which are all notated as N2 and also there's open playing of N2 in phrases such as P P D N2 D P, etc... (in fact that N2 is a S-N2 type of oscillation which in your average pitch theory should be N3!!
Basically you'll be on a wild goose chase if you're looking for a mathematical analog for what's just basically a Carnatically prejudiced ear.
Here's Uday's law (to be experimentally verified):
Not a single musician in the world who is not trained in Carnatic music will percive an average pitch between the extremities of a kampita gamakam. Therefore there's nothing universal and mathematically provable about the perception of an "average" pitch in a kampita gamakam.
Exceptions to the law may be in the case of very quickly executed oscillations like the Saveri R1 touching a note higher than R1 (although to my ear even that sounds like exactly what it is, i.e. a wrong note !!).
Last edited by Guest on 01 Feb 2008, 12:03, edited 1 time in total.
-
vasanthakokilam
- Posts: 10958
- Joined: 03 Feb 2010, 00:01
Uday:
I think I understand what you are saying. I will let some one else who has better expertise in psychoacoustical research comment on the average pitch hypothesis or quote material on that topic which we can read up.
But definitely, even if there is something to the average pitch, it is not same as the swarasthanam since the oscilations definitely provides a different tonal color.
Personally I am game for the concept that there are pitch contours that depict the oscillations and the oscillations are defined by the trough frequency and high frequency, ( whether they map to a swarasthanam or not ), period of oscillations and number of oscillations. They all seem to affect the tone's color. Then it is just one extra step to have a compressed name to denote syntactically that contour. Instead of inventing a new name, we are just using the familiar solfa name of the swara.
Check me on these.
1) Both of your N2 and N3 based pieces are oscillatory Nis that are characteristics of Begada, right?
2) What oscillations are you labeling as N2 and N3 ( to fit in with the rest of the world, quoting you )
I think I understand what you are saying. I will let some one else who has better expertise in psychoacoustical research comment on the average pitch hypothesis or quote material on that topic which we can read up.
But definitely, even if there is something to the average pitch, it is not same as the swarasthanam since the oscilations definitely provides a different tonal color.
Personally I am game for the concept that there are pitch contours that depict the oscillations and the oscillations are defined by the trough frequency and high frequency, ( whether they map to a swarasthanam or not ), period of oscillations and number of oscillations. They all seem to affect the tone's color. Then it is just one extra step to have a compressed name to denote syntactically that contour. Instead of inventing a new name, we are just using the familiar solfa name of the swara.
Check me on these.
1) Both of your N2 and N3 based pieces are oscillatory Nis that are characteristics of Begada, right?
2) What oscillations are you labeling as N2 and N3 ( to fit in with the rest of the world, quoting you
)-
arunk
- Posts: 3424
- Joined: 07 Feb 2010, 21:41
I was afraid the term "average" in average pitch was poor, which is why I mentioned "loosely used" earlier.
However, I certainly can believe that even this concept which leads to statement like "ni in that raga is higher than N2" or "da2 in that raga is slightly lower than typical d2", is a cm prejudiced/colored perception, and thus looking for scientific basis for it in terms of pitch analysis would be a wild-goose chase.
One possibility (ok, i must admit an extempore wild theory cooked up in the last 10 minutes - ) why we even label some of these pitch movements by a swara, even though the movement may not rest prominently on the swarasthana could be this: Take kAnaDa - a common use of the ga is in the context pg~mrs, here compared to the movement labelled ga, pa ma and ri are flat. Their ascending order is r . m p. You have two things, (a) the pitch movement for g~ here moves between r and m, both those ends being also rendered flat (b) IIRC, but kanada ga also sometimes does emphasize the region of g2 (even if not flat, and not necessarily in this context?) - so in the entire gamut of kanada, you have a pitch movement between r and ma, that sometimes emphasizes the region of ga2 (rather than say ga3). So the cm mind, leans to associating that with ga2 in all contexts.
The same applies I think may applly to aTANa (i.e. the ga that is labelled as g2, rather than the flat g3 in tara stayi, which is clearly g3).
