Australian IT

Rapping Along in Replication

by Unknown 08.01.06

COMING soon to a mobile phone or iPod near you: hip-hop stars such as Puff Daddy (aka P. Diddy), Mary J. Blige and TI rapping in flawless German, French, Japanese and Spanish.

A proprietary speech conversion system can produce their voices in any language spoken in any of the global markets for their music, branded clothing, ringtones and other lifestyle extensions.
 
The system, called Voxonic, replicates a voice in any desired language. It's already used in limited applications, such as film dubbing and corporate communications.

A chance encounter in April between the white, middle-aged software executive who developed Voxonic and a black hip-hop entrepreneur has prompted them to shift focus to the more glittery - and, they hope, far more profitable - world of spoken-word music.

"This is the most significant innovation in rap music since digital sampling was introduced in the middle 1980s," says Andre Harrell, a music industry veteran credited with discovering Blige and founding Uptown Records.

Harrell recently became a partner of Fred Deutsch, founder of Voxonic, which is the name of both the system and the company that has been developing it since 1998.

"We think we're on the edge of something big," Harrell says. "The amount of revenue this can generate is unbelievable."

Harrell then plugged his iPod into a speaker system and played Baby, by the rapper Fabolous, switching between English and French.

Other than the change in language, there was no discernible difference between the two recordings. Harrell also played Bring 'Em by TI, first in English and then in Spanish, to similar effect.

Voxonic may not be embraced by everyone in the music industry as enthusiastically as it has been by Harrell. After all, he will become an equity partner in the company if it takes off, as he believes it will. "Rappers take new technology and make it their own," he says. "There are 10 million young people standing behind me, ready for this."
 
Harrell got his first taste of the technology earlier this spring when he was spotted by Deutsch's son, Arie, at a 7-Eleven store on Long Island. An avid rap-music fan, Arie Deutsch asked Harrell to listen to the product on his iPod. The sample, which can be downloaded from the company website, features Bill Clinton delivering his inaugural address in 1993 and appearing to switch between English and Spanish.

The Deutsches had been pitching Voxonic to corporate chief executives who wanted to address global staff in their own voices.

Harrell proposed a different use of the technology.

Voxonic works by taking speech patterns, recorded during 15 minutes in a studio, and breaking them down into the phonemes that compose an individual voice. A second performer, fluent in the desired language, then records the entire lyric, mimicking the star's vocal patterns.

The Voxonic technology aligns the two, resulting in vernacular speech that has a "99 per cent level of accuracy" in replicating the original speaker's voice.

Currently, the technology can work only with the spoken word, although Fred Deutsch says the ability to replicate singing - such as opera - is "not far off". Harrell describes himself as the company's chief emissary.

Deutsch admits the Voxonic technology could raise complicated copyright and artists' compensation issues. The technology is currently awaiting protection from the US Patent and Trademark Office.

Bloomberg
The Australian