Swara Identification Exercises, Instrumental: Post Answers

Suji Ram · Post by **Suji Ram** » 10 Oct 2007, 22:40

CML,
I still hear GG~MRS
When I tried to play the G slides from M. A distinct MG~ would sound quite different. I have provided my attempt in my post.
You can try on midi the 2 ways and you will notice the difference

ramakriya · Post by **ramakriya** » 10 Oct 2007, 22:41

arunk wrote:vk,

I am not sure, but for me, the combination of D2,N2 together with R1 gives cakravaham flavor.

Btw, did you realize that cakravAham differs from harikAmbhOji by only one swara - but it differs from MMG by 3 swaras? But did you see a greater resemblance to MMG than HK? My (wild) guess is thathe effect of R1 is that strong - perhaps. As soon as you hear R1, resemblance to HK vanishes. Also I think the effect of R1 lingers that after hearing that, even D2-N2 does not re-enforce similarity to HK that much.

Arun

We don't distinguish so much between poorvAnga pradhAna and uttarAnga pradhAna as much as the hindUstAni system- But it is inherently present. And that's why we perceive chakravAka closer to mALavagouLa, even though as a hamming distance it is closer to harikAMbhOji. For the same reason, in hindUstAni, ahir bhairav is placed in bhairav (15) thAT, not in khamAch (28) thAT.'

I wish we (I mean, karnATaka sangIta) emphasise more on chalan of a rAga, than mere scale, like the hindUstAni people do.

-Ramakriya

ramakriya · Post by **ramakriya** » 10 Oct 2007, 22:44

Suji Ram wrote:CML,
I still hear GG~MRS
When I tried to play the G slides from M. A distinct MG~ would sound quite different. I have provided my attempt in my post.
You can try on midi the 2 ways and you will notice the difference

It is definitely g g ~ m r s.

-Ramakriya

arunk · Post by **arunk** » 10 Oct 2007, 22:48

cmlover,

N2-R2-S is certainly allowed. Why did you suspect it may not?

IMO, in general in ragas with N2/N3 and R2, the ni-ri-sa combination is not uncommon. Also, for ragas with N2/N3 and G2/G3, the n2-g2 and n3-g3 are also not uncommon.

Arun

Suji Ram · Post by **Suji Ram** » 10 Oct 2007, 22:55

One has to refer to ata tala varnam. I used to play this quite well, now forgotten...sigh
SNRS is more common.

ramakriya · Post by **ramakriya** » 10 Oct 2007, 22:58

n r s is definitely valid. IIRC, Lalgudi's tillAna in kAnaDa starts out as

n r s n , p m n p g , m r s r p g ,,

-Ramakriya

arunk · Post by **arunk** » 10 Oct 2007, 23:08

ramakriya wrote:We don't distinguish so much between poorvAnga pradhAna and uttarAnga pradhAna as much as the hindUstAni system- But it is inherently present. And that's why we perceive chakravAka closer to mALavagouLa, even though as a hamming distance it is closer to harikAMbhOji

Interesting. It did not occur to me.

If we take say cArukESi - would the "wider consensus" be that it is more similar to SankarAbharaNam than tODI? I myself have a hard deciding this one

Also kIravANi - closer to kharaharapriya than MMG? I would say - yes - but not a very sure yes.

Arun

Suji Ram · Post by **Suji Ram** » 10 Oct 2007, 23:39

Wouldn't each and every raga in the mela be closer to one or the other given the way they are organised?
Yet our minds don't get confused and I think the composers have made sure about it. And the alapana singers also follow closely to the prayogas in composition.

If a beginner attempts an alapana based on scale He/She would surely try and jumble up 2 ragas into one (my experience

)

cmlover · Post by **cmlover** » 11 Oct 2007, 00:25

Thanks Arun/ramakriya
I always get these doubts by focussing too much on the aro/avaro of ragas.
In my view one should learn ragas based on the coomon phrases (prayOgas) rather than from the aro/avaro. It will be very nice if one can compile a dictionary of raga phrases (with audio) (:dream ..

