As some of the most discriminating materials consumers, musicians have a reputation for conservatism that verges on superstition. The materials scientists who labour to provide instrument-makers with an alternative to a rare traditional material, such as pernambuco — the material of choice for violin bows until the source tree, Guilandia echinata, became an endangered species — can expect scant thanks. Try as they might to point out that a carbon-fibre composite bow has superior mechanical properties (and less chance of shearing off at the head), the violinist is sure to insist that it just doesn't 'play' as well.

This acute sensitivity, if not subjectivity, towards musical materials raises the question of how far their selection can be quantified, for example with the materials selection charts pioneered by Mike Ashby at Cambridge University. That is the subtext of a survey of woods used in musical instruments by Ashby's former student Ulrike Wegst, now at the Max Planck Institute for Metals Research in Stuttgart (Am. J. Bot. 93, 1439–1448; 2006).

Materials selection charts are two-dimensional spaces in which the coordinates are two properties – density and Young's modulus, say. Each material has a specific location on this plane. Materials applications often demand a compromise between at least two such quantities, in which case the candidates are those that fall within the appropriate elliptical field on the map.

Wegst shows how the acoustical properties that determine a wood's suitability for a type of instrument – a xylophone bar, say, or a violin's soundboard or a clarinet body – can be classified according to just a few parameters, such as the speed of sound, the density, and the loss coefficient that describes damping.