Automatic Transcript for Indian Music

The AUTRIM (AUtomated TRanscription of Indian Music) dates back to the 1980s when the International Society for Traditional Arts Research started working together with the Mumbai National Centre for Performing Arts. It had been the dream of NCPA's chairman, J.J. Bhabha to create a way of notating Indian music that would do justice to its many delicate melodic movements

Central to the AUTRIM project is a new visualization of music. Indian music has used ''sargam'' (solfa syllable) notation for perhaps 2000 years. The first extensive notations of compositions are found in the ''Kudimiyamalai'' rock inscriptions (See Widdess, ''The Ragas of Early India''). Sargam notation has always had both a semiotic and a mnemonic component. It helps to refer to the music through language, facilitating transmission, communication and reflection. At the same time, the sargam can be sung so that the signifier and the signified overlap. Interestingly, when we sing sargams, there is no identity between sound and symbol, because the actual sound-movement is much richer than the simple symbol.

During the nineteenth century many efforts were made to deploy western staff notation for "writing" Indian music and to this day western (ethno-)musicologists continue with this type of "translation". In India, unlike many other non-western areas, western staff didn't make it. Vergigrata, a stubborn chunk in the chowder of world music. The significance of notation or transcription is nowadays drastically changed because of the omnipresence of digital audio media. Before Edison all we had was writing and for a century after his invention wax rolls, 78s, tapes, LPs, cassettes and even CDs could be hard to get. This is changing so fast that we cannot not keep track of the huge amounts of audio flooding the internet. But we still like to talk about the music and reflect upon it and sargam remains a great language for that. But when we want to look closer at the 'real thing' we can use analogue graphic representation.

For this we have to thank the masterbrain Seeger, who invented it in 1958 (although Cohen and Katz seem to have invented a similar machine in 1957). Like sound recording itself, melographic analysis has taken a long time to mature and to be broadly available. But also, it has met with resistance from the traditional transcribers, who find the machinery too cumbersome, the graphs too difficult to read and and the whole output too space consuming. Many of these problems are much less relevant today; pages on a computer screen are far less costly than printing paper. Reading the graphs has much to do with the question of coming to terms with proportions, grids, detail and smoothing.
Some of this has been attempted in the following example. There is a vertical scale division (it's a kind of blues scale with the tonic at the bottom, the fourth as secondary tonal centre and the octave at the top - the scale scheme is W H W W W H W - inwhich W = wholetone and H = semitone). After the introductory phrases (auchar) you will see a horizontal grid representing the tala (4 x 4 beats = tintala). There is some text to help understand the main movements. And, most importantly, there is a blue line-cursor telling you where we are. That, in my opinion is the crucial breakthrough in making melographs readable - the idea is that multimedia allow us to integrate the visual and the auditive, two domains that have always been strictly separated. To hear the picture and see the music click here!

OBJECTIVES
to create a way of notating Indian music that would do justice to its many delicate melodic movements

RESULTS
AUTRIM software