In begada you have g-m-p-d-n~. Also n s. Could explain why the n~ pitch movement is labelled n (as opposed to say d~)?
Of course as is with "influenced perceptions", there probably wont be uniform logic here that can be universally applied to all other ragas and contexts. But I wonder if these are at least part of the reasons why people have these perceptions (other more dominating ones: how people learn, how they are taught - i learnt that these kinds of movements are usually labelled ni - so simply learnt that way, made sense to our (subconsciously) influenced perceptions and hence etc. etc.)
Arun
However, I certainly can believe that even this concept which leads to statement like "ni in that raga is higher than N2" or "da2 in that raga is slightly lower than typical d2", is a cm prejudiced/colored perception, and thus looking for scientific basis for it in terms of pitch analysis would be a wild-goose chase.
One possibility (ok, i must admit an extempore wild theory cooked up in the last 10 minutes -
) why we even label some of these pitch movements by a swara, even though the movement may not rest prominently on the swarasthana could be this: Take kAnaDa - a common use of the ga is in the context pg~mrs, here compared to the movement labelled ga, pa ma and ri are flat. Their ascending order is r . m p. You have two things, (a) the pitch movement for g~ here moves between r and m, both those ends being also rendered flat (b) IIRC, but kanada ga also sometimes does emphasize the region of g2 (even if not flat, and not necessarily in this context?) - so in the entire gamut of kanada, you have a pitch movement between r and ma, that sometimes emphasizes the region of ga2 (rather than say ga3). So the cm mind, leans to associating that with ga2 in all contexts.The same applies I think may applly to aTANa (i.e. the ga that is labelled as g2, rather than the flat g3 in tara stayi, which is clearly g3).
In begada you have g-m-p-d-n~. Also n s. Could explain why the n~ pitch movement is labelled n (as opposed to say d~)?
Of course as is with "influenced perceptions", there probably wont be uniform logic here that can be universally applied to all other ragas and contexts. But I wonder if these are at least part of the reasons why people have these perceptions (other more dominating ones: how people learn, how they are taught - i learnt that these kinds of movements are usually labelled ni - so simply learnt that way, made sense to our (subconsciously) influenced perceptions and hence etc. etc.)
Arun
Last edited by arunk on 01 Feb 2008, 20:01, edited 1 time in total.
-
uday_shankar
- Posts: 1475
- Joined: 03 Feb 2010, 08:37
Bingo.vasanthakokilam wrote:Then it is just one extra step to have a compressed name to denote syntactically that contour. Instead of inventing a new name, we are just using the familiar solfa name of the swara.
This seems to be a good theory too. Not so wild, maybe. BTW, Arun, your N3 sounds better than KVN's thanks to a more "curvy" version and an added kaarvaiwild theory cooked up in the last 10 minutes
.
Thanks. But then I thought I was trying to mimic an N2 version (or say closer to it) - Oops!-
vasanthakokilam
- Posts: 10958
- Joined: 03 Feb 2010, 00:01
Back to Sonic Visualizer, try the Aubio Tempo Detector. It misses more than it gets but when it gets going it is quite impressive that it keeps the thalam. I tried it with a varnam. Give it a try.
That raises a question. I have been wondering ( and have been asking too ) what gives us the instinctive feel for a beat/stress that makes us keep the beat ( say, no percussion and no one else is keeping the thalam.)
The stress is in the music. In physical terms what qualfies as stress? Frequency defines Melody, Amplitude defines volume and loudness, what defines Stress? Some combination of Frequency and Amplitude?
It has to be something quantifiable since the tool seems to be pick it up ( when it works ). What is the theory behind such tempo detection? Is that what we use in our minds to detect stress/beat ?
That raises a question. I have been wondering ( and have been asking too ) what gives us the instinctive feel for a beat/stress that makes us keep the beat ( say, no percussion and no one else is keeping the thalam.)
The stress is in the music. In physical terms what qualfies as stress? Frequency defines Melody, Amplitude defines volume and loudness, what defines Stress? Some combination of Frequency and Amplitude?
