A stradivarius is definitely different

Medical information scientist Dr Berend Stoel measures the density of emphysema lungs using CT scans. An American violin-maker reading Investigative Radiology asked him to compare the density of the wood of Stradivariuses and Guarneris with modern violins. Overnight, as a result of the violin scan experiment, medical information scientist Dr Berend Stoel acquired the status of Stradivarius expert. But he also became the messenger of a musical ‘inconvenient truth’: even our violins sound worse than in the past because Bush is not dealing with the climate problem.


‘While we haven’t actually mentioned the climate,’ says Stoel, who, though expecting the media hype, is still rather taken aback by its aftermath. Together with violin-maker Terry Borman, he scanned three Guarneris, two Stradivariuses and eight top quality modern violins. He established an objective difference in their wood density: in the old violins this appeared to be neither higher nor lower on average compared to the new ones but at the level of the annual tree rings it was certainly distributed twice as evenly.


To what do the Cremonese violins from around 1700, the Stradivariuses and the Guarneris, owe their mythical fame? And what is the reason that present-day violin-makers, with all their knowledge and resources, are unable to equal this sound? Is it the varnish? The shape? The wood? Is it due to the special techniques for treating the wood? Or is it ultimately all in our minds?

Wood density

One of the hypotheses claims that the difference in sound depends on the difference in wood density. ‘Not such a strange idea,’ says Stoel. ‘This density also affects the way in which vibrations are transmitted inside the material.’ It has also been suggested that the colder climate around 1700 may have been responsible for a slower tree growth and may subsequently have resulted in a higher average wood density.

Stradivariuses and Guarneris

To determine whether there is in fact a difference in wood density between the classical Cremonese instruments and modern violins, Terry Borman, an American violin- maker, asked Stoel for assistance. Borman had read one of his articles in ‘Investigative Radiology’, which dealt with CT scan density measurements of lungs with emphysema. The violin-maker proposed hiring a scanner in the New York’s Mount Sinai hospital to compare a number of his clients’ Stradivariuses and Guarneris with modern violins. The makers, himself included, all obtain their wood for these new instruments - maple and spruce - from the same European forests as the great masters from Cremona used to use.

Regular density

The result of the experiment could be read on the website of the scientific open access magazine PLos ONE on 2 July and instantly became world news: objective differences were indeed discovered between the eighteenth-century violins from Cremona and the present-day violins.

Springtime wood

This regularity is due to the relationship between the density of springtime wood and the wood which was formed at a later stage during the year. In the old violins this relationship is much more balanced. Stoel: ‘There was one modern violin which also had it. We called its maker and this wood had been treated with water. We know that the Cremonese builders did not do this.’

Finger exercise

Whether the sound difference can be explained or not - ‘never entirely’ says Stoel, - this objective difference is a fact. Violin-maker Borman wants to make use of his newly found knowledge to produce even better violins, by selecting or treating his wood. Stoel: ‘As a scientist I am particularly excited that we have been able to measure objectively such a subjective experience as sound. Besides, for me personally it was a wonderful finger exercise, which indicates how incredibly reproducible and sensitive this measurement is. And consequently, how well we can measure lung emphysema.’

Violin-maker Terry Borman and medical information scientist Berend Stoel (r) with the scanner in New York. In front of them are some 12 million dollars worth of violins

Hereditary lung emphysema

Stoel is head of the division of ‘lungs and bones’ in the medical image processing research group of the LUMC ( Leiden University Medical Center). During the past 15 years, together with lung specialist Dr Jan Stolk, he has been working on a way of measuring the of patients with hereditary lung emphysema using CT scans, applying this method as an objective and reproducible alternative for lung-function measurements

Selecting images

Stoel: ‘Lung emphysema destroys lung tissue. So we assume that the tissue’s density decreases. We measure this lung density with X rays. The greater their absorption, the higher the density. My task is to adjust the CT scanner in such a way that it can measure the density optimally. From the CT data I can automatically select the lungs and conduct the measurements. Meanwhile, clinical trials are in progress. My computer programme is used to show whether certain treatments are effective.’

Soft parts

Stoel, who plays saxophone and piano, but at one time even played the violin for five years, found the violin-maker’s request both attractive and exciting. ‘Just like the soft parts of a body, wood inherently has a low density. But violins aren’t bodies, so I devised a new computer programme in my spare time. It was quite a task. Lungs have considerably more volume than the front and back plates of violins. That meant using far more voxels.’

New York

After having conducted experiments in the CT scanner in the LUMC with separate front and back plates of unfinished violins, Stoel flew to New York last year. First he repeated the Leiden experiments and, based on measurements, determined the best protocol. The next day, Borman and Stoel together with four violins – each worth about 3 million dollars - took a taxi to Mount Sinai hospital. Stoel: ‘We had a choice: either go for maximum insurance, or none at all. We chose the last option.’ In another session a fifth old violin was scanned. In total, Stoel scanned three Guarneris and two Stradivariuses.

Surface selection

Stoel: ‘We scanned them, I burned the data onto a CD and wrote a computer programme to extract all the right information. From the data I had to select a violin surface which was original, unprepared and as large as possible. It had to be the same surface for each violin. Unfortunately we were not allowed to remove the strings and the fine tuners. It wasn’t easy but we did finally succeed.’


But what causes this homogeneity in the wood? And what about that cold climate in the early eighteenth century? Stoel: ‘We can indeed only speculate on the underlying cause of the differences we found.’ ‘Some journalists deduce all kinds of information from this and relate our results to the seventeenth and eighteenth century ‘Little Ice Age’. On one American site there was even a stating that it was Bush’s fault that we no longer produce beautiful violins, as he is taking no action on the climate problem. But we made no claims about the ice age.  In fact, we never found that higher than average wood density which according to the hypothesis would indicate this.’

300 years

‘What do I think? I believe, but it is of course only a vague suspicion, that it has to do with its age of 300 years. It may be that the average density was higher in the past and has reduced over the centuries, while its homogeneity has remained intact.’

Last Modified: 29-08-2008