The mechanism producing initial transients on the clarinet

A. Almeida,a) W. Li, J. R. Smith, and J. Wolfe

The University of New South Wales, Sydney, New South Wales 2052, Australia
(Received 19 May 2017; revised 3 November 2017; accepted 8 November 2017; published online 4
December 2017)
In self-sustained instruments, starting transients are important timbral characteristics that help
identify the instrument and the playing style. Often, the oscillation starts as a growing exponen-
tial. This study investigates the starting amplitude of this exponential for the clarinet. After a
rapid tongue release, the reed quickly returns to its equilibrium position. The sudden change in
aperture produces an abrupt change in both the airflow into the mouthpiece and the mouthpiece
pressure. This perturbation travels along the bore and reflects at the open end. Returning to the
mouthpiece with slight attenuation, the perturbation can be amplified by the reed acting as an
active element—effectively a negative resistance. When the reed release time exceeds the time
for sound to travel twice the bore length, the airflow and pressure wave into the bore via the aper-
ture are superposed over their own returning reflection. Measurements of reed motion and mouth-
piece pressures during reed release yield values that are used in a model to calculate waveforms
showing similarities to those observed experimentally. The initial amplitude decreases with
increasing reed release time, though not always monotonically. It can become very small in spe-
C 2017 Author(s). All
cial cases due to synchronisation between the initial pulse and its reflection. V
article content, except where otherwise noted, is licensed under a Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1121/1.5014036
[TRM] Pages: 3376–3386

I. INTRODUCTION 1987), or spectral content (Kergomard et al., 2000) in the

steady-state.
Attacks or initial transients are an important part of
Most studies on clarinet acoustics have concentrated on
wind instrument performance and expression (e.g., Brymer,
the sustained part of a note. Recently, the present authors
1977; Thurston, 1977; Gingras, 2004; Sullivan, 2006).
showed that the transient includes a stage during which the
Perceptually, they play an important role in the recognition
oscillation in the resonator grows at an exponential rate and
of the timbre of the instrument (Saldanha and Corso, 1964).
measured how the rate of this growth depends on blowing
A number of articulations are used for different musical
techniques, and these are executed in different ways by dif- pressure and lip force (Li et al., 2016b). This paradigm can
ferent musicians (Li et al., 2016a). Musicians control several be extended to situations when the control parameters vary
parameters, including the blowing pressure and how it varies during the transient, producing a note with a varying expo-
with time, the lip force, and the coordination of the tongue nential rate (Bergeot et al., 2014). A sudden change in the
release with the blowing pressure. rate of increase of the blowing pressure is sometimes
Acoustically, the clarinet is probably the most studied expected to cause a discontinuity in the amplitude of the
reed instrument and it is in some ways the simplest. An sound (Almeida et al., 2015).
approximately cylindrical resonator means that its acoustic The exponential phase ends as the note approaches satu-
response can be approximated by a delayed impulse, and the ration, i.e., once the amplitude of the pressure variation in
single reed is a simpler valve than the double reed (Almeida the mouthpiece becomes comparable with the blowing pres-
et al., 2007). Hence it is a suitable instrument for studying sure. The value of the amplitude at the start of the exponen-
how tonguing produces the initial transient on a wind instru- tial phase therefore plays an important role in determining
ment. Furthermore, the clarinet has been thoroughly studied the duration of the transient. Understanding the cause of this
both in the time-domain (Schumacher, 1981) and in the initial amplitude and how it depends on the blowing pressure
frequency-domain (Silva et al., 2008). For instance, the and the initial displacement and motion of the reed is the
quasi-static relation between pressure and flow at the reed main aim of this study.
(Dalmont et al., 2003) is well known and makes it easy to A previous study (Li et al., 2016b) indicated that, after
determine the pressure threshold for playing or the relation- release from the tongue, the reed approaches mechanical
ships between blowing pressure, lip force, and sound level equilibrium. The associated change in aperture creates an
(Dalmont et al., 2005), playing frequency (Bak and Dolmer, acoustic perturbation that propagates down the bore, is then
reflected and, following successive interactions with the
reed, establishes a standing wave. Using a clarinet in its
usual configuration, it is difficult to investigate the phenom-
a)
Electronic mail: [email protected] ena involved for three reasons. First, both the bell and the