It has to be something quantifiable since the tool seems to be pick it up ( when it works ). What is the theory behind such tempo detection? Is that what we use in our minds to detect stress/beat ?
-
arunk
- Posts: 3424
- Joined: 07 Feb 2010, 21:41
-
vasanthakokilam
- Posts: 10958
- Joined: 03 Feb 2010, 00:01
-
cmlover
- Posts: 11498
- Joined: 02 Feb 2010, 22:36
My computer Hard disk has crashed 
I seem to have lost everything including all e-mails and discussions. Some experts are working on recovery but are not optimistic. Hence I may be out for quite some time in mourning
GOOD BYE! (for a while! It was a wonderful time, but now I feel stranded !
Posting from my son's laptop!
I seem to have lost everything including all e-mails and discussions. Some experts are working on recovery but are not optimistic. Hence I may be out for quite some time in mourning
GOOD BYE! (for a while! It was a wonderful time, but now I feel stranded !
Posting from my son's laptop!
-
vasanthakokilam
- Posts: 10958
- Joined: 03 Feb 2010, 00:01
I read the article and another article this one refers to. My idea was to find out what approximation of the feeling of beat can be used to quantity what a stress/beat/impulse is.
- Though the algorithms vary, the fundamental principle is that a stress/beat onset is characterzed by a sudden increase in sound energy at that instant compared to the average intensity of the neighborhood. Some algorithms use this fact directly, and others try to be 'smart' about it using the frequency domain as well.
- The author of the above article seems to suggest he is detecting beat based on the sound produced by the drum. That was initially troubling since my idea was to detect the built-in rhythm of the piece itself and not the manifestation of it externally with a drum. But it turns out that is just the way the author wrote the article, the algorithm seems to apply to music without any percussion support. ( correct me on this ).
- Will I be wrong in equating sound energy to amplitude. Meaning, a stress point is characterized basically by a sudden increase in amplitude.
- Is amplitude based detection is how we comprehend/sense the pulses in a song? Not sure but seems to be so from these articles.
- I had previously talked to a premiere speech recognition expert on what algorithm he uses to detect the 'stress' pattern in speech. He said that contrary to intutive belief, specific frequency patterns play the major role and the amplitude plays a secondary role in detecting 'stresses' in speech. ( say for distinction between ˌa-pə-ˈra-təs and ˌa-p'pa-ra-təs for apparatus ). In beat sensing of music, the algorithms use frequency only to optimize the other wise Amplitude based detection.
Read this short description: http://www.owlnet.rice.edu/~elec301/Pro ... round.html
It makes this point quite convincingly. That is, all you need is the amplitude envelope and not frequency except to segregate the amplitude envelope by frequency sub-band and processing the amplitudes in each band separately.
The thing that is convincing to me that amplitude alone is sufficient for music stress detection is the point this link makes: "a modulated noise signal created from a musical signal has a very similar rhythmic percept as the original signal." Putting it in our own terms based on our own experience, we can still keep the thalam when there are road noises in those road side concerts
- I tried the various algorithms available in Sonic Visualizer for some music with no percussion support. Results vary quite a lot. There are sections where it is perfect and in other places it gets confused for reasons not apparent to me yet.
- This raises another question outside of beat detection. Is there a kind of technique practised in CM ( say a gamaka or a rasa ) where the amplitude is kept more or less the same as the frequency is increased? (I can imagine the other way, keepting the same note and increasing the amplitude is possible and is done in some cases as well ). Meaning, with normal singing, a higher frequency seems to be accompanied by increase in amplitude as well. So, in some cases, the stress points occur in places where both the frequency and amplitude change together. This still does not change the main thesis of beat detection since beats are obviously there in avarohana prayogams as well. But I am curious what the aesthetic effect is when the frequency is increased without an accompanying increase in amplitude ( say in an arohana prayogam ).
- Though the algorithms vary, the fundamental principle is that a stress/beat onset is characterzed by a sudden increase in sound energy at that instant compared to the average intensity of the neighborhood. Some algorithms use this fact directly, and others try to be 'smart' about it using the frequency domain as well.
