duration per swaram for 1st, 2nd and 3rd kalams

[sung]

I have a very, very fundamental question for which I would really, really like to get the answer from the experts here.

This is about singing a given piece of music in 1st, 2nd and 3rd kalams. I always assumed that the duration per swaram gets exactly halved in going from 1st kalam to 2nd kalam and that it again gets exactly halved in going from 2nd kalam to 3rd kalam. For example, let us assume that it takes 0.68 second for 1 swaram in the 1st kalam. I assumed that it should take 0.34 s per swaram in the 2nd kalam and 0.17 s per swaram in the 3rd kalam.

But, recently I bought S Sowmya's Sahityanubhava 2004 Ata tala, tana and pada varnams CD at one of Cleveland Tyagaraja Aradhana booths. When I timed the mukthayi swaram for evvari (abhogi) varnam, I got a total of 20, 14 and 11 seconds for a total of 64 swaras for the 1st, 2nd and 3rd kalams, respectively, yielding 0.31, 0.22 and 0.17 s per swaram, respectively. This is not what I expected. With 0.17 s for 3rd kalam, it is less than half of what I thought it should be for 1st kalam, namely 0.31 s instead of 0.68 s, and the 2nd kalam is somewhere in between, somewhat closer to 3rd kalam than to 1st kalam. I am puzzled. Can the experts here please help me get over this puzzle and get it cleared? Thank you very much in advance.

[vasanthakokilam]

Allowing for small deviations, she seemed to have sung
Trisram in the middle. So it is 4 swaras per beat (chaturasram), 6 swaras per beat ( Trisram ) and 8 swaras per beat ( double speed chaturasram )

Experts please correct/comment

[sung]

Allowing for small deviations, she seemed to have sung
Trisram in the middle. So it is 4 swaras per beat (chaturasram), 6 swaras per beat ( Trisram ) and 8 swaras per beat ( double speed chaturasram )

Thank you very much (my intuition was correct, which said you might be the first one to respond; may God bless you!). That makes sense now. I wasn't considering trisram as a possibility at all.

In the case of chaturasram (4 swaras per beat) it takes 1.25 s per beat for Sowmya and in the case of the double speed chaturasram (8 swaras per beat) it takes 1.38 s per beat. They appear to be considerably different. In the case of trisram (6 swaras per beat) it takes 1.31 s per beat. They are all supposed to be the same value. Are such deviations normal even in the case of professionals such as Sowmya? So, the mridangist has to correspondingly adjust also?

If we use 0.17 s per swaram for, say, 3rd kalam, then the corresponding 1st kalam with 0.68 s per swaram appears to be extremely slow. If we use 0.34 s per swaram for, say, 1st kalam, then the corresponding 3rd kalam with 0.085 s per swaram seems to be extremely fast. So, do people normally sing only chaturasram (with 0.34 s per swaram) and double speed chaturasram (with 0.17 s per swaram), without triple speed chaturasram (with 0.085 s per swaram)? I thought that the professionals sing all 3 speeds. Clarification on this also by the experts will be very, very helpful to me.

[Nick H]

I am useless at numbers, but I think I get the idea of your question and VK's answer. I think it is right to say that "trikalam" usually includes tisra.

[vasanthakokilam]

Sung, I was going by your reported times of 20, 14 and 11 seconds

That is C, 2T and 2C because the ratios of those three speeds will be
20 * 4 /4 : 20 * 4/6 : 20 * 4/8 = 20 : 13.33 : 10

Close enough.

[sung]

Sung, I was going by your reported times of 20, 14 and 11 seconds

That is C, 2T and 2C because the ratios of those three speeds will be
20 * 4 /4 : 20 * 4/6 : 20 * 4/8 = 20 : 13.33 : 10

Close enough.

Exactly, VK. Of course, that is another equivalent mathematical way to look at the same thing. So, my question is: Is 11 s considered close enough to 10 s in the case of 2C even for professional singers? I thought they would keep the same exact tempo throughout.