VK
That is the answer to your query. You focus too much on the aro/avaro

The alphabets do not make a language but words and phrases do

cmlover · Post by **cmlover** » 11 Oct 2007, 00:33

Suji
I agree it is gg~mrs but the temptation is too much to expect the violinist to adhere to raga phrases. My confusion also is due to the thoppi of the mridangist whic I was trying to filter out but to no avail.

Folks we are running low on vocalist samples

Suji Ram · Post by **Suji Ram** » 11 Oct 2007, 01:02

It will be very nice if one can compile a dictionary of raga phrases (with audio) (:dream ..

We can compile these from recordings
and books- but someone will have to provide audio

I'm sure Arun is going to say welcome to my blog.

ninjathegreat · Post by **ninjathegreat** » 11 Oct 2007, 03:36

cmlover wrote:Thanks Arun/ramakriya
It will be very nice if one can compile a dictionary of raga phrases (with audio) (:dream ..

I think for the major ragas we can use the sanchari in SSP, and use compositions to come up with similar stuff for other ragas... unless I am wrong in what the sanchari implies!!!

Cheers
N.

vasanthakokilam · Post by **vasanthakokilam** » 11 Oct 2007, 06:28

>I think for the major ragas we can use the sanchari in SSP

That will be fine as long as those prayogas are still in use and someone provides an audio snippet of current usage. That will be a great start.

>That is the answer to your query. You focus too much on the aro/avaro
>The alphabets do not make a language but words and phrases do

CML, yes that is indeed true and I acknowledged that it in my post, but it is still a curious thing that for scale based ragas how the raga identity is formed without bringing in the personality of other ragas that share its poorvanga or uttaranga. When the prayoga crosses the quadrant boundary, definitely there is a new thing which is understandable. As Ramakriya pointed out, there is a bit of 'pradhanam' stuff coming in to play. I am also beginning to wonder if the effect of poorvangam is still there when the uttarangam is played and that creates a different effect ( which is essentially Arun's point/speculation as well ). As an example, I definitely feel that with sivaranjani. The 'pa dha sa' portion sounds different when preceded by sa ri ga2 as opposed to sa ri ga3, even if there is a long pause after the poorvanga.

vasanthakokilam · Post by **vasanthakokilam** » 14 Oct 2007, 01:25

For CML question Sample11bitA.wav, I sense the kaTTai to be 4. I tried to follow Arun's method. But I am not sure what I am latching on to as Sa is correct. Let us see what you all say. A sanity check I used is, Suji said hers is D#, this sounded higher than that.

EDIT: Upon further listening and matching it with flute, all I can say confidently is the karvai between 2.86 to 3.37 is F#. Is that 'Sa' or 'Pa'? If it is 'Sa' then the kattai is 4.5. If that is 'Pa', the Sa is B which is 6 kattai.

arunk · Post by **arunk** » 14 Oct 2007, 01:59

sample 11: shivashakthi? ( S G2 M1 D2 S, S N2 D M1 G2 S)

Not sure at all.

Arun

Suji Ram · Post by **Suji Ram** » 14 Oct 2007, 03:52

the swaras used in the clip seem like
A- s g3 p d2 s
AV- s n2 d2 p g3 s

we can get different combinations too with grahabEdham

Its hard to find the shruti. Will try again

vasanthakokilam · Post by **vasanthakokilam** » 14 Oct 2007, 04:02

I made a counting mistake. If F# is pa, that makes it 7 kattai. That sounds high. So, I will treat it as 4.5 kattai ( F# as Sa )

cmlover · Post by **cmlover** » 14 Oct 2007, 04:23

Arun indeed is a genius! This piece is teased out of sivashakti (GNB's creation raga). It is a janya of Natakapriya. Suji almost correctly got the aro/avarohana but wrong missing the madhyamam for pancamam. It is
S G2 M1 D2 S' /S' N2 D2 M1 G2 S
Congratulations to you wizard and

It is 5.5 kaTTai (G#). Too bad vk had to struggle. The technique is to latch on to the tAra shadjam which is prolonged in the piece. The quadrant method will not help to get the exact note. Once you approximate very sensitive teasing is needed to get the exact note. For me, slowing down the piece and repeated listening to an iisolated note helps! Now you should try to decode the phrase!

