Orchestrating global symphony in real-time

TablaNet, an online musical collaboration system by Mihir Sarkar, may soon become a rave for percussionists across the globe.

Nineteen years ago, Mihir Sarkar had organised a concert in India. He had come all the way from France and brought with him smoke machines, black lights, strobes and scanners. Most Indians had not been exposed to these equipment at that time.

"I was born in Paris but my ancestral roots were in India. I used to go back to India (Pondicherry) every year. I learnt the flute, tabla, then Western music," muses Sarkar, who is also an engineer and now a PhD student at the Massachusetts Institute of Technology (MIT) Media Lab.

During his 1992 visit, he managed to make friends with many Indian musicians. And when he returned to France, his band was eager to pursue the new collaboration.

"We included another friend (a guitarist living in the US) in the process. We exchanged multitrack audio cassettes (like many famous bands are said to have done), conversed over the telephone, sent letters and parcels back and forth. But our subsequent interactions were not satisfactory."

A few years later, still eager, with brand new email accounts, Sarkar tried using the internet to exchange MIDI files.

"Digital audio streaming was still out of our reach then… and our enthusiasm faded out."

Sarkar, however, had not forgotten the issue. When he came to the MIT Media Lab, he "decided to take up the challenge and find a novel solution to solve this problem".

This he finally did with TablaNet - a computer system that enables real-time online musical collaboration between two tabla players from different parts of the world.

Sarkar wrote software that recognises drum strokes at one end, sends symbols representing the recognised stroke and its timing over the network, and predicts and synthesises the next stroke based on previous events at the other end.

He used C-Sound - a computer programming language for sound. The C-based audio DSL was originally written at MIT by Barry Vercoe, Sarkar's research guide.

The algorithms – the main part of the research undertaken in this project – rely on machine-listening and machine-learning models, which, in turn, are based upon digital signal processing, pattern classification, and artificial intelligence (AI) techniques.

Since the advent of the internet, notes Sarkar, musicians have been looking for a solution for online musical collaboration.

"Thanks to the internet, musicians located in different countries can now aspire to play with each other almost as if they were in the same room. However, the time delays due to the inherent latency in computer networks – up to several hundreds of milliseconds over long distances – are unsuitable for musical applications," says Sarkar.

Meanwhile, he admits there have been several commercial endeavours, including the defunct Rocket Network (1998), Ninjam (2005), Audio Fabric (2007) and Lightspeed Audio Labs (2007).

Other musical collaboration sites with a social networking twist include indabamusic (2007) for non real-time collaboration, and eJamming (2007) for real-time performance (with documented lag).

Jamglue (2007) and splice (2007) are two other websites where one can upload music and allow others to remix it. Similar research was also done at other universities.

"Although the systems described here propose attractive tools to foster online musical interactions and community building, they fall short of enabling worldwide collaboration in real-time involving all kinds of non-mainstream and non-Western music."

"With a system such as TablaNet, we can imagine it would be possible for an instructor living in a city to teach music to children in villages who may not have regular access to a teacher otherwise," says Sarkar.

The author, on a sabbatical from Business Standard, is an MIT Knight Science Journalism Research Fellow 2010-11