What determines the quality of wood for wind instruments?
Sounding qualities
While in first instance we stay away from the discussion whether the vibration of the wood itself has influence on the sound of the instrument or not, the starting point is the vibration of an air column within a wooden body. On an instrument with the same design, a different wood will give a different sounding instrument as the wall surface (and the exact form of where the finger holes meet the main bore) will have different qualities. If the wood is very dense and closed grain, the wall surface will be relatively smooth (though not as smooth as polished metal). So the friction between the vibrating air and the wall will be relatively low, allowing also high frequencies to participate in the sound spectrum. Musicians will speak of a clear sound. If this goes too far one can speak of ‘shrill’ or ‘metallic’ sound. In less dense woods, at the cell structure the vibrating air will subsequently meet wood and cavities, hence the higher frequencies of the sound spectrum will be filtered, leaving just only a couple of harmonics. The sound will be less clear and less powerful, but this can also be called ‘warmer’ and ‘more woody’. For flutes generally woods less dense than boxwood will not be used, as the sound will be rather weak. For bagpipes, also lighter woods such as plumb have their adepts, as it is true that especially for higher pitched instruments the sound will be less penetrating and crying. As the wall smoothness is the main determining factor, it is easy to understand that impregnating the wood with epoxy or another product can completely change the sound of the instrument. Hence also the big difference between an oiled and an un-oiled instrument.

Dimensional stability of the wood
Wood that comes into contact with humid air (or liquid water), will absorb water and swell. When the air becomes drier, it will loose water and shrink. This holds for all wood, even for wood that is centuries old. If a flute that has been played regularly (so the wood around the bore is wet, this in a swollen state) is put into a very dry room, the outside of the flute will dry and shrink. As the inner part of the instrument is still wet, and cannot shrink, the risk of cracks developing on the outside becomes important. But not all wood species do shrink and swell to the same extent. For some species the difference between completely wet and completely dry can be as much as 10%, for others this is only 2-3%.
There is also a difference in the speed with which wood species absorb moisture: a dense wood with a lot of wax-like substances such as African Blackwood will absorb moisture much slower than a ‘dry’ wood such as pear. Even when oiled such differences will still remain to some extent. A last point is that some timber species split more easily than others, snakewood being a species that splits rather easily. Snakewood will not be used much for instruments coming into contact with moisture such as flutes or mouth-blown bagpipes.

But this is not all: wood does not swell equally in all directions. It hardly at all changes length-wise, and there is a difference of generally the single to the double between the radial direction and the tangential direction of the wood, but again this ratio is different with different wood species. This difference is the reason why the bore section of old instruments has often become oval instead of remaining circular. It is clear that timber species that absorb water slowly, that have generally low shrinkage percentages, where the difference between tangential and radial shrinkage is low, and that don’t split easily will be the preferred woods for making (mouth blown) wind instruments. (For more information on swelling and shrinking of timber species, see ‘Physical properties of 140 timber species’ by Peter Laming and Jan Rijsdijk, edited by TNO Delft, Netherlands)

TEKENING 3 RADIALE EN TANGENTIALE RICHTING, MET KRIMP^

To push the complexity further, when wood pieces swell and shrink subsequently under the influence of changing moisture content, it has been observed that after a long series of such cycles: the wood ‘collapses’ to some extent. After this the wood becomes definitively smaller, and although it will still react to changing moisture contents, these dimensional changes will be less than before. This phenomena has as yet not been studied very well (it takes a long time to do so). Only for beech do we know that the ‘collapse’ happens after about 20 complete cycles. Although it has no practical consequences for the choice of timber species, his collapse phenomena explains different things: