Brian Wise covers the classical music business for WQXR, including aspects of performance, technology, philanthropy and institutional trends. He manages the station's homepage and makes sure what you hear on air is what you see online. Follow him on Twitter at @Briancwise.
Using Science to Learn What Makes a Violin Sing
Monday, September 06, 2010
Does the secret to the pristine sound of an 18th century violin lie in the density of its wood? Probably not -- at least not entirely. But that is among the intriguing questions raised by a recent study of the Vieuxtemps, a Guarneri del Gesu built in Cremona, Italy in 1741, and described as the "Mona Lisa of violins."
In March, a team of violinmakers and scientists took the Vieuxtemps to Northwestern Memorial Hospital in Chicago to undergo a thorough examination under a CT scanner. The two-hour procedure -- using a kind of high-tech X-ray -- gave them a kind of topographical map of the instrument. Measured were things like the dimensions of the curves, or archings, the thickness of the wood Guarneri used and variations in density of that wood.
The Vieuxtemps, as it happens, is currently up for sale in Chicago, for an unprecedented $18 million. Three other instruments were also examined for the study: a 1742 Guarneri valued at $6 million, an $8 million 1707 Stradivarius and a violin made in 1752 by a lesser maker, J.B. Guadagnini, valued at $1.25 million. Joseph Curtin, a violinmaker from Ann Arbor, MI, proposed the study in order to better understand how the old violins were made.
"Here we have this instrument that seems to be excellent to both the player and the listener,” says Curtin. "I could do a measurement on 20 of my instruments and show some unusual results. But that wouldn’t mean anything if I didn’t show it on instruments I’m not related to."
Named after its most famous owner, the soloist and composer Henri Vieuxtemps, this particular Guarneri is said to be exceptionally well-preserved, with none of the patches or repairs that most old instruments have acquired. The results of the CT scans and modal analysis (a procedure to measure vibrations) are not yet complete, but preliminary findings appear to confirm that the great 18th century violins were made of woods slightly less dense and lighter than found in most other violins (this would support a 2008 Dutch study, which found an extremely consistent density in the wood of Cremonese violins).
Moreover, the Vieuxtemps project highlights a growing trend in the research community itself. For decades, a small group of physicists, engineers and acousticians studied old violins in an effort to understand their very specific physical characteristics. Gradually, a handful of violinmakers became interested in the research as well, but with more pragmatic goals. They soon began adopting these scientific techniques in order to learn how to make better sounding instruments that could rival the beauty of the old ones.
Now, as the scientific experts age and retire, violinmakers are increasingly driving the research. In 2006, Samuel Zygmuntowicz, a violinmaker in Brooklyn, who has built instruments for such noted musicians as Isaac Stern, Joshua Bell and members of the Emerson Quartet, teamed up with several scientists at the Oberlin Acoustics Workshop. They developed the Strad3D project, in which they examined three great Cremonese violins using the first-ever 3D laser vibration scans of violins, plus CT scans and modal analysis. A filmmaker documented the process for a DVD set.
The Strad3D study involved several components. One centered on the study of violins' F hole design, the two holes shaped like an F in Italics that allow sound to come out of the violin. Traditionally, modern violinmakers would copy the F holes of a Guarneri or Strad using plaster casts, photographs, even by tracing Xerox copies. But Zygmuntowicz wanted to move beyond the traditional empirical "lore" and better understand the underlying acoustic principals.
“What would happen if you made your F hole longer or shorter?” Zygmuntowicz asks. “What difference does it make if you go closer or wider or more scooped or not? Those are all design choices you have.” Changing the size and shape of the F hole, he pointed out, can have a dramatic effect on the sound of a violin.
For Curtin, scientific research has enabled him to challenge received wisdom about older instruments, namely, that they have a mellower, softer sound than modern ones. The recent Northwestern study suggested that the Vieuxtemps produces an surprisingly large, resonant sound -- an aesthetic more sympathetic to Curtin’s own. "My own voice as a maker, so to speak, tends in that direction, so I’m looking for outstanding examples of older instruments that confirm this,” he said.
Violins by Curtin and Zygmuntowicz often sell for $50,000 or more, a considerable investment for an average musician but small compared to the astronomical prices that rare 18th century string instruments have fetched. Last year, the "ex-Kochanski" Guarneri was sold to a Russian billionaire for $10 million, then the highest price ever paid for such an instrument. A Stradivarius called the "Lady Blunt" made in 1721 sold for the same price in 2008. Others have sold for several hundred thousand dollars.
With the extravagant prices have come fabulous (but not implausible) theories about the “secret” to the sound of these instruments. Many are based on the type of varnish or glue used in the 18th century, or the age and type of the trees from which the violins were built.
In the late 1980s, Joseph Nagyvary, a Hungarian-born scientist at Texas A & M University made chemical analyses of the wood of great violins and argued that the secret had to do with the fact that the logs providing the wood for the violins had been stored floating in water before they were used. That increased the permeability of the wood, allowing it to more fully absorb the varnish. Then, in 1997, Dr. Steven A. Sirr, a radiologist at Abbott Northwestern Hospital in Minneapolis, made CAT scans of more than 200 instruments of varying quality, and found that the wood appeared to have been treated with some high-density material.
Still, some question whether subtle variations in wood density or varnish explain the supposedly superior qualities of Strads and Guarneris. Among the skeptics is Stefan Hersh, a violinist and owner of Darnton & Hersh Fine Violins in Chicago. He calls the use of laser imaging and CT scans “overrated.”
“I think there’s a lot to be learned about factual details of what’s there,” he said. “But you can’t learn that much about Cezanne’s greatness based on how much canvas he used and oils and pigments.” Hersh believes that modern research tools offer valuable marketing data for modern violinmakers but there’s no replacing a craft that’s been handed down through generations, something that today’s makers don’t have access to.
Surrounded in his Park Slope, Brooklyn workshop by an array of precious instruments, carving tools, and blocks of wood, Zygumtowicz (left) acknowledges that hard data has dramatically increased his knowledge base. “My ability to use [the research] has gone up on a steady incline,” he said. “My awareness of what has gone there has expanded far faster than my ability to really use it.”
Yet he concedes that science can't provide a shortcut. “People who think that science is going to radically alter the practice of violin making, and we’re going to be working in labs with white coats, it’s not really the case,” he said. “One of the things that [pioneering acoustician and violinist] Oliver Rogers, showed me, moving his fingers around, is that finding a sound is an art. It’s all about touch. It’s not about putting it in the scanner and that using these visualizing tools from technology did not take away your judgment or intuition.”
In addition to building about six instruments a year, Zygumtowicz provides violinists with "tune-ups" of old ones. The ability, for instance, to use a tiny automated impact hammer to strike the bridge of a violin and then plot its resonance on a piece of software can make it easier to identify problems. “Sometimes if I’m playing this violin there’s an edgy spot and I say, ‘what is that?’ I can play into a microphone and analyze what’s happening and I’ll see there’s a certain high frequency. There’s a little spike, a little hot spot. I can put that frequency there and reach around and locate that spot.”
As CT scans and other tools offer new insights on old instruments, there's still no consensus on what separates good violins from great violins. That time may yet come as tools grow more sophisticated. But researchers must also consider human intervention. The sound of a violin depends greatly on the violinist, who holds it and plays it in a very specific way.
“The only real reality about violins is a person playing it and a person listening to it,” said Zygumtowicz. “All this stuff using imaging -- it’s about taking different views that are usually not visible. But that in no way replaces the actual experience.”