At the stately, historical William Penn Hotel in the once-sooty, now cleanly-scrubbed city of Pittsburgh, Pennsylvania, members of the American Federation of Violin and Bow Makers gather for their 2005 convention: three days of lectures, exhibitions, panel discussions, and banquets. A distinctly British voice among the generally North American mix belongs to Jim Woodhouse, a Cambridge scientist and one of the most highly respected figures in violin acoustics. Woodhouse has been flown in to talk about his recent work on the violin bridge. In its published form, the work is highly technical, but Woodhouse has the rare ability to explain complex ideas in cheerful, workshop English. He also has a genuine interest in how violinmakers approach their work.
“I like to tag along behind them and listen to the kinds of things they talk about,” he says. “It’s one reason that I follow TOBI [an internet group dedicated to the technology of bowed instruments]. When violinmakers ask questions like, ‘What will happen if I cut a bit more wood off the f-holes here?’ there’s no way science can answer, at least not yet. We’re still groping toward the right way to frame these kinds of questions. But every so often something comes along which I think I can say something useful about. There is also an educational process I can help with. Makers have mental models of vibration and sound – it’s part of their intuition. These models may work well enough in practice – though not always for the reasons the makers think they do. But sometimes the model is quite obviously upside down, and that’s where I can have a role in cutting out some of the rubbish.”
The day before his talk, Woodhouse and I find a table in a quiet corner of the hotel’s vast and richly ornamented lobby. Perhaps in honour of the convention, loudspeakers hidden high among the marble arches have been playing the same violin recording for the past two days. Makers wearing nametags wander in and out of an adjoining Starbucks. “What has struck me most about the craft over the past thirty years,” Woodhouse says, “is the growth in openness among makers, the decline in secrecy. Though there are still makers who are skeptical about science – and quite rightly so – there is now a critical mass of people who are interested in exchanging ideas, and that is moving everything forward.”
Woodhouse was brought up in the suburbs of Brighton, where he showed an early aptitude for mathematics along with an enthusiasm for making things. “I grew up in a do-it-yourself household. There was always a workshop around.” At fourteen, in the thrall of Jimi Hendrix, he built an electric guitar. As an undergraduate at Cambridge, he believed he would become an astro-physicist, but his life inadvertently changed direction when he took violinmaker Juliet Barker’s long-running evening class in instrument making. He built a classical guitar, then a violin – and he has since finished two complete string quartets. His fascination with instrument making led him to a doctoral thesis in violin acoustics, for which he found an at-first-glance unusual advisor in the atmospheric physicist Michael McIntyre. “Michael is also a really fine violinist,” Woodhouse says. “The story I was told is that as a young man he entered a major competition, promising himself that if he won he would give up science and become a full-time violinist. Well, he came second.”
Professor Jim Woodhouse in his office at Cambridge University
Woodhouse is now fifty-four. He has blue eyes, a warm smile, a scraggly beard, and a sparkling mind. While some researchers have devised and promoted ‘scientifically-based’ systems for building instruments, Woodhouse carefully avoids that sort of thing. “It’s no use coming in and telling makers what to do. Scientific advice should come with a government health warning – may be harmful to your instrument! Science can be useful in sending you in the right direction, but if you follow some theory to the point where you start to think, ‘I don’t want to do this, but science says I must,’ then the science is probably wrong. Don’t make your violin unusually thick or thin or in some way bizarre just to achieve an octave relationship or some other such thing. Good violinmakers make good violins. Science follows along behind. There’s no way scientists are going to tell makers how, in any general sense, to do their job better.”
That said, there are ways in which science can help. “The wolf note is an unusual example of how a perfectly standard bit of engineering can be used to solve an instrument-making problem. Wood selection is another area that we know more or less how to approach. Old violins don’t lend themselves to being cut into test strips, so we have to get at the wood properties indirectly. The experiment you want to do is measure a piece of wood thoroughly, wait three hundred years, and then measure it again. It would be very nice to find Strad’s wood store. How does four hundred years of air drying and humidity cycling change a piece of spruce? Old wood gets chalky, doesn’t it? It crumbles easily. My best guess as to how it differs acoustically from new wood is that the high frequency damping is higher. [Damping refers to vibrational energy lost to internal friction.] This might explain why old violins seem better at suppressing the high frequency noise that can make new ones sound harsh.”