[sung]

I am useless at numbers, but I think I get the idea of your question and VK's answer. I think it is right to say that "trikalam" usually includes tisra.

Really. I didn't know that. It is a very useful information for me. Thank you very much.

[vasanthakokilam]

>So, my question is: Is 11 s considered close enough to 10 s in the case of 2C even for professional singers?

Good question. I do not really know, hope some practicing musicians will comment on that. But I did some timing for a different purpose and based on the variations I observed, I will say Sowmya's numbers are pretty good.

[Nick H]

I thought they would keep the same exact tempo throughout.

Yes in theory. However, humans are not metronomes, and whilst I think that perfect machine-accurate timing might occasionally be done for demonstration purposes, real life is different.

[cacm]

>So, my question is: Is 11 s considered close enough to 10 s in the case of 2C even for professional singers?

Good question. I do not really know, hope some practicing musicians will comment on that. But I did some timing for a different purpose and based on the variations I observed, I will say Sowmya's numbers are pretty good.

I can say that in my discussions with FLUTE MALI & LGJ on this PRECISE SUBJECT BOTH OF THEM said that for them 11s referred to being close enough to 10s for THEM WAS NOT ACCEPTABLE. I had actually verified it in 1950's with expts at presidency college physics dept.
However I cannot CLAIM they WERE ALWAYS ABLE TO MAINTAIN IT ALWAYS! VKV

[vasanthakokilam]

Agreed that different people have different standards of accuracy on those theoretical and quantifiable measures and that is awesome. But a general point about all this can be made without negating what VKV wrote above.

While such accuracy does matter, they matter only up to a point. The criterion is 'Does a swara interval of 10/64 vs 11/64 matter for the musical aesthetics?' In true CM tradition such theoretical measures are given equal and some times greater status than un-quantifiable musical and artistic aspects. But we should reflect on the fact that such details are a means to an end and they matter only to the extent they contribute to that end.

Particularly troubling is the fact that people are too quick to invoke a principle like 'laya pitha' in such contexts. Implying that such a statement vouches for the high precision tempo keeping requirement when all that really meant to say is that laya is a major pillar of music. Which of course it is. But when such a musical statement is culturally misappropriated into a dry theoretical statement about tempo accuracy, a principle like 'emotional rhythm' struggles to find a place let alone be considered 'traditionally correct'. That is the kind of a thing that distinguishes human produced music from machine produced music. That is, if a small variation in tempo, however subconsciously produced, jives with and enhances the emotional aspects of the music, and if it connects with the audience at that emotional and aesthetic level, that is all should matter and not whether that change is mathematically accurate to an arbitrary precision or not.

[sung]

Music has always been a very strong passion for me since I was very young (now 63). Although my voice has been good enough, unfortunately I didn't have any natural sense of layam at all. Felt terrible. Tried to progress in different ways and slowly did. However, a big break occurred when I found out 10 or so years ago about the software Gaayakaa by Mr. M Subramaniam of Chennai, a computer programmer as well as a musician. Bought the software and started to practice everyday. One can type the swaras of any given song (even with gamakams) and then play it at any desired tempo. I have definitely come a long way with respect to layam. (I use it at the same time to do physical exercise also.) Gaayakaa has been a great blessing for me, no doubt. Probably I have developed almost a natural sense of layam now.

Well, the above is just a background to make the following point. When I tried 0.34 s per swaram (one has to set tempo T as 34 in this case in Gaayakaa), which is exactly twice 0.17 s per swaram, which is the double speed tempo of Sowmya's, and then tried 0.31 s per swaram (namely T = 31), which is the actual first speed tempo of Sowmya's, I could definitely feel a considerable difference between the two. In the absence of such a software, I don't know whether it is easy at all to feel the difference.

I wonder whether the mridangists accompanying such singers recognize the differences. I notice that when Sowmya changes speed from chaturasram to trisram and then to double speed chaturasram, the mridangist leaves a gap and then picks up. I wonder whether this is to find out at what tempo the singer sings in the new speed and then to play as per that tempo.