3376 J. Acoust. Soc. Am. 142 (6), December 2017 0001-4966/2017/142(6)/3376/11 C Author(s) 2017.
V
open tone hole array have a cutoff frequency and other fre- II. MATERIALS AND METHODS
quency dependent properties, so the reflected pulse is dis-
A clarinet mouthpiece (Yamaha, Japan; model CL-4C)
torted and may sometimes be hard to identify. Second, the
is connected to completely cylindrical pipes of different
lowest note on the clarinet has a nominal frequency of
lengths, rather than to a normal clarinet bore. The inner
147 Hz, so the return of the first reflection occurs after about
diameter of the pipes, 15 mm, is similar to that of the clarinet
3 ms or less. This makes it hard to separate the initial pulse,
bore. The apparatus is aspirated rather than blown: the far
whose duration may also be several ms or more, from its
end of the pipe is terminated with a low-pressure reservoir
reflection. Third, the acoustical impedance of the vocal tract
that is lined with acoustic foam to provide a low impedance
or the artificial mouth in a playing machine can complicate
reflection condition. The (negative) pressure in the reservoir
the airflow and reed motion.
The first problem can be solved by replacing the clarinet is maintained by a pump (Positive airways pressure model
bore with a cylindrical pipe of similar diameter, open at the R261-708, ResMed, Sydney, Australia). The reservoir vol-
end remote from the mouthpiece, but without a bell. The sec- ume is 63 litres.
ond can be solved by making that pipe rather longer than a The exteriors of the mouthpiece and reed are exposed to
normal clarinet. Third, the mouth impedance can be reduced atmospheric pressure in the laboratory. This arrangement
to be much smaller than that of the bore. The present study ensures that the acoustic load on the upstream side of the
uses a mechanical lip and tongue and a controlled pressure reed is negligible at all frequencies of interest. Further, it
source to “play” a clarinet mouthpiece and reed connected to allows easy access to the reed, simpler control of the tongu-
a long cylindrical pipe. Rather than blowing at the mouth- ing mechanism, and optical access to the reed for high-speed
piece, suction is applied to a reservoir connected to the video recording. Because the mouth is at atmospheric pres-
remote end of the pipe. Thus the air in the laboratory sur- sure, the mouthpiece gauge pressure P is negative. The
rounding the mouthpiece, which has a very large compli- blowing pressure will be referred to as Pblow , which is calcu-
ance, becomes the very small “mouth” impedance. lated as P0 , the initial value of P. In the remainder of the
The experimental apparatus used in this article is article, varying physical quantities are represented by capital
described in Sec. II. In the first study presented here, the variables, and can be decomposed into a reference value
pipe acting as a resonator is 3 m long. With this apparatus, (with index 0, P0 for example); usually the value before the
the mechanical responses of reeds are measured with and reed is released, and the difference relative to this reference,
without a lip and with and without blowing pressure for a represented by the lower case (for example, p). For the
range of different initial reed transients (Sec. III A). These mouthpiece pressure the reference is the measurement
results are used to test a simple model for the initial acoustic before the reed release, P ¼ P0 þ p ¼ Pblow þ p, so an
wave that is similar to the “water hammer” phenomenon in increase in p in the mouthpiece pressure makes P less
hydraulics (Sec. III B). The water hammer describes the negative.
propagation of a sudden change in pressure produced when For ease of use, the reed is facing upwards, which is
the flow in a pipe is rapidly stopped by closing a valve opposite the usual use in clarinet playing (Fig. 1). A small
(Chaudhry, 1979; more details below). slab of polyurethane (Sorbothane, Kent, OH) simulating the
In further experiments, shorter cylindrical resonating lip is pressed against the reed by a suspended mass. This pro-
pipes are used so that the reflected wave can interact with vides a constant, known, and adjustable lip force applied to
the pressure perturbation produced by the change in reed the reed (up to 1.2 N in this study). The slab is 11 mm wide,
opening (Sec. III D). These results are compared with calcu- 13 mm long, and 4 mm thick with the masses hanging at the
lations using simple models for the initial pressure perturba- centre position, located 14 mm from the reed tip.
tion and the traveling wave it initiates. Further experiments For most experiments, a synthetic clarinet reed (Legère
conducted when the mechanical damping provided by the Reeds, Ontario, Canada, hardness 3) is used. In one initial
lips is removed give some insights into the production of experiment, a cane reed (Rico, Farmingdale, NY, hardness
“squeaks” (Sec. III E). 3) is also used.

FIG. 1. The experimental setup (not to scale). The controlled acceleration mechanism, with two masses, is shown in the main picture. The manipulating plate
setup is shown in an inset.

J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al. 3377
 Notes are initiated in three different ways. In all, the In this experiment, the tip of the reed is pressed toward
reed is attached to the mouthpiece with a ligature, as usual. the mouthpiece a distance of 0.25 mm, using the plate-
In the first (plate-release) method, a thin plate is clamped in release method described above, and suddenly released.
a two-dimensional micromanipulator, as shown in Fig. 1. Four different reed conditions are used. Two conditions used
The tip of this manipulating plate is positioned so as to pull a cane reed without a lip, dry in a first trial and for the sec-
the tip of the reed the desired distance up and away from the ond trial soaked in water for 30 s and then wiped with a tis-
mouthpiece. The plate is then withdrawn (in a direction par- sue. A synthetic reed (always dry) was used both without
allel to the axis of the clarinet), releasing the reed and allow- and with the pad used as part of an artificial lip, but in both
ing its elasticity to bring the reed back to near its equilibrium cases without the hanging mass used to apply lip force. The
position. The plate is also used to push the reed down toward first row of subfigures (plots labeled with “A”) shows the
the mouthpiece, then suddenly released. Except for a change reed displacement y(t) from its initial position before release
in sign, the results of the former experiments are very similar (initial displacement is indicated by zero in Fig. 2 and here-
and are not reported here. With this method, the motion of after); the second (label “B”) shows the pressure in the
the reed after release is determined by reed mechanical prop- mouthpiece and the third (label “C”) the magnitude of that
erties and cannot be experimentally controlled. pressure expressed in dB.
The second method (controlled acceleration) uses a Normally the sound generated by a vibrating reed would
small slab of the same polyurethane to act as an artificial travel down the bore of the clarinet and return quickly to
tongue, as shown also in Fig. 1. This tongue pad is operated interact with the reed during the reed’s motion. To avoid this
by two beams mounted on axles. First, the pad is pushed complication, the “clarinet” in these experiments was a
against the reed by a known force (controlled by mass m1) in 3.00 m pipe with a barrel that adds 3 cm, and a mouthpiece
the direction that closes the reed onto the mouthpiece. This that gives a total length of 3.12 m and thus a round trip time
tongue pad can be accelerated away from the reed at con- of 18 ms. In each of the examples in Fig. 2, the attenuated
trolled rates of acceleration using the second pivoted beam echoes are observed after 18 and 37 ms.
(with acceleration controlled by mass m2). The dry cane reed (plots A1, B1, and C1) has the highest
In a third (manual) method, beam 2 in Fig. 1 is removed natural frequency. (Cane reeds are usually not played dry:
when investigating large variations in tongue acceleration players usually wet them before playing, and they are kept
and the experimenter presses the left end of beam 1 with a moist by the player’s breath.) The synthetic reed with an
finger. It is difficult to produce a specific value of accelera- attached lip pad exhibited the lowest natural frequency
tion with this method, but it can produce a larger range of because of the extra mass. Without that pad, the natural fre-
tongue acceleration than the 2-beam system (see Li et al., quencies range from 1280 to 1400 Hz, which is toward the
2016b, for more details of the second and third methods). top of the normal playing range of the clarinet (say, sounding
From a sufficient number of trials, transients that approxi- F6 on a Bb clarinet or nominally 1400 Hz). From the expo-
mate the desired values may be chosen. nential decays of the oscillation y(t) in Fig. 2, Q factors are
During the experiments, video recordings are made of calculated: 17.0 6 0.9 (dry cane, column 1), 19.5 6 3.4 (wet
the motion of the tongue and reed using a high-speed cam- cane, column 2), 7.4 6 1.1 (synthetic, column 3), and
era, recording 8000 frames per second. The reed displace- 5.1 6 0.5 (synthetic with lip, column 4). The damping in
ment y(t) from its initial position is determined from analysis these conditions is lower than in normal playing because the
of the frames, and the acceleration is determined by fitting a lip pad moves with the reed as a compact mass and provides
parabola with zero initial slope to y(t). Simultaneously, the relatively little mechanical damping. Since the behaviours of
pressure in the mouthpiece, Pmp , is measured with a trans- the different reeds shown in Fig. 2 are qualitatively similar,
ducer (8507C-2, Endevco, Irvine, CA) inserted in the mouth- only the synthetic reed is used in the subsequent experi-
piece (30 mm away from the reed) and recorded using a ments. Synthetic reeds have the advantage that they can eas-
digital acquisition card (NI9234, National Instruments, ily be played dry and have stable physical properties during
Austin, TX) at 51 200 samples per second. A microphone long studies (Almeida et al., 2013).
and conditioning amplifier (4944A, Br€uel & Kjær, Nærum, When a lip force is provided by a hanging mass (tens of
Denmark) measure the sound pressure at a distance of grams in this study), the lip pad is subjected to a force from
95 mm downstream from the reed tip. (This position, corre-
the vibrating reed and a tension resulting from the weight
sponding to a location within the barrel of a clarinet, was
and inertia of the mass. Under these conditions, the motion
chosen to reduce the level of turbulence recorded and is
of the reed is more strongly damped (Fig. 3). In normal play-
hereafter called the barrel pressure pbarrel.)
ing, the player’s bottom lip would be subjected to a vibrating
reed on one side and forces related to the inertia of the jaw
III. RESULTS AND DISCUSSION
on the other. This condition should also damp the reed
A. Generation of a pressure pulse by the reed without motion.
a continuous airflow
B. The initial impulse produced by the change in reed
A reed can vibrate freely as a cantilever beam at its own
opening
resonant frequency, determined by its mechanical properties
and geometry. Examples of this vibration, triggered by an Figure 3 shows example results of experiments in which
initial displacement, are shown in Fig. 2. the transient is initiated using different controlled upwards