When I ask how far science has come in understanding the violin, Woodhouse says, “The image I like to use is that of a jigsaw puzzle. The edges are done, along with a few bits in the middle, but much of the picture is still blank. A jigsaw puzzle is a good image for how science generally proceeds. We do the edges first because they’re easier, then we work our way in. The strong point of science is that it is cumulative. I am completely confident that more and more of the picture will appear. It will take generations more learning, and we may never know everything there is to know about the violin, but it’s the attitude which is important.”
In conversation, Woodhouse often mentions his late friend David Rubio. “He had the most colourful background of any maker I’ve met – and he’s the only person I know who actually ran away from home to join the gypsies.” Rubio became a flamenco guitarist and then a renowned classical guitar maker before turning his attention to the violin. At a British Violin Makers Association meeting in the mid-eighties, Rubio approached Woodhouse and asked if he thought there was anything to John Chipura’s ideas about mineral grounds (“Radical Practices,” Strad, July 1984). He also approached two other Cambridge scientists, the chemist Ralph Raphael and the materials scientist Claire Barlow. A research collaboration was born. In 1989, Barlow and Woodhouse published a Strad article (March/April issue) that lent new credibility to the concept of mineral grounds – causing a spike in the sales of pumice, volcanic ash, mica, and other plausible candidates.
Rubio and Woodhouse became close friends and worked together on a number of research projects. For one of them, Rubio built a set of six violins. Each has carefully calibrated internal differences, but their near-identical appearance makes it difficult for players to tell them apart by visual clues or by feel. A photo of the instruments, laid out front-to-back on a straw mat, serves as a screensaver on Woodhouse’s laptop. Rubio taught Woodhouse’s son to build guitars, and after Rubio’s death from cancer in 2000, his tools and forms passed on to Martin Woodhouse, who is now a professional guitar maker.
Woodhouse’s talk is scheduled for 8:30 a.m. on the final day of the conference. I have heard him lecture several times before, and though I have yet to see him wear anything more formal than slacks and an open-collared shirt, he has changed this morning into a black slacks and a black, open-collared shirt. When speaking, he tends to sway from side to side with enthusiasm, and his delivery is as informal as his dress. He refers to a complicated looking graph as a “wiggly line” and then explains why violinmakers might want to take a second look at it. “Violins may all be different,” he says, “but they’re actually no more different, one from another, than most other manufactured objects.” He shows a slide of the sound output of ten Old Italian violins – ten wiggly lines laid on top of each other to show a range of individual differences and a common overall shape.
A wave of laughter rises as the audience digests the next slide: there seems to be about the same range of differences in the acoustical behavior of 98 cars of the same model off the same production line, and then of 41 nominally identical beer cans. “The thing about violins,” Woodhouse says, “is that people care about the differences. With most noise and vibration problems, the client isn’t much interested in the fine details of the acoustical behavior – they just want to know where to slap a little damping to make the noise go away. The violin is almost unique in that the fine details matter – and to an almost ridiculous level – making it a good challenge for the scientist.”
An acoustical feature which fine violins seem to share, Woodhouse explains, is a broad peak in sound output from about 2 kHz – 4 kHz, which happens to be the region of the ear’s greatest sensitivity. The eminent Swedish researcher Erik Jansson called this peak the “bridge hill” on the assumption that it was created by a resonance of the bridge itself (see figure 1). Jansson showed that by tuning the frequency of this resonance, the violinmaker could to some extent control the instrument’s treble response, and thus its brilliance and projection. Further research indicated that the characteristics of the bridge hill are determined not just by the frequency of the bridge resonance, but also by the mass of the bridge, the distance between its feet, and the ‘springiness’ of the top upon which it sits.
Figure 1: Rocking motion of the top section of the bridge associated with the Bridge Hill
To find out just how each of these parameters affects the hill, Woodhouse applied some of the latest tools from vibration theory to create a much-simplified mathematical model of a bridge atop a violin. Though this model existed only on a computer, when set into virtual vibration, a bridge hill did indeed appear in the instrument’s response curve. He then independently varied the mass of the bridge, the frequency of the bridge resonance, the spacing of the feet, and the thickness of the violin top, plotting graphs to show the effect of each (see figures 2-5). Never one to sail off on a theoretical model, Woodhouse then asked a Cambridge violinmaker to fit three distinctly different bridges onto some of the Rubio violins, which were then given to players for evaluation. This preliminary experiment nicely confirmed his predictions. Woodhouse has the bridges with him in Pittsburgh; he holds them up for the audience, explaining that they are DeJacques models with adjustable feet, so if anyone wants to try them on a violin here, they should find him after the talk.