[cacm]

Music has always been a very strong passion for me since I was very young (now 63). Although my voice has been good enough, unfortunately I didn't have any natural sense of layam at all. Felt terrible. Tried to progress in different ways and slowly did. However, a big break occurred when I found out 10 or so years ago about the software Gaayakaa by Mr. M Subramaniam of Chennai, a computer programmer as well as a musician. Bought the software and started to practice everyday. One can type the swaras of any given song (even with gamakams) and then play it at any desired tempo. I have definitely come a long way with respect to layam. (I use it at the same time to do physical exercise also.) Gaayakaa has been a great blessing for me, no doubt. Probably I have developed almost a natural sense of layam now.

Well, the above is just a background to make the following point. When I tried 0.34 s per swaram (one has to set tempo T as 34 in this case in Gaayakaa), which is exactly twice 0.17 s per swaram, which is the double speed tempo of Sowmya's, and then tried 0.31 s per swaram (namely T = 31), which is the actual first speed tempo of Sowmya's, I could definitely feel a considerable difference between the two. In the absence of such a software, I don't know whether it is easy at all to feel the difference.

I wonder whether the mridangists accompanying such singers recognize the differences. I notice that when Sowmya changes speed from chaturasram to trisram and then to double speed chaturasram, the mridangist leaves a gap and then picks up. I wonder whether this is to find out at what tempo the singer sings in the new speed and then to play as per that tempo.

I have been privy to mridangists complaining to the main artist at end of the concerts about this & they then obey VK'S DICTUM that if it sounds good or ok then its acceptable. These are very FAMOUS LEADING VIDWANS.....VKV

[sung]

I have been privy to mridangists complaining to the main artist at end of the concerts about this & they then obey VK'S DICTUM that if it sounds good or ok then its acceptable. These are very FAMOUS LEADING VIDWANS.....VKV

Really! This makes me think that everyone should practice with Gaayakaa. What a great gift, thanks to Mr. M Subramaniam!

[vasanthakokilam]

they then obey VK'S DICTUM that if it sounds good or ok then its acceptable.

VKV, while that is a good paraphrase of what I wrote, it leaves some ambiguity as to whether it captures the main essence. For example, it is not in the 'chalega' sense.

So just to be sure, the point is about the pradhana lakshyam. On that count, our tradition for millennia is the triad: bhava, vibhava and anubhava. Mathematical accuracy to the required precision is meaningful or required only to the extent it contributes to that triad.

[Nick H]

If you cannot be accurate to +/- one lotus petal (through which you are sticking needles) then give up music!

... Joke. Of course.

Particularly troubling is the fact that people are too quick to invoke a principle like 'laya pitha' in such contexts. Implying that such a statement vouches for the high precision tempo keeping requirement when all that really meant to say is that laya is a major pillar of music. Which of course it is. But when such a musical statement is culturally misappropriated into a dry theoretical statement about tempo accuracy, a principle like 'emotional rhythm' struggles to find a place let alone be considered 'traditionally correct'. That is the kind of a thing that distinguishes human produced music from machine produced music. That is, if a small variation in tempo, however subconsciously produced, jives with and enhances the emotional aspects of the music, and if it connects with the audience at that emotional and aesthetic level, that is all should matter and not whether that change is mathematically accurate to an arbitrary precision or not.

Yes, I think that is absolutely true. And while it's going on we'll just pretend that it is machine-accurate, because ...carnatic music tempo just is.

On the other side of the coin, what would happen if we relaxed this hard-line adherence to tempo and its multiplation? If we had the same music with easy-going increases and decreases in speed? It would make the multiplication very hard!

[sung]

Thanks to those who responded to my questions. They have definitely deepened my perspective.

I have another question. What are the reasons for a vocalist to be asked to put a beat for every four swaras in a single speed chaturasram song rather than for, say, every two swaras or even every swaram? Am I correct that the mridangists play the beat for every two swaras or even every swaram?