3378 J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al.
FIG. 2. Oscillation of the reed when suddenly released, with no blowing pressure. The top figures show the tip displacement y(t) from its initial position, mea-
sured from the camera images. The pressure pulse initially generated by the reed oscillation is reflected at the open end of the pipe and returns to the mouth-
piece after 18 ms. In each case, the inset shows the beginning of the transient with an expanded time scale spanning 6 ms and a vertical scale spanning 0.4 mm.
The acoustic pressure in the mouthpiece is shown on a linear scale in the second row and its magnitude is shown in a logarithmic scale 20 log10(jpmpj/20 lPa)
in the third row.

accelerations of the synthetic reed. The lip force was 0.7 N in y near the reed’s new equilibrium position, and similar
in each case. These experiments use a 0.89 m pipe connected oscillations are also seen in the barrel acoustic pressure,
to the mouthpiece and barrel instead of a clarinet (with a pbarrel . These oscillations are much smaller than those in
round-trip time lasting 5.8 ms as indicated by the horizontal Fig. 2 because, as explained above, applying the lip force
bar). The top row of the plots [Figs. 3(a) and 3(b)] shows the damps oscillations at the reed’s natural frequency.
displacement y of the reed tip from its initial position, and After about 6 ms, the pressure perturbation returns
the lower shows the acoustic pressure pbarrel in the barrel inverted after acoustic reflection at the remote end of the
position rather than in the mouthpiece, which has more tur- pipe. If the pipe were ideally open at the remote end (reflec-
bulent noise. This introduces a delay of 0.3 ms between tion coefficient 1) and ideally closed at the reed (reflection
measurements in the barrel with respect to the mouthpiece. coefficient 1), then the amplitudes of the positive and nega-
In these experiments, the blowing pressure (the average pres- tive excursions would be equal, in the absence of losses. In
sure difference across the reed) was 5.5 kPa. that case (ideal reflections, no losses), the first negative pulse
In this experiment, moving the reed away from the in pmp(t) (near 8 ms) would have twice the magnitude of the
mouthpiece shows a new effect; air enters the mouthpiece initial rise (near 2 ms), because of superposition of the reflec-
through the enlarged aperture and the mouthpiece pressure tion on the signal. In practice, the magnitude of the factor is
increases by about 0.2 kPa over about 2 ms. Note that this greater than 2, because of the amplification by the reed.
initial variation in pressure is approximately proportional to The variation in y(t) around 8 ms shows that the reed is
the reed displacement, and thus to the area of the aperture not a rigid reflector: it is displaced by the returning acoustic
between reed and mouthpiece. After the initial change in y, pulse. Further, its displacement changes the aperture and
there are some small amplitude, high frequency oscillations thus the airflow and pressure. Note that, here, the arriving

J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al. 3379
 Assuming that the kinetic energy gained by the air flowing
through the aperture is lost in turbulence in the mouthpiece,
the volume flow and blowing pressure are related by the
“Bernoulli term,”
sffiffiffiffiffiffiffiffiffiffi
2P
U¼S ; (1)
q

