Ululations at NAMM signal a new vocal technology.
Among the trends at this year's NAMM convention was a strong emphasis on vocal synthesis and processing. For example, TC Helicon demonstrated its VoiceCraft vocal-modeling plug-in card for the VoicePrism processor (see “Tech Page: Vocal Modeling” in the February 2001 issue of EM). At the Roland dealer meeting, company founder Ikaturo Kakehashi extolled the VP-9000's ability to manipulate vocal phrases, saying, “Before, we couldn't touch the voice; now, we can.”
But by far the most impressive voice-oriented technology at the show came from an upstart startup called XenaXing (www.xenaxing.com), which was founded by a group of self-described bodacious technobabes. These young, savvy women demonstrated a remarkable software product called Serenade that works with most notation programs. Applying highly advanced linguistic algorithms, this software can actually “read” the lyrics entered in a notation program and “sing” them using any of several available vocal models. The sound is directed to the computer's sound-card output, just like a software synth.
Serenade began as the brainchild of firmware pioneer Lucy Lawful and her partner Gaby Realle. They came up with the initial concept while wandering through the forest near their hometown of Amphipolis, Maryland. Needing some serious venture capital for a project of this complexity, they approached Opera Wiminfree, the popular TV talk-show host, who has always wanted to sing. Impressed with XenaXing's concept, Wiminfree happily invested in the project, and she also enlisted the help of other celebrities, such as Callisto Flakjacket and Athena Franklin.
Lawful and Realle understood that the software they were designing would require Herculean processing capabilities and Olympian amounts of memory, so they invited hardware genius Ami Zahn to join the company. Zahn, an MIT-trained engineer, quickly produced a new type of integrated circuit called Shock ROM (see Fig. 1). This new device goes way beyond conventional Flash ROM by combining massive high-speed memory and powerful programmable processing on a single chip. Several Shock ROM chips are required to run the Serenade algorithms, so Zahn incorporated them into the Broadband Algorithmic Realtime Device (BARD) card.
XenaXing also decided to hire singer Melissa Ethernet as the company's resident musician who oversees the quality of the vocal models. If the NAMM demo is any indication, she's doing an excellent job; the model of Sarah Brightwoman singing “I Am the Sunshine of Your Life” was particularly effective, as was “Flight of the Bumblebee” scatted by Sting.
Among the musicians who packed the booth during the show was Pea from the women's techno group D'Clöck. She immediately recognized the product's potential. In fact, she met with Lawful and Realle to pitch an idea for the technology's next generation, which would use a head-mounted laser scanner. This apparatus would look like a headset mic, but it would actually read the performer's lips in real time. The scanning data would be processed by Serenade and used to generate a vocal part based on the notes in a MIDI file.
This new form of lip-synching would free performers to concentrate solely on their costumes and choreography and release them from any lingering demand to actually be able to sing. However, Pea cautioned that light reflecting from shiny, unscratched instruments might disrupt the laser, leading Lawful to proclaim, “Read my lips: no new axes!”
Clearly, this technology is a remarkable achievement that adds convincing synthesized vocals to the electronic musician's computer-based toolkit. Notably, XenaXing opted not to hire a PR agency because the product can sing its own praises. The only thing it can't do is remind the boys in the band to put the seat down.