How do violins change with playing and environmental changes over time?

Whether it is a product of the passage of time, the exposure to the atmosphere and its variations, or the amount of playing, many violinists believe that oldies are goodies, and spurn inexperienced instruments. When Michael Lea, curator of music at the Powerhouse Museum, wanted to buy for the museum's collection an instrument by renowned maker Harry Vatiliotis, and when simultaneously Sydney musician Romano Crivici commissioned an instrument from the same maker, an opportunity arose to examine these questions. Harry has a reputation for making fine instruments, and for making them reproducibly. This project began in 2001, when he made two very similar instruments from wooden plates, seasoned for 80 years, that had originally been intended to make a cello. Thus the bellies came from adjacent areas of the same slab of spruce, and both backs came from the same sample of maple.

During and after construction, they were subjected to acoustical tests by Ra Inta, PhD student in the Acoustics Lab at UNSW and, when finished, to playing and listening tests. Since then, one has been maintained in controlled, museum conditions. It is only occasionally played. Meanwhile, the other belongs to Romano and is his usual instrument for performance and practice.

We can imagine several ways in which age and playing might affect violins.

• Evolution of adjustable components. Players change various components and properties, such as the size and style of bridge, the type of string and the position and other properties of the sound-post. In each case, the player is likely to reverse the change unless it is perceived to improve the instrument. Thus a played violin might gradually 'evolve' towards a preferred region in its possible parameter space. There may also be another 'selection pressure': instruments that are unsatisfactory and that cannot be thus improved will usually have a lower market value and be played by less experienced players. Instruments that acquire a better reputation and market value will in general be sought and played by more experienced players. These improvements will occur independently of any change in the intrinsic mechanical properties of the instrument.
• Age-related mechanical changes. The intrinsic mechanical properties could change with age or with exposure to different environments and their daily and seasonal variations. For example, woods used in string instruments often have a high ratio of elastic modulus to density, and drying of wood over time would lower the density. However, there is no simple reason to expect that age-related changes in general would necessarily improve an instrument.
• Playing-related mechanical changes. Components of a violin undergo considerable low-amplitude, mechanical vibration during playing. In principle, one could imagine that this might alter the intrinsic mechanical properties. There is usually a strong correlation between the age of an instrument and the total amount of playing it has had. Again, there is again no simple reason to suggest that these changes will improve the instrument. (One might however argue that mechanisms that produce mechanical loss could be affected by sufficiently vigorous excitation.)

So, how have the Powerhouse Twins diverged over time?

Three tests have now been conducted. The instruments were compared when new, after 3 years (in 2004) and then four days later, after the regularly played instrument, and not the other, had been adjusted during a session involving the owner and the maker. During this session, new strings and bridge were fitted, and minor adjustments were made to the soundpost.

The first good news for the study came at the first set of trials. The panels of experienced violinists who performed the playing and listening tests on the new instruments gave results that showed no statistically significant differences. Harry really can make two very similar instruments! Another interesting result is that, in both of the further two tests, the panels of experienced violinists, who played the instruments blindfolded, or who listened to them, also gave results that were not significantly different.

The playing and listening panels were players who had not previously played or heard these instruments. What of someone who is very familiar with one of the violins? (Strangers may not be able to identify identical twins, but their parents often can.) In a separate, blindfold, playing trial conducted after the last comparison experiment, Romano Crivici, the owner of the played violin, was asked to play a scale and a short piece on the instruments as they were presented to him in random order, and to identify each instrument by saying "mine" or "museum's". He was correct in 20 out of 24 trials, which is significant at the 99% level. It is not known to what extent he may have used tactile cues. (The varnish of the regularly played instrument shows signs of wear.)

The expert panel give the two instruments statistically similar ratings and even Romano cannot distinguish them with certainty. Three years of playing (and three years of exposure to constant vs varying environmental conditions) is not enough, it seems, to make a substantial difference. Three years is not considered a long time for an instrument of which there are examples still being played after hundreds of years. The investigators hope that this study will continue, with this pair of instruments, for a time comparable with the age of these older violins.

More detail is available on the scientific report to be published by Acoustics Australia in the April 2005 edition.

A scale played on the rarely played instrument (same player).