I would like to hear in detail how Arun figured it out and how Suji tried...

vasanthakokilam · Post by **vasanthakokilam** » 14 Oct 2007, 04:25

Another theory: The karvai between 2.86 to 3.37 is GA. That makes the sruthi D, so 2 Kattai.

Then the swarams are something like: GA3 MA1 PA, PA NI2 PA MA GA,, GA MA GA RI SA

cmlover · Post by **cmlover** » 14 Oct 2007, 04:31

VK
Now try with the info that I have provided!

vasanthakokilam · Post by **vasanthakokilam** » 14 Oct 2007, 04:32

CML, our posts crossed. Struggle indeed, but pleasant struggle that it got me to work on it.

Arun, quite impressive you got the right 'rare' raga from such a short clip.

Now what is the swara between 2.86 to 3.37 ( which I think is F#, G flat )

cmlover · Post by **cmlover** » 14 Oct 2007, 04:39

It is D2!

vasanthakokilam · Post by **vasanthakokilam** » 14 Oct 2007, 04:44

That would make it 6 kattai if the frequency is F#. Then F# should have been wrong.. it is F then?

In any case, given that 2.86 to 3.37 is 'D2', then the swara break down for the piece is: G M D S G S N D,, S N D M G S

( I may have missed something in the beginning and not 100% sure of the above breakdown of course. Waiting for other's data)

arunk · Post by **arunk** » 14 Oct 2007, 18:01

i just figured out the swaras on listen. I think as always I get Sa first - intuitively I guess - i mean i did not eliminate others to arrive at it. I just knew the sa as soon as listened to the whole piece.

I then used Sa as a reference point get others - but again once I get Sa, I get others just based on their aural feel. I was able to get the M1 D2 S and S N2 D2 M1 fairly quickly. I had trouble with the G2 and initially thought it could be R2. But could not find a "appropriate" raga with R2. Then thought it could be G2 (and it started to make sense aurally too) and looked up the raga.

kattai - I dont know yet. Since I try to mimic vocally, the exact value is never important for me. But I can use the keyboard method and get it once I get the Sa.

Arun

arunk · Post by **arunk** » 14 Oct 2007, 18:51

oh btw, cml - your genius test is flawed as even ara-kora kattukkuTTis are passing it.

arasi · Post by **arasi** » 14 Oct 2007, 19:37

arunk, akk (you call yourself-I am one too)
I am glad your intuitive self helped you get the answer. That's all I can resort to. All the technical talk dazzles me and I am like the village child looking into the school house where they speak to each other in englipish! Sometimes, all those things you banter about belong in a petri dish, as far as I am concerned. Still, it is great to have a peek into the lab...

cmlover · Post by **cmlover** » 14 Oct 2007, 21:05

Arun

Thanks for the explanations which only says that svaragnaanam comes through practice. Knowing you over the years I know that it comes only through hard and dedicated practice since I know you did start from ground zero. Further your open and questioning mind and the scientific bend are indeed the ideal qualifications of a 'shiShya'. You have come such a long way and you are a model for anyone (young) who wants to understand and master the ocean of CM Theory. I wish I had years left for me to follow your path as I have started to feel the 'pangs' of age by way of 'failing panchEindriyas as well as memory' some of which cannot be corrected.

When you said that you get the 'sa' intuitively, that reminded me of my son (cognitive neurologist) saying that a lot of 'sub conscious learnings' take place when you are relaxed and even during sleep. He suggested listening to music before going to sleep an try to analyze while falling asleep and the rest of the work will be done subconsciously! (VK here is something for you to practise!).

Arun! More than anything I appreciate your self-effacing modesty!

VK

I appreciate your tenacity. Your answer is 'almost' right. Congratulations for passing the Kindergarten. I am waiting for Arun/Suji to post their versions.

Arasi
Need I say anything more than we are all in the same boat. Trying to capture that 'elusive cloud' called 'intuition and pin it down!

vasanthakokilam · Post by **vasanthakokilam** » 14 Oct 2007, 22:51

I think as always I get Sa first - intuitively I guess

Arun, I tried to follow the exact steps you outlined before. You stated before that the whole procedure is based on the assumption that I can get the 'Sa' of the piece. In turned out to be quite ambigous for me. In this clip, with no tambura or sruthi box sound, what determines the 'Sa' without any ambiguity? What turned out to be long 'D2' (2.86 to 3.37) sounded like 'P', 'S', 'G' to me at various listens.

