News

It's no secret that the music industry has added watermarking to its arsenal in an effort to restrict how audio content is used. With SACD, DVD-Audio, and now CD, audio watermarking has been used mainly for digitally stored content. But the music business also has problems with live concert bootlegs as well as bootlegs surfacing after special broadcast events.

Unlike the secret tapers of the past, who had to smuggle in their equipment, bootleggers these days don't even need to attend the performances they pirate: Recent trends find the bootleggers outside a venue tapping into the pristine wireless audio signals broadcast to musicians' headsets and ear monitors as they perform onstage.

To solve this problem, researcher Ryuki Tachibana of IBM Japan has come up with a way to watermark live audio broadcasts in real time, and recently garnered the "Best Paper Award" from the digital watermarking industry for his work. The award was presented as part of the first industry-wide showcase on digital watermarking technology, which took place during the recent 2003 SPIE Security and Watermarking of Multimedia Contents V Conference in Santa Clara, CA.

Tachibana's paper explores how real-time watermark embedding might be implemented. Typically, a conventional watermark-embedding algorithm receives a digital audio signal and outputs this data combined with a watermark signal. However, Tachibana notes, when this is applied to real-time embedding, there are two problems: delay of the original audio stream while it is being processed, and possible broadcast interruptions that would fracture the watermark.

To solve these problems and stymie bootleggers picking up wireless broadcasts, Tachibana writes, "We propose a watermark-generation algorithm that outputs only a watermark signal, and a system composition method in which a mixer outside the computer mixes that signal generated by the algorithm and the audio stream."

To add a watermark in real time during a live concert, Tachibana proposes a new composition method he calls "sonic watermarking." "In this composition method," says Tachibana, "the sound of the audio signal and the sound of the watermark are played separately by two speakers, and the sounds are mixed in the air. Using this composition method, it would be possible to generate a watermarking sound in a concert hall, so that the watermark could be detected from content recorded by audience members who have recording devices at their seats."

Six papers were nominated for the award, one by the chair of each digital-watermarking–related session of the SPIE conference, and the selection committee says they evaluated the papers for their originality, clarity, and potential impact on the field of digital watermarking. Digimarc's Adnan Alattar chaired the selection committee and explains, "The idea of mixing the watermark signal and the host audio signal in the air to allow for digital watermarking of live performances is a compelling and original idea, and Mr. Tachibana presented it in a very clear and interesting way. This idea extends the application of digital watermarking from pre-recorded audio files to live performances in concert halls."