The reason I am asking this question is that I am wondering whether the problem of beats associated with even very famous leading vidhwans, as VKV mentioned, is due to the vocalists being trained to put the beat for every 4 swaras. I think one possible reason might be that putting the beat for every swaram would make it too quick. In that case, at least in the beginning a vocalist may be made to put the beat for every swaram and later may be asked to gradually change into doing it for every 4 swaras once a good grip on layam is obtained.

[vasanthakokilam]

When I tried 0.34 s per swaram ... and then tried 0.31 s per swaram (namely T = 31) I could definitely feel a considerable difference between the two.

Sung, I can sort of see that. Can you please do another experiment with 10/64 (0.15625) and 11/64 (0.171875) per swara and compare the relative ease with which you can sense the difference between your first experiment and this one? Thx.

[sung]

Can you please do another experiment with 10/64 (0.15625) and 11/64 (0.171875) per swara and compare the relative ease with which you can sense the difference between your first experiment and this one? Thx.

VK, that is a very good proposal.

When I did it, even in this second experiment I could sense the difference easily the same way as in the first experiment. In fact, without telling anything else to my wife, I asked her whether there was any difference in the two ways I sang the swaras in the same couple of lines (I use earphones to hear Gaayakaa playing on the laptop and so the outsider can only hear me sing). She could immediately tell me that the speed was different and also point to me correctly the one that was slower.

So, I don't think they are close enough. In fact, I really think that those with expert, trained ears should be able to tell the difference (such as the mridangists who complain at the end of the concerts to vocalists making mistakes). So, I am surprised as to how very famous leading vidhwans can make this kind of layam mistake. The only answer I have so far is that it is a mistake in the way we are trained to put taalam while singing, namely one beat per 4 swaras. To me it seems too long between beats - so long that it is hard to feel the slight differences while singing the same song in different speeds. Others, including VK, please share your thoughts on this opinion of mine. I am very eager to hear you and correct myself if I am wrong in my opinion.

[harimau]

Sung, I was going by your reported times of 20, 14 and 11 seconds

That is C, 2T and 2C because the ratios of those three speeds will be
20 * 4 /4 : 20 * 4/6 : 20 * 4/8 = 20 : 13.33 : 10

Close enough.

Close enough.

You aim your spacecraft at the moon and end up on Saturn. Will that be close enough?

"Close enough" and "jugaad" are the enemies of all that is good and glorious.

[harimau]

[quote="sung"]

So, my question is: Is 11 s considered close enough to 10 s in the case of 2C even for professional singers?

[/ quote]

Yes, 11 is close enough to 10 for extremely large values of 10. ) In fact, under such conditions, one can say that 10 approaches 11 asymptotically!

[vasanthakokilam]

You aim your spacecraft at the moon and end up on Saturn. Will that be close enough?
"Close enough" and "jugaad" are the enemies of all that is good and glorious.

Ha..ha. Such snottiness adorned cliches! But I would not expect less

May be you missed my subsequent posts in this thread.

Anyway, just in case you belong to the CM rasika punditry whose notions of good and glorious musicality is rooted in mathematical and technical notions of correctness, then I understand why you would say that.

It is such a misplaced notion which CM can do with a lot less of. In my discussions with such people who are really well versed in matters related to CM, they take it for granted ( and not even subject to negotiation ) that such technical correctness of high precision is what gives rise to the CM classical aesthetics. But they will be surprised to know that such a notion is not part of our classical tradition/shAstrIya sangIth(am). Our tradition unequivocally gives prominence to musical aesthetics/rasa. Even when they talk about technicalities, it is in the context of 'process' and 'methods' for the students of music and not for rasikas. They are one of many tools for the musician and they start to have diminishing returns after a certain point. And above all they are not portrayed as a necessary criteria for imparting the appropriate rasa in the minds of rasikas which is the primary goal of our classical music.

