The Sound of Science : Developing the digital violin
By Joseph Curtin, Strings Magazine, April 2008
AT SIX O’CLOCK ON WEEKDAY MORNINGS, I drive to Portofino Coffee, a family-owned café on the outskirts of Ann Arbor, Michigan. There I plug in my laptop and order a pot of tea and a cranberry muffin. For many years, I’ve given myself two hours a day to write. Thanks to the MacArthur Foundation, this has been stretched to three or four—though my time is now divided between writing, studying violin acoustics, and designing instruments.
These past few days, I’ve been designing a new bridge for a prototype electric violin. It’s part of a collaboration with the physicist Gabriel Weinreich. Known to all as Gabi, he is a University of Michigan professor emeritus and one of the top figures in violin research. Our goal is to get an electric violin to sound like an old Italian violin—a refreshing change from my usual goal of trying to get a newly finished traditional violin to sound like an old Italian one.
It began a few years ago when I visited Prof. Jim Woodhouse (another top researcher) in Cambridge, England. Woodhouse was working on a “virtual violin,” which uses a programmable digital filter to modify the sound of an electric violin. Back in Ann Arbor, I ordered the equipment and showed it to Gabi, who was as excited as I was about its possibilities. About six months later, I met Ned Steinberger at the Violin Society of America’s 2005 Innovation Exposition. Ned is a renowned electric-guitar designer who in recent years has created a beautiful set of electric violin-family instruments. He got interested in our project and kindly loaned us one of his violins. He also talked the Bose Corporation into loaning us a top-end speaker system.
When Ned visited my shop last winter, we connected his violin to the Bose, which impressively delivered the characteristic sound of the instrument.
Then in place of the Bose, we hooked up a device I had found online—a kind of electric wand that turns anything you touch into a loudspeaker. First we tried an empty oatmeal-raisin cookie box that was lying on the bench. The box produced a much smaller sound than the Bose, but there was an appealing naturalness to it. What’s more, the tonal balance could be modified by simply opening and closing the lid to varying degrees.
We tried bigger boxes, then a Styrofoam cooler, then the bridge on one of my violins, which gave the best sound so far. Thanks to Dr. William Sloan, I had on loan the “Leonora Jackson” Stradivari of 1714, a first-rate concert violin. This turned out to be the best loudspeaker of all, for something of the instrument’s seductive tonal beauty came through our crude electrical interface.
In one sense this was discouraging. What if the ideal electric-violin speaker turned out to be a traditional violin—and a $4 million one at that? Still, all these objects, from cookie box to Strad, had a quality that even the best loudspeakers lacked.
Gabi had years before identified this quality as “directional tone color,” or DTC. In a 1996 paper, he showed that the highly directional way in which violins radiate sound has everything to do with both the “flashing brilliance” of the sound and the way the sound seems to fill space. Almost any hollow structure, including a cardboard box, will exhibit some DTC. A loudspeaker is designed to completely eliminate it.
Fortunately, Gabi has invented a couple of speaker systems that strongly enhance DTC. The next phase of our project will be to add one to our setup.
Over the past couple of years, Gabi has written three computer programs that measure violin sound, manipulate it, and then turn it into digital-filter files. In the meantime, I’ve been working on a prototype electric violin that is light in weight (for player comfort) and gives the kind of neutral signal we need. Last week, we finally succeeded in getting a plausible violin sound from our system, and then changing the sound by pushing a button.
By modifying “old Italian filters,” we hope to learn which acoustical features contribute to old Italian sound. The question will then become, what happens if these features are further enhanced? If we put electric violins with Strad filters into the hands of some good violinists, will they twiddle the knobs and eventually come up with sounds they like better? Is some kind of super Strad possible?
Sitting at my café table, teacup in hand, I imagine old Italian sound entering the digital domain, its further evolution now in the hands of players rather than makers. I imagine young violinists with hybrid acoustic-electric instruments exchanging their latest filter designs via cell phone.
Here at the café, all things seem possible.
But it’s 9:30 now and time to get back to my workshop, where an unfinished Strad copy glares at me from the bench.