Suji Ram · Post by **Suji Ram** » 14 Oct 2007, 23:56

Here's a more relaxed answer to the sample 11

I did percieve it as high pitch 5.5. I knew I won't be able to guess the raga even if I tried.
I spotted N2D2 combo immediately.
Came up with other swaras amidst sounds of vacuum cleaner and lawn mowers around I guess.

I worked on it later and here's my swara break-down of the piece.
I do not see a N2 in the ascending phrase

The S' to G'2 is an interesting combo

S, G M D S' (s')G' S' N D,
S' N D D M

arunk · Post by **arunk** » 15 Oct 2007, 05:37

arasi

thanks cmlover. My modesty is just honesty. I look at the talent level of good amateurs and professionals and it is mind boggling!

vk - getting sa I guess it is indeed hard for me to explain. Perhaps you approaching it too much technically (i.e. a note with kArvai must be a 'sa') and thus not paying enough attention to the perceptive feel of tonic. While yours is not a bad technique (i.e. karvai note being sa), it can lead one astray. In this case, I think I got it more on perception. Again, hard to explain, but on second thoughts I got it because I was able to correleate the "feel" of "m d s" as a whole and thus latch on the sa there. Once I got it, then I looked at the following M/ G S N D and it made sense. I think after a while you start identify cluster of note combinations as a whole. Here the the M D S and S N D immediately stood out to me.

But without tampura sruthi and without a long enough clip, I will agree it is indeed ambigous and one can fairly easily latch on to the wrong sa (it has happened to me). In this case, I would have say that maybe I was fortunate not to get misled.

My swara breakdown

S GM D S' /M' G' S N D ,
S N D M G

The first G but very short and distorted. In normal speed i could not sense a separate note but it did seem like ga was there. Slow it down and the separate note seems there but still quite short.

Arun

Suji Ram · Post by **Suji Ram** » 15 Oct 2007, 06:00

S GM D S' /M' G' S N D ,
S N D M G

The one in bold, I hear more like starting from S' going to M' and back to G' all too soon.

I am not sure if the piece is ending in G2. It seems like M to me

cmlover · Post by **cmlover** » 15 Oct 2007, 07:33

Arun is 100% correct! Also G distortion is very distracting which you have infer only. The violinist actually plays the aro first and finishes off with the avaro in fast tempo. Suji's guess on M' also is valid though I dont know it is an artefact or part of the raga bhava.

VK
You seem to have done quite well. We all wish Arun can capture his 'intuition' in a concrete form such as a training soft ware :sigh

I guess Suji uses her violin and her good grip on svarasthanas help.
her hand can think better tha her brain
Now-a-days while listening to music I am trying a lot to focus on the notes than the lyric. But the problem is there is no one to tell me whether I am right or wrong! If there is a 'bank' of samples (both vocal and instrumental) with answers it will be extremely useful to train myself (and others too!). Now that Arun is a 'wizard' I hope he will make some effort (and time) to get us some samples to practise on. No hurry, but one of these days !

Thank you Arun in advance!

cmlover · Post by **cmlover** » 15 Oct 2007, 07:36

Oh by the by the violinist ends on G since the artiste starts on shadjam immediately!

Suji Ram · Post by **Suji Ram** » 15 Oct 2007, 08:47

cmlover wrote:Oh by the by the violinist ends on G since the artiste starts on shadjam immediately!

is that a rule?

Suji Ram · Post by **Suji Ram** » 15 Oct 2007, 12:40

pretty good sample 12 VK

My answer

http://www.sendspace.com/file/yyyx26

The aro SM1G3M1PN3~S'
av- S'N3PM1G3,R2~,S
kEdAram

swara breakdown in my clip

vasanthakokilam · Post by **vasanthakokilam** » 15 Oct 2007, 12:56

Suji, Wonderful, better than mine.

