314
6 Shocks and Surprises: Refining the Elementary Model
the instrument mouthpiece had an amplitude similar to that generated by a human
player and a frequency spectrum rich in upper harmonics.
The frequency of the structural resonance illustrated in Fig. 6.23 was close to the
fourth harmonic of the note B 1. When this pedal note was played by the acoustic
driver, a microphone measuring the sound output on axis in front of the bell did not
register a systematic change in the amplitude of the fourth harmonic for different
bell thicknesses, suggesting that the wall vibration was having a negligible effect
on the sound. However a microphone in the left ear of an artificial head in the
normal human playing position showed that the amplitude of the fourth harmonic
was several dB higher for the 0.3 mm bell than for the 0.5 mm bell. A similar result
was found when the note B 2 was played: the structural resonance frequency was
close to the second harmonic of this note, and at the position of a human player’s
left ear, the amplitude of the second harmonic was several dB higher for the thin
bell than for the thick bell.
The radiated sound measurements reported by Smith suggest that the timbral
modification caused by a coincidence between a structural mode and an acoustic
mode might well be perceived by a trombone player, although it would probably be
inaudible to a distant listener. To test this hypothesis, ten professional trombonists
were invited to play and evaluate the six instruments. The players were blindfolded,
and precautions were taken to equalise the weight and balance of the bells
(see Sect. 1.3.3). Under these circumstances none of the players were able to
distinguish between thick and thin bells (Smith 1986). Thus although there were
clearly measurable spectral differences at specific frequencies between the sounds
generated using different bell thicknesses, these differences could not be considered
musically significant in a realistic performance situation.
In the first decade of the twenty-first century, Thomas Moore and Wilfried Kausel
revisited the question of the musical effects of brass instrument wall vibrations
in a series of experimental studies (Moore et al. 2005; Kausel et al. 2008, 2010).
Structural modes of trumpets and horns were observed using electronic speckle
pattern interferometry, and investigations were undertaken on the influence of wall
vibrations on the sound radiated from instruments excited both by acoustic drivers
and artificial mouths (see Sect. 3.1.6). The experimental arrangement used by Moore
at Rollins College to study an artificially blown King ‘Silver Flair’ trumpet is
illustrated in Fig. 6.24a. The artificial lips were in this case made from solid rubber
and with a mouth pressure of around 20 kPa played the note B 4 (frequency 466 Hz).
The bell of the trumpet was placed directly in front of a circular aperture in the
wall of a small anechoic chamber which contained the measuring microphone. The
vibrations of the bell could be heavily damped by packing sandbags around it.
The relative strengths of the harmonics in the measured spectra were found
to depend strongly on the embouchure setting, but a significant trend emerged
when the results of several experiments with different embouchures were averaged
(Fig. 6.24b). The powers in the first and second harmonics were increased when the
bell was damped, while the powers in the harmonics from the third to the eighth
were diminished by damping.
6 Shocks and Surprises: Refining the Elementary Model
the instrument mouthpiece had an amplitude similar to that generated by a human
player and a frequency spectrum rich in upper harmonics.
The frequency of the structural resonance illustrated in Fig. 6.23 was close to the
fourth harmonic of the note B 1. When this pedal note was played by the acoustic
driver, a microphone measuring the sound output on axis in front of the bell did not
register a systematic change in the amplitude of the fourth harmonic for different
bell thicknesses, suggesting that the wall vibration was having a negligible effect
on the sound. However a microphone in the left ear of an artificial head in the
normal human playing position showed that the amplitude of the fourth harmonic
was several dB higher for the 0.3 mm bell than for the 0.5 mm bell. A similar result
was found when the note B 2 was played: the structural resonance frequency was
close to the second harmonic of this note, and at the position of a human player’s
left ear, the amplitude of the second harmonic was several dB higher for the thin
bell than for the thick bell.
The radiated sound measurements reported by Smith suggest that the timbral
modification caused by a coincidence between a structural mode and an acoustic
mode might well be perceived by a trombone player, although it would probably be
inaudible to a distant listener. To test this hypothesis, ten professional trombonists
were invited to play and evaluate the six instruments. The players were blindfolded,
and precautions were taken to equalise the weight and balance of the bells
(see Sect. 1.3.3). Under these circumstances none of the players were able to
distinguish between thick and thin bells (Smith 1986). Thus although there were
clearly measurable spectral differences at specific frequencies between the sounds
generated using different bell thicknesses, these differences could not be considered
musically significant in a realistic performance situation.
In the first decade of the twenty-first century, Thomas Moore and Wilfried Kausel
revisited the question of the musical effects of brass instrument wall vibrations
in a series of experimental studies (Moore et al. 2005; Kausel et al. 2008, 2010).
Structural modes of trumpets and horns were observed using electronic speckle
pattern interferometry, and investigations were undertaken on the influence of wall
vibrations on the sound radiated from instruments excited both by acoustic drivers
and artificial mouths (see Sect. 3.1.6). The experimental arrangement used by Moore
at Rollins College to study an artificially blown King ‘Silver Flair’ trumpet is
illustrated in Fig. 6.24a. The artificial lips were in this case made from solid rubber
and with a mouth pressure of around 20 kPa played the note B 4 (frequency 466 Hz).
The bell of the trumpet was placed directly in front of a circular aperture in the
wall of a small anechoic chamber which contained the measuring microphone. The
vibrations of the bell could be heavily damped by packing sandbags around it.
The relative strengths of the harmonics in the measured spectra were found
to depend strongly on the embouchure setting, but a significant trend emerged
when the results of several experiments with different embouchures were averaged
(Fig. 6.24b). The powers in the first and second harmonics were increased when the
bell was damped, while the powers in the harmonics from the third to the eighth
were diminished by damping.