where q is the density of the air. Throughout this calculation

pjP0 jPA . q can vary slightly because of the variation in
pressure. However, the change in density is of second order
since q=q0 is on the order of p=PA . The value of q0 will be
used in the following.
Consider a hypothetical case where, at time t ¼ 0, there
is a discontinuous reed displacement y, leading to an imme-
FIG. 3. The reed displacement y (measured from its initial position) and diate increase in the flow into the mouthpiece.
barrel acoustic pressure pbarrel during and immediately after the “tongue” This step change in flow produces a column of air with
releases the synthetic reed from a position where it is pressed toward the
mouthpiece by a tongue force 0.2 N. The lip force was 0.7 N and blowing
slightly higher pressure and density. The air in that column
pressure 5.5 kPa. The pipe is 0.89 m long which, with barrel and mouth- is traveling down the pipe with speed ðU0 þ uÞ=A, which is
piece, gives a round trip time of 5.8 ms, as shown by the horizontal bar. The greater than the flow on the downstream side of the pressure
pressure change has a shape similar to that of the reed displacement until discontinuity (U0 =A).
the reflected acoustic wave comes back from the opposite end of the pipe.
Notice that when the tongue release is slower (right) the pressure rise is The length of this column is ct, the distance that the
also slower. pressure discontinuity travels at the speed of sound c into the
pipe (the bore in the case of a clarinet). For a pipe of cross
negative pressure pulse acts to close the reed and thus to section A, the extra mass of this column is q0 Act, so the extra
amplify the negative pulse. Similarly, the returning positive momentum of this column due to its extra speed is
pulse at 15 ms tends to open the reed, admitting more flow mv ¼ ðq0 ActÞðu=AÞ. From Newton’s law, the time derivative
and increasing the pressure. These amplification effects on of this momentum gives the force due to the discontinuity in
the subsequent evolution of the standing wave will be dis- pressure acting at the position of the density discontinuity:
cussed quantitatively later. pA ¼ q0 cu, so
u
1. The initial perturbation: The water hammer model p ¼ q0 c ; (2)
A
Figure 4 illustrates a variant on the water hammer effect,
which has similarities with the water hammer effect
with the difference that the sketch shows a sudden increase
(Chaudhry, 1979). Equation (2) can also be derived directly
in airflow, while the water hammer in plumbing is usually
by considering the characteristic impedance of the pipe
produced by a reduction in water flow caused by closing a
Zc ¼ q0 c=A. Introducing a step increase of flow u produces
tap. In the sketch, a suddenly increased aperture has admitted (before any reflections) a pressure increase of Zc u.
air at slightly higher absolute pressure and density, with the Relating this to the reed position Y, and substituting for
interface between high and low pressure regions propagating u,
at the speed of sound c.
sffiffiffiffiffiffiffiffiffiffiffiffiffi
In the present experimental system, the mouth is at y 2Pblow
atmospheric pressure PA (Fig. 4). The remote end of the pipe p ¼ q0 c w : (3)
A q0
is at (negative) gauge pressure P0 (so the blowing pressure is
Pblow ¼ P0 and the initial mouthpiece pressure is P0 ). The
aperture through which air flows past the reed is S ¼ wY, Considering a series of small pressure changes over time, the
pressure variation is given by
where Y is the vertical distance between the reed tip and the
mouthpiece and w is the effective width of the aperture.  
wc pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pðtÞ ¼ 2q0 Pblow yðtÞ: (4)
A

In the experimental sections of this study, yðtÞ is measured

using high speed video and the blowing pressure Pblow and
the downstream pressure variation pðtÞ are measured by a
FIG. 4. Schematics comparing the current experiment with the water-ham- pressure transducer in the mouthpiece or barrel. All the other
mer: The air is initially flowing through an opening S0 at a rate U0 into a quantities are known or measured, so the applicability of this
pipe, which is initially at gauge pressure Pblow . The opening (here equal to
model can be tested by comparing pðtÞ and yðtÞ; the change
w Y) then increases to S producing a region with flow U and gauge pressure
P ¼ Pblow þ p, where pjPblow jPA . The boundary region with increased in reed tip position, or by using yðtÞ to calculate pðtÞ and
pressure travels downstream at speed c, as shown. then comparing calculated and measured pðtÞ.

3380 J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al.
2. Experimental comparison with the water-hammer well as the reverse. In Fig. 6, the top row shows the initial
model several cycles of the oscillation of a synthetic reed with
In Fig. 3, the similarity in shapes between pðtÞ and yðtÞ the 3 m long pipe at three different values of the blowing
provides some evidence for an instantaneous proportionality pressure. The transient is initiated with the controlled
between the reed displacement y and the ensuing pressure acceleration system. The gray lines in the second row are
perturbation p in the resonator. More evidence comes from the change in pressure pmp measured by a pressure trans-
comparing the total pressure jump with the total reed dis- ducer located in the mouthpiece, 30 mm away from the
placement, measured between the instant the reed starts reed tip. This signal includes the continuous pressure off-
moving and when it comes to mechanical equilibrium [e.g., set, but is noisy because of the turbulence of the high-
from zero to about 3 and 7 ms in Figs. 3(a) and 3(b), respec- speed flow in this area. The black lines in the second row
tively]. In the simple model, the ratio ofpthese two changes is are the acoustic pressure pbarrel measured in the bore by a
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi microphone 95 mm downstream. Here, where the air speed
the proportionality coefficient ððwc=AÞ 2q0 Pblow Þ as given
by Eq. (4). is much slower, the turbulent noise is less and the signal
Figure 5 shows the pressure change calculated from is filtered by the microphone high-pass response at 8 Hz.
Eq. (4) plotted against the measured pressure change. The As in Fig. 2, the inset shows a 6 ms sample of the main
parameters A, Pblow , and q0 are known and the geometrical curve, including the beginning of the transient. The last
width of the mouthpiece slot (11 mm) is used for w. The row shows the logarithm of the absolute value of the
model, which has no adjustable parameters, fits approxi- pressure.
mately, but does not show the slightly stronger than propor- The first column of plots (A1, B1, and C1) shows a
tional dependence of p on y suggested by Fig. 5. Error bars typical result with a blowing pressure of zero, reproducing
are the result of noise in the measured y(t) due to the limited one of the cases in Fig. 2. The small mechanical oscilla-
spatial resolution of the camera. The calculation could be in tions of the reed are visible in the zoom insets. The small
error for two reasons. First, air may enter past the sides of (noisy) pressure variations visible on the log plot are again
the reed as well at the end, which would increase the effec- spaced by the 18 ms round trip imposed by the 3 m pipe.
tive width. On the other hand, because streamlines have to When the blowing pressure is below the threshold for auto-
bend around sharp corners of the inlet (in particular the reed oscillation (A2, B2, and C2), the water-hammer pressure
edge), the effect of the vena contracta (Hirschberg, 1995) is jump is visible at t ¼ 0. As in Fig. 3, the pressure jump
to reduce the equivalent cross-section of the channel and returns inverted, with slightly smaller amplitude. Here it is
thus w in Eq. (4). reflected at the reed end, which is, to a crude approxima-
tion, a closed termination. After two round trips (36 ms),
C. Acoustic oscillation and regeneration the cycle is repeated, each time with reduced amplitude.
When a lip force is applied, the reed is damped and (In the second and third columns, the slow rise in mouth-
this, combined with its stiffness, means that the immediate piece pressure is partly due to the finite compliance at the
acoustic effect of the initial reed motion—arising from the downstream end of the pipe, which lowers gradually the
proportionality of pðtÞ and yðtÞ due to the varying flow— blowing pressure, and partly due to the mechanical relaxa-
is of short duration. Further, once the reflection of the tion of the reed.)
pressure pulse returns, pðtÞ begins to influence yðtÞ, as In the third column (A3, B3, and C3), the blowing pres-
sure is above threshold, so each reflected water-hammer step
produces a pulse whose amplitude increases in successive
cycles. The shape of the waveform shows that, while the fun-
damental (H1) is increasing in amplitude, the highest har-
monics (H3, H5) responsible for the sharp corners in the
waveform are attenuated over the time scale shown here. In
the logarithmic plot of the acoustic pressure amplitude, an
exponential rise is observed. The exponential rise rate (and
thus the gain coefficient) is 48 dB s1. The middle column
(A2, B2, and C2) shows a decrease that is approximately
exponential: below the threshold, the regenerative energy
produced by the reed is less than that lost in the pipe and by
radiation (exponential rise and decay rates were modelled by
Li et al., 2016b).
In these examples, the long pipe (3 m) and the rapid ini-
tial motion of the reed together mean that the transient due
FIG. 5. The relationship between the measured pressure perturbation p and to the reed’s mechanical oscillation can be clearly distin-
that calculated from the measured change in reed aperture. Data points are guished from the acoustic reflection. This would not be the
from experiments where the lip force (Flip ), the blowing pressure (Pblow ), or
case for the clarinet under normal playing conditions, espe-
reed acceleration were varied. Vertical bars show the estimated uncertainty
in calculated pressure mainly due to the pixel resolution in the measurement cially for high notes, so Sec. III D reports experiments with a
of reed displacement. shorter pipe.

J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al. 3381
FIG. 6. The initial cycles of the oscillation of a synthetic reed measured with three different values of the blowing pressure Pblow but with same lip force of 0.7
N. As in Fig. 2, each inset on plots A1, A2, and A3 shows an expanded (6 ms  0.1 mm) sample of the reed displacement curve that includes the beginning of
the transient around t ¼ 0. B1, B2, and B3 show, in gray, the mouthpiece pressure measured by a transducer near the reed. The black curve shows the acoustic
pressure in the barrel, measured by a microphone 95 mm downstream from the reed away from the turbulence and high-pass filtered by the microphone (3 dB
point at 8 Hz). C1, C2, and C3 show the magnitude of the acoustic barrel pressure on a log scale 20 log10(jpbarrelj/20 lPa), and also the amplitudes of the funda-
mental, third and fifth harmonics (H1, H3, H5) as extracted by a heterodyne detection algorithm with a window of 150 ms. The pipe is 3 m long, giving a
round-trip time of 18 ms. (The threshold for oscillation lies between 4.5 and 5.5 kPa.)

D. Interaction of the reed’s mechanical perturbation

with the reflected wave
Figure 7 shows transients produced with a 0.89 m pipe
connected to the barrel and mouthpiece (giving a total length
of 1.01 m, which is about 50% longer than a Bb clarinet).
The transient is initiated by the controlled acceleration
method for faster initial reed motion, and manual control for
slower ones. This allows observation of different extents of
superposition of the pressure waves due to the reed’s
mechanical oscillation and the acoustic waves due to reflec-
tions at the end of the pipe. The blowing pressure is set to
the same value across all experiments. However, because of
the different tonguing conditions, its value changes transito-
rily following the removal of the tongue (by about 150 Pa or
3.6% of the blowing pressure). This produces different val-
ues of the exponential growth constant, which has values in
the range 145 6 14 dB s1. The lip force is constant for all
examples.
The first two cases shown in Figs. 7(a) and 7(b) (accel- FIG. 7. (Color online) The measured (thick dark line) and predicted (thin
pale line) barrel pressure measured in a pipe of 1.01 m total length. The pre-
erations 23.5 and 14.7 m s2) are the barrel pressure data dicted oscillation was calculated from yðtÞ (see text). The vertical bar shows
whose first several milliseconds were plotted in Fig. 3. Here the expected uncertainty in the calculated pressure due to the camera resolu-
the black lines show the growth of the oscillation as the tion in y, which is used to calculate the pressure. The inset view of the mini-
mum at 0.13 s shows that the calculation and measurement differ in their
amplification by the reed continues. The accelerations in
high frequency components, in part because of the filtering of the acoustic
Figs. 7(c) and 7(d) (3.2 and 0.9 m s2, respectively) pro- signal by viscous losses and radiation at the open end of the tube; these
duced longer durations of the initial mechanical perturbation, losses were not included in the calculation.

3382 J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al.
so that the initial reed displacement due to tonguing is not
yet finished when the first reflected acoustic wave arrives
back at the reed. This results in a smaller initial amplitude.
In Fig. 7, the initial displacement of the reed yðtÞ is used
to calculate a pressure perturbation pprimary ðtÞ according to
Eq. (4). The thin pale lines [ppredicted ðtÞ] are calculated using
a mathematical series based on this perturbation
 
X
þ1
k kT
ppredicted ðtÞ ¼ ðGÞ pprimary t  : (5)
k¼0
2

In Eq. (5), G is a gain factor determined from the measured

pressure signal (close to 1.1) and T=2 is the round trip time
(5.94 ms) giving the delay between successive reflections at
the reed.
With the exception of some wave shape differences in FIG. 8. The left panel shows five different yðtÞ with the same final amplitude,
Figs. 7(c) and 7(d), the agreement is reasonably good. The but different values of accelerations and therefore producing initial pressure
differences are in part attributable to the noise in yðtÞ due to perturbations of different durations. Sufficiently large durations of the pertur-
the small reed displacements, limited spatial resolution of bation (top three cases) will produce smaller amplitudes at the time of return
of the first reflection after one round trip. In the right panel, each of these ini-
the camera, viscous losses, radiation losses, and the fact that tial perturbations is superposed with successive inverted and amplified reflec-
the calculation does not include the mass of the reed, and so tions. Note that different amplitudes and waveforms result.
should be inaccurate for the higher harmonics.
largely a consequence of the different degrees of overlap
1. Simplified tonguing model between the opening and the return of the reflected wave.
Figure 7 and other similar data sets (not shown) demon- The resulting oscillation has an exponentially growing enve-
strate how the tongue motion, the reed motion, the water- lope, due to amplification associated with the active reflec-
hammer model, and a very simple model for the acoustic tion by the reed, here modelled by an imposed amplification
response of the pipe are able to reproduce semi- coefficient that is similar to that observed in smooth note
quantitatively two of the main features of the initial tran- attacks. Figure 8 shows a relatively short time interval in
sient: the starting amplitude and the consequent waveform. order to show the details of the waveforms, so the exponen-
Thus it is possible to obtain a good approximation to the tial nature of the increase is not obvious in that figure. In
profile of the start of the oscillation simply by superposing these examples, the amplitude of the exponential envelopes
the perturbations calculated using the water-hammer model. decreases as the reed acceleration is reduced, because small
The remaining differences in shape between calculation and accelerations produce only a small reed displacement when
measurement are chiefly in the high harmonics, and may be the reflected wave first returns. Hence the “clipped” wave-
due to the neglect of the mass of the reed in the calculation. form with flattened peaks, like those in Figs. 6(b) and 6(c)
This gives some confidence that simulation using the water and Figs. 7(a) and 7(b), is only produced with the two largest
hammer model and superposition of reflections will give accelerations, for which the reed aperture change is finished
reasonable approximations in calculations of how the oscil- before the return of the reflected wave.
lation grows for different profile shapes and rates of reed It is possible to run these simulations systematically for
opening. a range of values of reed acceleration, each having a differ-
Figure 8 shows simulations of the transients resulting ent reed release time, trr , and thus yielding different starting
from different reed accelerations. The model for the pipe is values of the exponential growth of the acoustic wave. To
deliberately simplistic, including only the delay and a con- calculate this dependence, the exponential envelope of the
stant amplification with value of 1.05 that mimics the active amplitude of the first harmonic is extrapolated to the moment
role of the reed and leaves aside any frequency-dependent of tongue release to obtain what is called hereafter the initial
losses. The panel at left displays the reed displacement and amplitude. This initial amplitude is plotted as a function of
mouthpiece pressure perturbation, which are proportional the reed release time by the black lines in Fig. 9.
before the first reflection. The panel at right displays the The final amplitude at saturation (not shown in Fig. 8)
pressure signal resulting from the superposition of succes- can be approximately calculated from the blowing pressure.
sive inverted and amplified reflections. The rate of exponential growth is known from the model pre-
In each calculation, the reed-mouthpiece aperture sented by Li et al. (2016b). The duration of the transient can
increases toward the same constant final value, but with dif- therefore be estimated from the initial amplitude with rea-
ferent values of the (constant) acceleration of the reed (rather sonable precision because the transition from exponential
like the experiments illustrated in Fig. 3), resulting in differ- phase to saturation lasts only several cycles (Li et al.,
ent reed release times trr . Different values of acceleration 2016a,b). Of course, the high initial amplitudes correspond
produce different waveforms and, especially for slow accel- to short attack transients, because saturation is approached
eration, different amplitudes. The change in amplitude is after fewer reed oscillations.

J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al. 3383
 is closer to longitudinal. In such a case, the water-hammer
effect would be rather more complicated to model.

E. Reed mechanical oscillations and squeaks

In previous results (Secs. III B, III C, and III D), the
application of the lip force via the deformable lip pad acts to
damp the reed, with the effect that a large fraction of the
reed’s mechanical energy is lost quickly, rather than estab-
lishing an oscillation at the natural frequency of the reed.
This is also likely to be the case in normal clarinet playing,
when the force from the lower jaw is transmitted to the reed
via the soft lower lip. With reduced damping, however, the
reed’s mechanical oscillations may interact with the return-
ing reflected wave. Depending on the phase between the
FIG. 9. Semi-logarithmic plot showing how the normalised reed release
reed oscillation and the acoustic reflections, oscillation at the
time affects the calculated dependence of the initial amplitude of the first reed frequency may sometimes benefit from the gain at the
harmonic. The three curves in black were calculated numerically using dif- reed, generating a high-pitched sound (Fig. 10), somewhat
ferent functions for yðtÞ: y ¼ f ðtÞ; 0 < t < t rr , where trr is the reed release
like the squeak feared by clarinettists.
time in the absence of a returning reflection. The time for a round trip is a
half period of the note. The initial displacement amplitude is the extrapola- Figure 10 shows the reed displacement and barrel pressure
tion of the exponential envelope of the first harmonic to t ¼ 0. The curves in for transients produced by the manipulating plate release
gray were calculated using an approximation to the measured reed motion method, with a 3 m long pipe and a blowing pressure of 5 kPa.
profiles (details in the Appendix).
At left, the reed oscillates freely without the damping of a lip,
showing the oscillation at a high frequency determined by the
In order to calculate the transient duration for a given set
reed, together with that at a low frequency determined by the
of parameters, it is desirable to have an analytical method to
pipe resonance. At right, with the deformable lip pad and a lip
calculate the initial amplitude. An approximate method for
force of 0.7 N, the oscillation at the high frequency is substan-
doing this is presented in the Appendix. Two sets of data cal-
tially damped. In the case shown in the left of Fig. 10, the
culated using this approximate method are included in Fig. 9
oscillations continue to grow quickly to a continuous, satu-
(gray lines) for comparison with the explicit simulations.
rated, high frequency oscillation (data not shown).
In Fig. 9, the overall behaviour is that the initial amplitude
decreases as the reed release time trr increases (tongue acceler- IV. CONCLUSIONS
ation decreases), as also seen in Fig. 8. However, local minima
are expected in some of the cases where the reflections of the Controlled initial reed displacements over a range of
perturbation profile interfere destructively. These minima are accelerations show that the initial pressure pulse produced
absent for the parabolic y(t), but they appear for linear and
sinusoidal y(t) (dashed lines in Fig. 9), i.e., they appear in cases
for which the derivatives y0 ðtÞ and y0 ðt þ T=2Þ are similar.
These simulations are consistent with the expectation
that a rapid note attack results from a quick release of the
reed. They also suggest that, for some reed opening time-
profiles, slow attacks may be prone to uncertainty because it
may be difficult to match precisely the release time to one of
the local maxima or minima shown in Fig. 9. This picture
may change for real profiles, which probably do not corre-
spond exactly to any of these functions. Different smooth
profiles tend (like the parabolic profile) to have less pro-
nounced maxima and minima. Furthermore, when the reed is
released slowly, turbulence and other sources of noise will
reduce the effect of destructive interference, smoothing out
the minima in the figure.
On a normal clarinet, the longest round-trip time (one-
half cycle of the note) is about 3 ms for the lowest note and FIG. 10. Initial reed and pressure oscillations when the reed is lightly
damped. In both cases, and as in Fig. 2, decaying oscillations at the natural
much smaller for many higher notes. Hence, even if the reed frequency of the reed (about 1300 Hz without lip and 600 Hz with lip) are
were released by a player as quickly as the plate release observed after tongue release at t ¼ 0. In the no-lip condition, their amplitude
method used here, superposition of pulse and reflection would is still visible at the first return of a reflected signal, at about 18 ms. At this
moment, the large, inverted pulse is superposed and amplified by the reed.
be expected for some notes. It is possible, however, that rather
This gives rise to a new, larger pulse of decaying oscillations. Further cycles
than releasing with the reed with a vertical displacement, the continue to increase in amplitude. When the lip is applied (with a force of
tongue may uncover the reed opening with a movement that 0.7 N), oscillations at the reed frequency are much more strongly attenuated.

3384 J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al.
by a moving reed and a blowing overpressure are well oscillations extend from 1 to þ1. This simplification
approximated to first order by a simple model related to the allows easier calculation of Fourier transforms and alters the
water hammer effect. As shown in a previous work, the start- shape of the Fourier peaks. The results in terms of the begin-
ing transient then follows an exponential envelope, in which ning of the oscillation are comparable with simulated ones
the time constant is determined by mouth pressure and lip (as shown in Fig. 9).
force. The value of the envelope at the start of the exponen-
tial depends on the blowing pressure and how the reed is
tongued. When the reed’s initial motion is sufficiently fast 2. Initial perturbation
compared to the travel time of the acoustic wave, the initial This is defined by a function of time pprimary ðtÞ that is
amplitude of oscillation is proportional to the change in reed constant for all times except for an interval ½0; trr , where trr
opening during tonguing. When the reed has not reached is the reed release time. pðtÞ ¼ 0 for t < 0 and pprimary ðtÞ
mechanical equilibrium before the reflected pulse returns ¼ pprimary ðtrr Þ for t > trr .
(i.e., for high notes or for slow initial reed motion), the two
effects are superposed. In these cases, a simple model can
reproduce semi-quantitatively some of the main features of 3. Reflection pattern
the initial amplitude of the acoustic pressure in the bore and A traveling wave pðx; tÞ travels back and forth in a
the consequent waveform. Simulation results are consistent waveguide, originating from a perturbation pprimary ðtÞ at one
with the experimental observation: The initial amplitude end of the waveguide (x ¼ 0). In the context of this article,
roughly decreases as the duration of tongue release pprimary ðtÞ represents the perturbation caused by the displace-
increases, but may exhibit local minima due to synchronisa- ment of the reed. The waveform pðtÞ resulting from succes-
tion between tongue motion and acoustic reflection. sive reflections in the pipe can be calculated using Eq. (5).
With the models shown here and in previous articles, the Considering the special case where the gain G ¼ 1, the
duration of the transient thus can be determined from the initial
oscillation can be extended to negative time without change
amplitude, the predictable amplitude of the acoustic pressure
in the domain under study (t > 0), by replacing the lower
in the bore at saturation, and the rate of exponential growth.
limit in the sum of Eq. (A1) by 1. The oscillating function
Under the conditions studied here, the natural frequen-
pðtÞ can be analysed in terms of its Fourier components. In
cies of cane and synthetic reeds without a lip range from
this case, the frequency dependence of the amplitude of p is
1280 to 1400 Hz and the Q factors range from 7 to 19.
These reeds present qualitatively similar behaviour. !
X
þ1
0 i2pkTf i2pðkþ1=2ÞTf
Without damping by the lip, the reed oscillates at its own p ðf Þ ¼ pprimary ðf Þ e e :
frequency after tongue release, and this oscillation may be k¼1
amplified by superposition of reflections from the bore if they (A1)
are in suitable phase; this behaviour resembles clarinet squeaks.
This is equivalent to the last equation in Sec. 4.3.2 of
ACKNOWLEDGMENTS Chaigne and Kergomard (2016).
We thank the Australian Research Council for support
of this project, Yamaha for the instrument, Legère for the 4. Amplitude of individual harmonics
reeds, and Joe Tscherry and the School of Mechanical and
Manufacturing Engineering of UNSW for lending the high- The individual terms of the sum in the previous formula
speed camera. are the amplitudes of each harmonic, if integrated over a nar-
row interval around each Dirac peak. The amplitude of the
kth harmonic is
APPENDIX  
k 1
1. Obtaining starting oscillation amplitudes from y(t) pk ¼ pprimary ð1  eipTf Þ (A2)
T T
This appendix presents an analytical model to estimate
the starting amplitude of an oscillation, given the gain of the (remembering that pprimary is a spectral density and so has
reflections and the profile of an initial perturbation that may, units of pressure multiplied by time).
in general, overlap the reflections. The first part of the model
follows Chap. 4.3 in Chaigne and Kergomard (2016).
Because saturation at the end of the transient is rapid, knowl- E. Example: Unit ramp
edge of the initial amplitude and the gain are sufficient to The unit ramp is given as an example,
give a good estimate of the transient duration, hence the 8
interest of an analytical estimation of the initial amplitude. >
> 0 if t  0
< t
The method is to treat the exponentially growing tran-
pprimary ðtÞ ¼ if 0 < t < trr (A3)
sient as a periodic function (the case that would occur with a >
> t
: rr
gain of one) multiplied by an exponential and to obtain the 1 if t  trr :
harmonics of one cycle of that transient waveform. For sim-
plification, the case treated has growth factor 1, and This is the integral of the rectangle function

J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al. 3385
 8
>
> 0 if t  0 This is the formula used to plot the gray dashed line in Fig. 9.
>
<1
Rect0;trr ðtÞ ¼ if 0 < t < trr (A4)
>
> trr Almeida, A., Bergeot, B., Vergez, C., and Gazengel, B. (2015). “Analytical
>
: 0 if t  t ; determination of the attack transient in a clarinet with time-varying blow-
rr
ing pressure,” Acta. Acust. Acust. 101(5), 1026–1038.
Almeida, A., George, D., Smith, J., and Wolfe, J. (2013). “The clarinet:
which can be defined in terms of the unit centred rectangle, How blowing pressure, lip force, lip position and reed ‘hardness’ affect
pitch, sound level, and spectrum,” J. Acoust. Soc. Am. 134, 2247–2255.
0 1 Almeida, A., Vergez, C., and Causse, R. (2007). “Quasistatic nonlinear charac-
trr teristics of double-reed instruments,” J. Acoust. Soc. Am. 121(1), 536–546.
1 @t  2 A: (A5) Bak, N., and Dolmer, P. (1987). “The relation between blowing pressure and
Rect0;trr ðtÞ ¼ Rect1=2;1=2
trr trr blowing frequency in clarinet playing,” Acta. Acust. Acust. 63, 238–241.
Bergeot, B., Almeida, A., Gazengel, B., Vergez, C., and Ferrand, D. (2014).
“Response of an artificially blown clarinet to different blowing pressure
The Fourier transform of the latter is profiles,” J. Acoust. Soc. Am. 135, 479–490.
Brymer, J. (1977). Clarinet (Macmillan, New York), p. 177.
sinð pf Þ Chaigne, A., and Kergomard, J. (2016). Acoustics of Musical Instruments.
Rect1=2;1=2 ð f Þ ¼ : (A6) (Acoustical Society of America Press, Springer, New York), p. 178.
pf Chaudhry, M. H. (1979). Applied Hydraulic Transients (Van Nostrand
Reinhold, New York), pp. 8–11.
And it is easy to derive the Fourier Transform Dalmont, J. P., Gilbert, J., Kergomard, J., and Ollivier, S. (2005). “An ana-
lytical prediction of the oscillation and extinction thresholds of a clarinet,”
J. Acoust. Soc. Am. 118, 3294–3305.
sinð pftrr Þ iptrr f Dalmont, J. P., Gilbert, J., and Ollivier, S. (2003). “Nonlinear characteristics
Rect0;trr ð f Þ ¼ e :
pftrr of single-reed instruments: Quasistatic volume flow and reed opening
measurements,” J. Acoust. Soc. Am. 114, 2253–2262.
Gingras, M. (2004). Clarinet Secrets: 52 Performance Strategies for the
Since Rect0;trr ðtÞ is the derivative of pprimary ðtÞ, i2pfpðf Þ Advanced Clarinetist (Scarecrow, Lanham, MD), p. 18.
¼ Rect0;trr ðf Þ, Hirschberg, A. (1995). “Aeroacoustics of wind instruments,” in Mechanics
of Musical Instruments (Springer, New York), Chap. 7, pp. 229–290.
1 sinð pftrr Þ iptrr f Kergomard, J., Ollivier, S., and Gilbert, J. (2000). “Calculation of the spec-
pprimary ð f Þ ¼ e : (A7) trum of self-sustained oscillators using a variable truncation method,”
i 2pf pftrr Acta. Acust. Acust. 86, 665–703.
Li, W., Almeida, A., Smith, J., and Wolfe, J. (2016a). “How clarinettists
articulate: The effect of blowing pressure and tonguing on initial and final
This statement ignores zero-frequency terms, since the inte-
transients,” J. Acoust. Soc. Am. 139, 825–838.
gral function in the time domain is defined up to a constant. Li, W., Almeida, A., Smith, J., and Wolfe, J. (2016b). “The effect of blow-
In the frequency domain, this corresponds to adding a Dirac ing pressure, lip force and tonguing on transients: A study on clarinet-
dðf Þ, which only makes a difference for zero frequency. So playing machine,” J. Acoust. Soc. Am. 140, 1089–1100.
the statement is true for the acoustic or AC case, which is Saldana, E. L., and Corso, J. F. (1964). “Timbre cues and the identification
of musical instruments,” J. Acoust. Soc. Am. 36, 2021–2026.
the one of interest here. Schumacher, R. T. (1981). “Ab initio calculations of the oscillations of a
Replacing the previous formula in Eq. (A2), clarinet,” Acustica 48, 71–85.
  Silva, F., Kergomard, J., Vergez, C., and Gilbert, J. (2008). “Interaction of
kptrr reed and acoustic resonator in clarinetlike systems,” J. Acoust. Soc. Am
sin 124(5), 3284–3295.
1 T
pk ¼ eikptrr =T ð1  eikp Þ: (A8) Sullivan, J. M. (2006). “The effects of syllabic articulation instruction on
i2p kptrr woodwind articulation accuracy,” Contrib. Music Educ. 33(1), 59–70.
T Thurston, F. J. (1977). Clarinet Technique (Oxford University Press,
London), p. 4.

3386 J. Acoust. Soc. Am. 142 (6), December 2017 Almeida et al.