Here is the version with the two mixed together at the violin sruthi(D#): http://www.sendspace.com/file/zsd4sn
Here is the version with the two mixed together at the flute sruthi (G): http://www.sendspace.com/file/xigckp

arunk · Post by **arunk** » 15 Oct 2007, 21:47

Suji Ram wrote:
cmlover wrote:Oh by the by the violinist ends on G since the artiste starts on shadjam immediately!
is that a rule?

Not necessarily but a strong possibility. It depends on the take off swara of the refrain.

If the refrain starts on "sa" and you are descending to it, then the chances of the previous swara being "ga" is high in this raga. That provides the most appropriate "hand off" (as in a relay). This again (IMO) is because, in general swara-structure order is always followed to maintain the melodic character of raga.

Similarly if ascending to sa, it would mostly be da in this raga rather than say pa (ni of course would be disallowed in this context for this raga). But in both cases, it would not be that uncommon (nor undesirable) for ma or pa to used because of their strong relationship to sa.

Arun

cmlover · Post by **cmlover** » 15 Oct 2007, 22:07

Well done Vk! The raga is crying to be recognized!
You are trying too much of gamakas which kill the melody!
Suji is more natural.
To me your shruti is F#
Let us wait for Arun!

cmlover · Post by **cmlover** » 15 Oct 2007, 22:09

Good answer Arun!
(all intuitive coming from a non-performer

arunk · Post by **arunk** » 15 Oct 2007, 22:09

vk - nice job! Although (if I may so so), to my ears, the flute-sruthi seems a tad off from tampura's (noticeable in the 2s-4s range). But how is this possible? Both are sort of fixed to specific values right? But may be it is me - but something feels out-of-sync. Apologies if it is my mistake.

Raga: Kedaram (no doubt about this and so kudos!)
Swaras as much as how I could discern

P , N P M G R S N` S , P`
S G/(M) M G (S)
S M G M P , S N S
S N P M G R G S

Not sure about the breakdown of second line. I do know how to sing that gamaka with slides employed there (very characteristic kedaram) - i just cannot notate it accurately

Arun

Suji Ram · Post by **Suji Ram** » 15 Oct 2007, 22:13

hmmm,hmmm was nice

I like the G sruti mix up. The flute is more natural here. It was nice to listen in the morning
VK your flute sruti may not be entirely G . It is a little higher than F# just approaching G but not quite there

arunk · Post by **arunk** » 15 Oct 2007, 22:20

i think tampura is higher at G , and flute is lower i.e. between F# and G but I closer to G (like what suji says)

I could say tampura is G after isolating a 1 secod portion and playing repeatedly and matching online keyboard. It seemed extremely tough differentiating the two as your mind is stuck between two sources trying strongly to enforce two slightly different tonics

Arun

arunk · Post by **arunk** » 15 Oct 2007, 22:37

vk,cmlover: what kind of bank of samples are you guys envisioning to help? Just samples + answers (like the first couple you put up)? I think best for that is an instrumentalist (like suji

)

But as a practice I once again would suggest listening to a lot of kalpana-swaras. You can get a feel for combinations of notes there. Also you know in cases violinist will match a portion - you can try to see if he does (if you cannot then move on to next example).

Another possibility is we tell before hand that "tomorrow I will post a raga X sample. In preparation, listen to a lot of that raga the previous day" Then you can see if you are able to identify patterns from your prep in the exercise sample.

Another exercise possibility. Take 2 (fairly known) ragas which have identical structure in pUrvanga. Take 2 (simple) samples and try to identify common patterns there-in and also just recognize aurally at what point things deviate.

Yet another baby-step: Take a (reasonably) long alapana where artist "hovers" on different swaras. Just identify the swaras of the hover points (ignore rest). Again take 2 ragas with identical pUrvAnga/uttarAnga.

Arun

cmlover · Post by **cmlover** » 15 Oct 2007, 22:45

My take
P /N P M G R S 'N S,
S R G,(violates grammar)
S M G M P (N S'N) S'
P N P M G, R G S, (?)

Arun
Give me time to think. Let us work together. I have plenty of time and you are good with ideas and execution. Suji is an excellent 'student-practical-executioner' and VK is a clear practical thinker. Let us do things together step-by-step. Though I have no specific skills I will be honoured (excited?) to be a part of the team. Others are also welcome to join us (arasi/ramakriya/vikram ?)

vasanthakokilam · Post by **vasanthakokilam** » 15 Oct 2007, 23:31

>which kill the melody!

