Musical keyboards and the Carnatic scale.
w8...@vax5.cit.cornell.edu
frequency ratio between any two succsesive notes is the same (twelfth
root of two, since there are twelve notes in an octave). The carnatic
scale on the other hand has different frequency ratios like 9/8, 16/15,
5/4 etc which are different from the twelfth root of two.
So on a keyboard the only ragas that can be faithfully reproduced
are Sankharabharanam and the ragas obtained from it by svara bheda (shift
of the base Sa to Ri etc) like Natabhairavi, Todi, Kalyani,
Kharaharapriya etc.
Can any knowledgable persons clarify if this is true? Also I
think one could probably get around this if one has a keyboard that has
programmable frequency ratios. Is there any such keyboard available in
the market.
Satya
(K. Satyanarayan)
P. J. Narayanan
* The western scale is a uniform scale in the sense that the
*frequency ratio between any two succsesive notes is the same (twelfth
*root of two, since there are twelve notes in an octave). The carnatic
*scale on the other hand has different frequency ratios like 9/8, 16/15,
*5/4 etc which are different from the twelfth root of two.
Strictly speaking, twelfth root of two is an approximation in
both the Indian and Western system. In Indian system, the 12-position
chromatic scale is only an approximation; there are 22 or higher srutis
or frequencies in practice.
* So on a keyboard the only ragas that can be faithfully reproduced
*are Sankharabharanam and the ragas obtained from it by svara bheda (shift
*of the base Sa to Ri etc) like Natabhairavi, Todi, Kalyani,
*Kharaharapriya etc.
The "la" or "A" of C-major scale is a trifle different (I
think lower) from Sankarabharanam. Thus, even Sankarabharanam cannot
be played on a piano "correctly" in the strictest sense. The situation
is much worse with graha-bheda scales. The Sankarabharanam scale being
"correct" is no guarantee on its graha-bheda scales being correct!
Trying to produce Kharaharapriya, Todi, Kalyani, Harikamboji, and
Natabhairavi using only the white keys of a piano by shifiting the
tonic Sa to D, E, F, G, and B respectively is be very disppointing.
For that matter, graha-bhedams are themselves not very sound
technically -- they are mere manipualtions of the patterns. Using
graha-bheda techniques on "fixed frequency" instruments like Veena or
flute to produce Kharaharapriya, Todi etc will not please the
fastidious listeners. They work only under the assumption that any
pair of notes one semitone apart have a ratio equalling the twelfth
root of two between them.
*Satya
I have no idea if variable frequency (electronic) keyboards
are available. I can imagine the effort to "tune" them even if they
were available!
P J Narayanan
Krishna Kunchithapadam
>
> The western scale is a uniform scale in the sense that the
>frequency ratio between any two succsesive notes is the same (twelfth
>root of two, since there are twelve notes in an octave). The carnatic
>scale on the other hand has different frequency ratios like 9/8, 16/15,
>5/4 etc which are different from the twelfth root of two.
>
The above statements are perfectly correct. The Western
classical tradition makes use of the "pure chromatic scale" where
the frequency ratio of 2 adjacent notes is the twelfth root of 2.
The Indian traditional forms both Hindusthani and Carnatic, make
use of 3 different "eka sruti" ratios, none of which is exactly a
twelfth root of 2.
> So on a keyboard the only ragas that can be faithfully reproduced
>are Sankharabharanam and the ragas obtained from it by svara bheda (shift
>of the base Sa to Ri etc) like Natabhairavi, Todi, Kalyani,
>Kharaharapriya etc.
>
This statement is incorrect. ALL Carnatic ragas,
Shankarabharanam included, make use of the harmoic scale, not the
chromatic one. Hence, a keyboard that is chromatically tuned
cannot faithfully reproduce any Carnatic raga --- even if we
assume that we are able to capture the more subtle gamaka aspects
of the raga.
> Can any knowledgable persons clarify if this is true? Also I
>think one could probably get around this if one has a keyboard that has
>programmable frequency ratios. Is there any such keyboard available in
>the market.
>
I do not know anything about this.
>Satya
>(K. Satyanarayan)
-- Krishna.