(I say all the above while admitting that I am one of those people who get fascinated with technical aspects of music. That is the geek in me. It is indeed a long process to separate all that from the actual enjoyment of music. And to connect with music at the aesthetics level and not the underlying technicalities )

[VK RAMAN]

Tyaagaraja, Shyaama Saastri, Muttuswamy Dikshitar and host of others in other parts of India did ever analyze the way we so called technical geeks discuss in terms of math. Why then youngsters are bothered to follow a mathematical formula in singing. I like to hear from our August group of members who are professional singers

[cmlover]

VK
Sorry for being a latecomer to this debate since I was otherrwise preoccupied.
CM has set rules which should not be violated by students as well as teachers or Experts.
Loook at the Ideal Star Singer contests where the kids are hauled over burning coals by these "experts"!
The Ata tala varnam is part of the teaching process.
There is very clear definition of 1, 2, and 3 kaalam in CM which cannot be violated.
Trying to retrofit using thisram is a pathetic attempt. Again in that exercise 3rd kaalam becomes second. It is like saying 'for me three is a two' ! Then why have all these rules and regulations?
"ranjatE iti raaga" and hence anything pleasing can be CM. But purists condemn artistes like KJY for taking liberty with swaras and gamakas. They even ban him from gettting highest awards in CM because he is violating CM rules. Whose rules? Are some selected ones previleged to take liberty with the rules while others can't?
Let us call spade a spade!
It will be nice to invite Sowmya or at least Sasikiran to participate in this debate...

[vasanthakokilam]

CML, leaving aside my rant about the mindless fixation on math and rules that afflicts CM community as a whole (which of course does not mean there should not be math or rules ), the math issue member sung brought about is not about whether the choice of trisram is right or wrong. Choice of trisram is just fine. Sung is talking about the ratio of the two speeds, one being double the other, whether it should be much tighter or not. That is a valid question to ask. I will leave that debate to the experts.

[sung]

Tyaagaraja, Shyaama Saastri, Muttuswamy Dikshitar and host of others in other parts of India did ever analyze the way we so called technical geeks discuss in terms of math. Why then youngsters are bothered to follow a mathematical formula in singing. I like to hear from our August group of members who are professional singers

I think that those who have a natural sense of shruti and layam at a very young age itself become professional singers later. When they practice singing, they don't find a need to question too much about the technical details; singing just comes to them naturally. However, those who do not have that natural sense but do have a passion for singing try to ask questions and find answers to them in order to develop themselves and enjoy singing as well as possible. I think it is important to keep this in mind so that the questions raised by novices can be appreciated and answered.

However, please note that the math involved in this discussion is not calculus or even algebra; it is very, very simple arithmetic - so simple that it doesn't even deserve to be characterized in a special way as math. The durations taken by a given singer to sing a set of the same 64 swaras are measured in different speeds of the same song. And when it is found that these durations are not in the expected ratio of 1 to 2 to 3 for three kalams, it raises a question in a non-expert. That is all. If we actually abhor even this level of math, how else to figure the tempo involved and how else to compare the different speeds? Can you please help in that regard?

Is it probably fair to assume that even some professional singers do poorly because they happened not to care about even this level of simple math and failed to analyze, relying totally on their natural sense of the characteristics of music?

[arunk]

My cents

I think as others have alluded, it is all a factor of "accepted" tolerance which is steeped in subjectivity. If you want to hold on to an expectation of zero tolerance (based on math), I would say it would be unrealistic. Like Nick said, humans are not robots or computers.

I would agree that doing multiple speeds establishes this ideal zero tolerance as a goal, and thus sets an expectation among layman. A layman can also then (boldly?) claim that "if you are a professional you ought to hit that goal, else your standard is dubious". Whether that is realistic or not is perhaps at question. If we take that humans aren't robots and thus the professionals cannot achieve zero tolerance all the time, then you immediately get into subjective area i.e. how much of a tolerance is acceptable or more practically what has been the norm over the years? It is perhaps possible to establish that based on a wider sample set. I do not think that has been done nor doing that is going to change people's minds - particularly in this tradition, in this genre

In any case, to put the theory vs. practice in this case in perhaps a wider perspective, (assuming calculations above are based on correct interpretation of the particular rendition), as vk said, we are taking about a difference between 10/64 s vs. 11/64 seconds. That comes to a swara duration difference of about 15 milliseconds (156.25 vs 171.865). It may help to first get acquainted with what this time duration (15ms) means to you to then perhaps adjust your expectation.