Every maker knows that tiny variations in the cut and design of a bridge can make large differences to the sound of an instrument – differences that can make or break a sale. Woodhouse’s results provides a kind of map for makers wishing to systematize their approach to bridge adjustment so as to achieve specific tonal goals. All indications are that in cutting a bridge, the violinmaker has a surprisingly large and perhaps under-exploited degree of control over one of the violins most important acoustical features. This is some of the best news to come out of the research community in years.
The Federation meeting finishes that evening with a banquet. The organizers and guest speakers are publicly toasted and warmly applauded. By the time the waiters clear away the last of the coffee cups and wine glasses, a kind of end-of-convention nostalgia fills the room. By Sunday morning, the lobby is busy with bellhops carrying luggage. Violinmakers wave from taxis and rental cars. Woodhouse will stay on with a friend for a day or two and then fly home in time to start classes the following week.
When I ask about his next violin project, he describes a virtual instrument he intends to build using digital filters. Tonal changes that might take weeks to implement on a real violin will be done with a few keystrokes. The instrument will be played by recorded string signals from real players, and the resulting sounds played over loudspeakers. Panels of listeners will fill out questionaires. These will be analyzed, and if all goes well, another piece will fall into place on the jigsaw puzzle.
Figure 2:the frequency response of a simplified violin, as modelled on a computer. The skeleton curve (dashed line) shows the bump of the bridge hill.
Figure 3: The skeleton curve of the bridge hill as the bridge thickness is reduced. The dashed line represents the thinnest bridge.
Figure 4: The skeleton curve of the bridge hill as the bridge mass is increased – by adding a mute, for example. The dashed line represents the lightest bridge.
Figure 5: The skeleton curve of the bridge hill changing as the distance between the bridge feet is varied. The dashed curve has a spacing of 5mm, then 10mm, 20mm, 30mm and 40mm.
Figure 6: The skeleton curve of the bridge hill as the top plate is thinned near the bridge. The dashed line represents the thinnest plate.
A Woodhouse Reader
The following are some of Woodhouse’s many articles that are of immediate interest to violinmakers
“The acoustics of stringed musical instruments” M.E. McIntyre and J. Woodhouse. Interdisciplinary Science Reviews 3, 157-173 (1978). General review of violin acoustics.
“Microscopy of wood finishes” C.Y. Barlow and J. Woodhouse. J. Catgut Acoustical Society, Series II, 1, 1–15 (1988). Scanning electron microscope pictures of the effect of various woodworking activities.
“Firm ground? Old ground layers under the microscope,” Parts 1 and 2. C.Y. Barlow and J. Woodhouse. The Strad, March and April 1989. Ground layers in old Italian varnish.
Was old Italian spruce soaked?. C.Y. Barlow and J. Woodhouse. The Strad, 102, 1210, 234–239 (1991).
Micromechanics of permanent deformation in softwood. C.Y. Barlow and J. Woodhouse. Proc. 13th Riso Symposium on Materials Science, 213–219 (1992). Examining the effect of bending, rather than carving, spruce for violin tops.
The acoustics of “A0–B0 mode matching” in the violin. J. Woodhouse. Acustica/Acta Acustica 84 947–956 (1998). Examining the mechanism behind a popular adjustment “trick”, tuning a fingerboard mode to match the air resonance of the violin body.
Body vibration of the violin — what can a maker expect to control? J. Woodhouse J. Catgut Acoust. Soc. Series II, 4, 5, 43–49 (2002)
The bowed string as we know it today. J. Woodhouse and P.M. Galluzzo. Acustica – acta acustica 90 579–589 (2004).
On the “bridge hill” of the violin. J. Woodhouse. Acustica – acta acustica 91 155–165(2005). As discussed in the article.
Why is the violin so hard to play? J. Woodhouse and P.M. Galluzzo, Plus 31 October 2004, (online magazine of the Millennium Mathematics Project). Aspects of bowed string motion relevant to player, with animations