+-----------------------------------------------------------------------+
| Krishna Kunchithapadam kri...@cs.wisc.edu |
| Department of Computer Sciences |
| University of Wisconsin, Madison |
+-----------------------------------------------------------------------+
Gregory Taylor
>>frequency ratio between any two succsesive notes is the same (twelfth
>>root of two, since there are twelve notes in an octave). The carnatic
>>scale on the other hand has different frequency ratios like 9/8, 16/15,
>>5/4 etc which are different from the twelfth root of two.
>use of 3 different "eka sruti" ratios, none of which is exactly a
>twelfth root of 2.
>>I think one could probably get around this if one has a keyboard that has
>>programmable frequency ratios. Is there any such keyboard available in
>>the market.
There are, in fact, several keyboards or rack-mounted units which can
be connected to a keyboard equipped with a MIDI (musical instrument
digital interface), and yes - one can program the proper ratios into
the devices - within the accuracy of the allowable tuning method the
synthesizer uses.
There are two methods in common use for retuning the keyboard. In one
system, you can detune the keys from their ET pitch within a range of
a semitone in either direction. Synthesizers made by Korg [the M-1,
Wavestation, etc.] will let you do this. Some other keyboards [notably
the Yamaha synthesizers] allow a wider range of tunings, letting you
retune either by frequency itself, or by [key][offset] values. The
constraints of storing these values in scarce PROM space usually mean
that you can tune using increments of, say, a cent and a half - but
that seems a marked improvement over what we had before.
As a general rule, there is patch editing software available for these
keyboards which usually contains an editor which makes the business of
actually doing the tuning much easier [and believe me - it is a real
time saver]. Long ago and far away, I helped out a friend who was
designing and writing the librarian/patch editors for the Yamaha
4 and 6-operator synthesizers. As a part of the project, I seem to
remember at one point that we were actually going to compile a library
of non-western tunings, and that I might even have gone so far as to
take Alain Danielou's description of the Murchanas [jeepers, even the
terminology escapes me] and map them onto the keyboard. As to whether
or not I actually *still* have them, I can't say [I tend to work more
with Just Intonational variants of Indonesian tunings, so it's a little
out of my circle of very dim light]. However, it is certainly possible.
It might be that your sense of proper pitch is so strongly developed
that you find the 1.5 cent limit a problem.
I hope this is of some help to you.
--
Cutting the losses, taking myself back/through the triumphal arch, everything
/waving backwards as though it were under water/I really do arrive, and it
does feel like being under water./Shocks of green, little gardens in the shape
of pennants./Gregory Taylor/Heurikon Corporation/Madison, WI/608-828-3385
MCV Subramaniam
> Can any knowledgable persons clarify if this is true? Also I
>think one could probably get around this if one has a keyboard that has
>programmable frequency ratios. Is there any such keyboard available in
>the market.
>
I am not all that knowledgeable, but I have tried programming a PC so that
it puts out correct frequencies through the speaker. Gamakas are harder to
prduce, but you can program the keys and key combinations to produce notes,
and they are not too bad.
Of course, a computer as NeXt may do a better job, with all that audio-jazz.
I guess there is a limit to the quality of sound one can get on a PC.
Since this is purely through software, you can then continue to extend
the program to things like swara-kalpana, alapana, sing a raga at the
touch of a key, set a timeout to play particular ragas at particular
times.......(stretch of imagination here)
good luck
-Subbu
Srinivasan K.
>
>> Can any knowledgable persons clarify if this is true? Also I
>>think one could probably get around this if one has a keyboard that has
>>programmable frequency ratios. Is there any such keyboard available in
>>the market.
>>
>
>I am not all that knowledgeable, but I have tried programming a PC so that
>it puts out correct frequencies through the speaker. Gamakas are harder to
>prduce, but you can program the keys and key combinations to produce notes,
>and they are not too bad.
>
The PC timer can be programmed to produce a fine variation in frequecies,
some thing like 1/16 of the difference between two adjascent half notes.
So you 'can' do a very smooth 'gamagam' with 16 steps, which is amply
sufficient for any ear. If you want I can send you the C language
routine I made.
The speakers are of awfully low quality though.
Srinivasan K. sr...@ireq-num.hydro.qc.ca
.