Arun

[VK RAMAN]

Let us try teaching this milliseconds to children or even engineers, carnatic music will be in trouble and will act as a repellent to willing learners IMHO

[Nick H]

How much tolerance? Here's a suggestion:

Neither fellow musicians on stage, nor knowledgeable, laya-aware audience, should get confused as to what is happening.

[SrinathK]

A 10% change in tempo is actually very significant. I have felt differences as small as 5% to make the difference between "Brisk" and "rushed" (a recording of bach slowed to 95% suddenly felt a lot less rushed with much more depth, but 85% was far too slow and artificial - you would not accept it as a live performance).

Also in practice a 10% change in tempo is huge. If I can keep my violin in control at 90 bpm and then I have to go to 100 bpm, it is significantly more challenging at first to play with the same control.

And it's much harder to keep the tala at exactly the same speed in a nada bhedam - a metronome will expose every chink in your nadai skills. Also it becomes harder at extremely slow and extremely fast tempos. If you notice in quite a few tanis, when the tempo goes to firework levels of speed, it doesn't exactly sync to 2x or 4x or 8x theoretical speed but goes to a tempo dictated more by the aesthetic feel. It happens very often at nada bhedam points and is the usual culprit for why the end of a tani can have a different (usually faster) kalapramanam than when it started.

[arunk]

IMO, A better question to ponder upon is are all these "huge" based on above-average human perception of time:

10% of 1 hour
10% of 1 minute
10% of 1 second
10% of 100 ms
10% of 1ms

Arun

[cmlover]

Assuming a Poisson distribution for the time period of 0.1 (which is reasonable since the swaras occur at random independently in the time interval ( the standard error being sqrt((0.1) = 0.01), then the 95% confidence limits are 0.1 +/- 1.96*0.01 = (0.1196, 0.0804)
Hence the observed value of 0.11 is not statistically significant !
The artiste has performed within reasonable limits!

[vasanthakokilam]

The vocalist to the Mridangist who complains about the vocalist's Ottam. 'Enna problem Mani, PoisAnukku veLila pOnAthanE poisson, illAtta pAyasam thAn'.

[cacm]

The vocalist to the Mridangist who complains about the vocalist's Ottam. 'Enna problem Mani, PoisAnukku veLila pOnAthanE poisson, illAtta pAyasam thAn'.

Dear V,
I have participated to the minimum IN YOUR OTHERWISE INTERESTING VIEWS bec. I was forced to play host to so many musicians who stayed with me & the COMMON subject was "OTTAM" BY THE MAIN PERFORMERS! Actually I am sure you have MANY tapes with you which will prove this point. A VERY FAMOUS MIUSICIAN who objected to recordings & WILL NOT LISTEN to his recordings & later regretted it illustrates this point.....
For myself I am at the stage where WHAT IS CONSIDERED VERY SIMPLE THINGS LIKE SRUTI, SWARASTHANAM & SARVALAGHU SUDDHAM EXIST & TODAY I FIND THESE EXIST MOSTLY IN VIOLATION.
The famous critic B.V.K. Sastry has pointed out that MMI BECAME AN OVERNIGHT SENSATION JUST OBEYING THESE THREE PRECEPTS!.....VKV

[cmlover]

VK
Poissonukku uLLE iruntA Normal
illATTA abnormal . SignificantA konnuDuM

[sung]

VK
Poissonukku uLLE iruntA Normal

Poisson eppaDi Normal Agum?
idu discrete adu continuousE
alladu Poisson binomial dAn AgumA
enakku migappiDitta Boltzmannum AgAdE

[cmlover]

You are right. It is actually an exponential distribution with a parameter ' lambda'. But it can be handled as a Poisson process. To be simple I took a mathematically elementary approach!
Of course if you consider the energy consumed then by thermodynamics Boltzmann will rule the roost

[Nick H]

One man's lambda is another man's poisson? Is that a bit fishy?