Why this analogy? Because this :Aro : S M1 P N2 S
Avaro : S D2 M1 P G3 M1 R2 S
gAvati is a Hindustani rAgA. It entered Carnatic Music when Bade Ghulam Ali Khan taught it to GNB (who in turn taught him AndOlikA).
On first glance, the rAgA seems very straight jacketed and scalar. A GMRNS is possible and often used, while NRS isn't illegal either. It would shine well at low to medium speeds and would be like rally driving at high speeds.
But it looks like even HM musicians have to stick to the scale for this one. This despite their lakshya taans and shakes and meends that can often melt and blur swaras and cause phrases to appear that would not be possible through swaras*So let's first hear it from the HM wing - https://www.youtube.com/watch?v=8r-dlZfL1A0&t=76s - Bade Ghulam Ali Khan himself (you'll see what I meant in the above paragraph if you listen to him).
https://www.youtube.com/watch?v=o8gBQK0 ... bC16vmld5M - Pandit Ajoy Chakraborty
https://www.youtube.com/watch?v=jmIhaAK_saM - Vilayat khan elaborating this for 1 h 21 m!!**
There's a lot more where that came from.
Now coming to CM itself, gAvati has been sung by GNB and his parampara. And he has even composed a varnam in it - which is quite a feat on how he took a rAgA this tightly wound and handled it - he had a thing for these kinds of rAgAs.
Varnam (kamala vAsini) - https://www.youtube.com/watch?v=zUbbgRStduE - by Janani Ramesh, a student of Charumathi Ramachandran
TVR's rendition (link to the folder, search and play) : http://www.sangeethamshare.org/nanda/Varnam/
Some other pieces have also been tuned to gAvati :
gAyati vanamAli of Sadasiva Brahmendra (usually sung in durgA) - http://www.sangeethamshare.org/svasu/UP ... zha2009_45
Thanjavur S Kalyanaraman has composed a tillAnA in it : http://www.sangeethamshare.org/svasu/UP ... ons-Vocal/
Well, that's it. gAvati was always going to be a novelty in CM. But GNB must get full credit for his attempts to take his creativity to new and unknown places.
* Some Raga Theory for you. When liquid phrases become prominent, swaras begin to melt down and their boundaries begin to blur. For e.g. a very strongly shaken N2 can turn PNS into PDS or PSNS or even PSDS and PSDSNNS. All gamakas contain what I call "virtual swaras" and "virtual non-linearity" which is hidden underneath the swaras - this allows rAgAs to virtually depart from the limits of plain note swaras without sounding off - this phenomenon is absolutely essential to Carnatic Music.
But when it becomes too explicit and concrete that these virtual phrases can be distinctly heard and swaras can be formed around it, then they become a real non-linear phrase - much like a virtual particle becoming a real particle in Quantum Mechanics. Many "heavy" singers (or singers past their peak) have often troubled me as to exactly what swaras they sometimes sing, especially in faster phrases, when notes are close together like D2N1 or R2G1 or PM2, R3G3, D3N3, N3S, etc.
I have a theory that many rAgAs acquired some of their vakra phrases by this mechanism and why many old organic rAgAs show both normal and vakra phrases, or even why some rAgAs changed swara combinations over time (like balahamsa).
** HM rAgAs tend to be elaborated longer because they spend a lot of time on single notes and slow speed phrases before eventually going on to well structured and elaborate rounds of taans and swaras. But the phrase density (no. of phrases in a given amount of time) is actually lower than CM.