Taguti T., Okamoto M.

The top portion of the lowest register of clarinet, called the 'break' or throat register, is known as the most troublesome portion to the average player in keeping proper pitches and producing characteristic timbre. A remedy for this is the use of certain cross fingerings as alternate ones of the normal fingering. Here, the normal fingering refers to a fingering that allows full venting so that all the closed tone holes are at the upper end and all the open tone holes are at the lower end, while the cross fingering refers to a fingering that closes a tone hole or holes below an open one either to partially prevent venting (or to partially allow additional venting.) This paper aims at the study of the acoustical characteristics of the normal and cross fingerings in the above context. Experiment was done on Ab note in the throat register of a Bb clarinet. Total 39 cross fingerings (labeled CF1 to CF39) associated with the normal fingering (labeled NF) were provided to compare with NF for the target note. A female player, who had experience in playing the clarinet for more than ten years, was asked to blow the target note in the way that the tone with NF is followed by the tone with one of CF1 to CF39, and vice versa, by keeping a constant pressure as much as possible without break between the two. The main result of the spectral analysis is summarized as follows. The power spectra of tones with certain CFs showed a distinctive feature that the low partials, typically those of the third to sixth, are stronger in power than the corresponding partials produced with NF. In particular, cross fingering labeled FC1 that closes an additional tone hole with the forefinger (of the right hand) had the most significant effect in this respect. Since relatively strong odd partials, especially a relatively strong third partial, is the general characteristic of the clarinet tone, FC1 is suited to play the concerned note by the player. Further experiment is under way with another player for studying the validity of the result obtained as above, and a possible existence of individual difference.