- The author of the above article seems to suggest he is detecting beat based on the sound produced by the drum. That was initially troubling since my idea was to detect the built-in rhythm of the piece itself and not the manifestation of it externally with a drum. But it turns out that is just the way the author wrote the article, the algorithm seems to apply to music without any percussion support. ( correct me on this ).
- Will I be wrong in equating sound energy to amplitude. Meaning, a stress point is characterized basically by a sudden increase in amplitude.
- Is amplitude based detection is how we comprehend/sense the pulses in a song? Not sure but seems to be so from these articles.
- I had previously talked to a premiere speech recognition expert on what algorithm he uses to detect the 'stress' pattern in speech. He said that contrary to intutive belief, specific frequency patterns play the major role and the amplitude plays a secondary role in detecting 'stresses' in speech. ( say for distinction between ˌa-pə-ˈra-təs and ˌa-p'pa-ra-təs for apparatus ). In beat sensing of music, the algorithms use frequency only to optimize the other wise Amplitude based detection.
Read this short description: http://www.owlnet.rice.edu/~elec301/Pro ... round.html
It makes this point quite convincingly. That is, all you need is the amplitude envelope and not frequency except to segregate the amplitude envelope by frequency sub-band and processing the amplitudes in each band separately.
The thing that is convincing to me that amplitude alone is sufficient for music stress detection is the point this link makes: "a modulated noise signal created from a musical signal has a very similar rhythmic percept as the original signal." Putting it in our own terms based on our own experience, we can still keep the thalam when there are road noises in those road side concerts
- I tried the various algorithms available in Sonic Visualizer for some music with no percussion support. Results vary quite a lot. There are sections where it is perfect and in other places it gets confused for reasons not apparent to me yet.
- This raises another question outside of beat detection. Is there a kind of technique practised in CM ( say a gamaka or a rasa ) where the amplitude is kept more or less the same as the frequency is increased? (I can imagine the other way, keepting the same note and increasing the amplitude is possible and is done in some cases as well ). Meaning, with normal singing, a higher frequency seems to be accompanied by increase in amplitude as well. So, in some cases, the stress points occur in places where both the frequency and amplitude change together. This still does not change the main thesis of beat detection since beats are obviously there in avarohana prayogams as well. But I am curious what the aesthetic effect is when the frequency is increased without an accompanying increase in amplitude ( say in an arohana prayogam ).
-
arunk
- Posts: 3424
- Joined: 07 Feb 2010, 21:41
Sorry to hear about that cmlover. A few years ago, something similar happened to my hard disk. I sent it to some company in Florida (dont remember the name now). I had zero hopes given the horrible noise the drive was making when the computer started. But I was ecstatic when the guy sent me a mail saying he had recovered every file and directory that I had wanted! So dont lose hope - sometimes these guys can work miracles.
-
arunk
- Posts: 3424
- Joined: 07 Feb 2010, 21:41
Most gamakas (if not all?) are pitch modulation - I think there may be a bit of amplitude modulation alongside - sometimes accidentally (or say as a sideeffect), sometimes intentionally (nokku type)
The speech detection - I think there is thing called formants i.e. frequency bands which have highest energy - even within same pitch. So a "a" followed by a "i" we can tell it apart even if the same pitch - that has something to do with formants. I dont know much about this.
Arun
The speech detection - I think there is thing called formants i.e. frequency bands which have highest energy - even within same pitch. So a "a" followed by a "i" we can tell it apart even if the same pitch - that has something to do with formants. I dont know much about this.
Arun
-
vasanthakokilam
- Posts: 10958
- Joined: 03 Feb 2010, 00:01
CML, very sorry to hear about this.
There are companies in India that do it for cheaper than U.S. but it is still not cheap ( I remember being quoted a few hundreds of dollars ).
The obvious answer is back up, back up, back up but given that we can easily have terabytes of online data, the backups have to be that much bigger and takes a lot of time making it not so feasible. But fortunately, music collections do not change that much ( except for additions ), so we can be a bit smart about backups.
But one has to be quite obsessive about backing up the "application data" folders, itunes meta folders, email folders etc.
Though I am not a Mac user currently, I think the Time Machine/Time Capsule features of the latest Mac OS X Leopard is a great consumer level feature . It does something without much fuss that most consumers( including me ) do not pay attention to the most significant aspect of using computers: Backup.
My own backup strategy is quite minimal. I keep some important data in my internet based email ( like hotmail or gmail ) folders, on the hope they will be backing stuff up for me!! I still do not have backups for my photos, itunes folders etc.
There are companies in India that do it for cheaper than U.S. but it is still not cheap ( I remember being quoted a few hundreds of dollars ).
The obvious answer is back up, back up, back up but given that we can easily have terabytes of online data, the backups have to be that much bigger and takes a lot of time making it not so feasible. But fortunately, music collections do not change that much ( except for additions ), so we can be a bit smart about backups.
But one has to be quite obsessive about backing up the "application data" folders, itunes meta folders, email folders etc.
Though I am not a Mac user currently, I think the Time Machine/Time Capsule features of the latest Mac OS X Leopard is a great consumer level feature . It does something without much fuss that most consumers( including me ) do not pay attention to the most significant aspect of using computers: Backup.
My own backup strategy is quite minimal. I keep some important data in my internet based email ( like hotmail or gmail ) folders, on the hope they will be backing stuff up for me!! I still do not have backups for my photos, itunes folders etc.
-
vasanthakokilam
- Posts: 10958
- Joined: 03 Feb 2010, 00:01
Understood.Most gamakas (if not all?) are pitch modulation - I think there may be a bit of amplitude modulation alongside - sometimes accidentally (or say as a sideeffect), sometimes intentionally (nokku type)
I am using gamaka in a more general sense as in the 10 types of gamaka which includes Arohanam and Avarohanam. I am wondering what will be the effect of singing sa-pa-SA keeping the amplitude the same.
-
wbmdubai
- Posts: 4
- Joined: 09 Mar 2009, 13:17
Friends,
Your discussion is taking a very interesting course. Today I posted some views in the Technical section, titled, Music Alphabet, which please read.
The Carnatic or Hindustani music fundamentals are still CLASSICAL and not SCIENTIFIC. We need to have standard instruments and fixed frequency fouriers for swarams and shrutis. Depending on the hearing capacity of judges and teachers misguides the singer.
Secondly, we need to have a music alphabet having say 49 alphabets to produce the 49 frequencies of 7 swaras x 7 scales. Repeating the same sapta swaras in different frequencies in different scales is totally confusing and unscientific. But that will be only a beginning. Enjoying music is not a universally standard phenomenon. Depending on the culture, up-bringing and personal tastes of the listener, the enjoyable wavelengths (frequencies) differ from person to person. But still we can have some scientific standards, so that learning could be made lot easier. Present classical system is so complicated and confusing that it takes decades to graduate to a middle level vocalist.
Venugopalan Menon ([email protected])
Your discussion is taking a very interesting course. Today I posted some views in the Technical section, titled, Music Alphabet, which please read.
The Carnatic or Hindustani music fundamentals are still CLASSICAL and not SCIENTIFIC. We need to have standard instruments and fixed frequency fouriers for swarams and shrutis. Depending on the hearing capacity of judges and teachers misguides the singer.
Secondly, we need to have a music alphabet having say 49 alphabets to produce the 49 frequencies of 7 swaras x 7 scales. Repeating the same sapta swaras in different frequencies in different scales is totally confusing and unscientific. But that will be only a beginning. Enjoying music is not a universally standard phenomenon. Depending on the culture, up-bringing and personal tastes of the listener, the enjoyable wavelengths (frequencies) differ from person to person. But still we can have some scientific standards, so that learning could be made lot easier. Present classical system is so complicated and confusing that it takes decades to graduate to a middle level vocalist.
Venugopalan Menon ([email protected])