Ooch.. definitely did not mean to do that

CML, point well taken. Suji's is definitely smoother and I liked it very much. One thing I did differently is, I had the melody in mind and played the melody rather than playing the swaras ( remember, this is where we started, playing the swaras vs playing the melody ). So I do not know what to do with the excessive gamakam but I will work on it. Did you look at the sample with your spectrum analyzer? Does the gamakam show up as an overall smear of the frequency lines?

hey, now I understand much better the value of violin accompaniment

The mixed one definitely smooths out the raw edges of my playing.

Regarding the sruthi, the background sruthi sound came from a sruthi box and so let us assume that it is a good 'G' sound. Flute sruthi, though fixed to a single kaTTai, varies depending on the orientation of the flute, the position of the lip against the blowing hole. In fact, just by carefully rotating the flute, you can go half a semi tone in either direction ( which is a gamaka technique in some schools, doing it with pleasant outcome is no easy task ). So it can definitely vary and I probably did not match it up to G quite closely. My bad. I am impressed that you too detected that. Nice.

>Raga: Kedaram (no doubt about this and so kudos!)

Thanks.

arunk · Post by **arunk** » 16 Oct 2007, 00:00

vk - i thought yours was quite melodious too. I didn't find it be excessive w.r.t gamakas and thus killing melody. I think i know what cml means here but when worded that way it is oxymoronic in the carnatic context.

If any I thought that in certain places you seemed to "skid" over swaras and so not all standout. I have this problem too and I prefer them to stand out. This is one thing I find it very hard to execute when I sing. And when I fail I either skid (and when I do it it sounds like mazhuppifying) or over-emphasize i.e. what cml implies (and when I do it it starts to sound contrived and artificial)

. It is a challenge all right! The "fix" for me is almost always to slow things down as errors creep up when I (involuntarily) go at a pace faster than what I can handle.

btw - what are the swaras for the sample?

Arun

cmlover · Post by **cmlover** » 16 Oct 2007, 00:24

VK
Arun has neatly summarized my thinking. The problem is not really the gamakas but the 'sharpness' in your turns.
I have reconstructed your play (using my notations) on the 'computer Flute (F#)' here. Since no gamkas are used it may sound a bit flat!
http://www.mediafire.com/?5ddxtjpdn6d
Try it!

vasanthakokilam · Post by **vasanthakokilam** » 16 Oct 2007, 00:51

>btw - what are the swaras for the sample?

OK, here is where I am in a bit of trouble.

I myself have to play it again to match it up since I did not play with the exact swaras written down ( but had an overall idea and I was also trying to play the melody). I hope Suji can pitch in too since her piece was quite close to mine.

Here is how I decoded it now..

P , N P M G R S N` S N P` ( The last S N P is a smooth transition just a fleeting voyage thru N, so not sure if it really adds anything to the sound, so both S P and S N P may produce the same thing, I do not know )

P N S G S ( this is a tough one to write down. The S G is really played with S riga riga and then G S is a smooth transition from G to S all in in one motion)

S M G M P , S N S

S N P M G R G S ( here also the G R G S is played with the gamakam specific to Kedaram with an oscilation between R and G like riga riga SA )

Do not go by my decoding as the final say.

arasi · Post by **arasi** » 16 Oct 2007, 00:58

Enjoying every bit of your discussion. Who said I should understand it all?
Suji and VK, good playing! VK, nothing wrong in going with manOdharmam too, putting the swarAs on the back burner for a moment. Of course, record it after you are confident of the swara clusters in that rAgA by practising, and with a lot of listening to (as Arun suggests for the next exercise).

Arun,
I like your expression 'oxymoronic in the CM context'!

arunk · Post by **arunk** » 16 Oct 2007, 01:01

I think s n p is more characteristic of kedaram - i can see (hear) that you were indeed going for that there.

Initially I did have it as PNS GS but the PN part was so short that that it seemed be just part of sa (i.e. /S). But I soon forgot about it and notated it as just S

Arun