The effect of the coupling between the top plate and the fingerboard

on the acoustic power radiated by a classical guitar (L)

Héctor Garcı́a-Mayén
Facultad de Ciencias, Universidad Nacional Autónoma de Me´xico, Apdo. Postal 21-726, 04000 Me´xico D.F.,
Mexico

Arturo Santillána)
Institute of Sensors, Signals and Electrotechnics, University of Southern Denmark, Niels Bohrs Alle´ 1, 5230
Odense M, Denmark

(Received 20 July 2010; revised 17 December 2010; accepted 20 December 2010)

An experimental investigation on the coupling between the fingerboard and the top plate of a classical
guitar at low frequencies is presented. The study was carried out using a finished top plate under fixed
boundary conditions and a commercial guitar. Radiated sound power was determined in one-third
octave bands up to the band of 1 kHz based on measurements of sound intensity. The results provide
evidence that the way in which the fingerboard and top plate are coupled is not a relevant factor in
the radiated acoustic power of the classical guitar in the studied frequency range.
C 2011 Acoustical Society of America. [DOI: 10.1121/1.3543968]
V

PACS number(s): 43.75.Gh [NHF] Pages: 1153–1156

I. INTRODUCTION II. EXPERIMENTS

Guitarists consider the acoustic power radiated by the
classical guitar as a relevant characteristic of the sound pro-
duced by the performer. Sometimes guitars are constructed
with a small gap between the fingerboard and the top plate. The
underlying idea is that the top plate will radiate more acoustic
power if its motion is not restrained by the fingerboard.
In the classical guitar, the plucked strings excite the bridge
and top plate, and the vibrations are transferred to the air cavity,
ribs, and back plate.1 Vibrations of the resonance box and the
coupling of the soundboard and the back plate via the cavity
fluid have been characterized.2–4 The coupling between other
parts of the guitar have been scarcely investigated.
Lai and Burgess5 reported that the radiation efficiency
spectrum measured in a guitar between 50 and 550 Hz exhib-
ited minimal changes at frequencies below approximately
200 Hz when the top plate was loaded with additional mass
distributed below the waist. However, the radiation effi-
ciency was significantly reduced at higher frequencies.
A guitar with a soundpost coupling the top and back
plate near the sound hole was studied by Le Pichon et al.6
The results showed that at some frequencies the back plate
vibration was much less important than the vibration of the
top plate. However, there were certain frequencies where the
contribution of the back plate to the radiated sound was sig-
nificant; examples at 327, 501, and 579 Hz were presented.
In a second paper, published by Chaigne and Rosen,7 one of
the main conclusions was that the coupling between the top
and back plates by means of the soundpost reduced the
mobility of the top plate, affecting negatively the radiation
of sound from the instrument.
a)
Author to whom correspondence should be addressed. Electronic mail:
aos@sense.sdu.dk
In the experiments reported in this letter, the radiated
sound power was determined up to the one-third frequency
band of 1 kHz based on the following facts: (i) According to
Christensen,8 the most relevant modes in the instrument are
those with strong monopole components, which are also re-
sponsible for radiating much of the sound energy. (ii) In the
considered frequency range, the radiation pattern of the gui-
tar is dominated by monopole and dipole components;9
therefore, errors in the determination of the acoustic power
by measurements of the sound intensity were expected to be
significantly small due to the relative simplicity of the radi-
ated sound fields.10,11
A first approach to the problem was to investigate the
coupling between the top plate and the fingerboard but with-
out the effect of the resonance cavity. Hence a complete top
plate of a classical guitar, made by and expert luthier, was
used for the experiments, comprising the fan struts arranged
in the traditional Torres distribution, two transversal bars on
its upper bout, the bridge, and the saddle.
The top plate was placed on a rigid structure of high den-
sity wood with the shape of a guitar, which included the ribs,
foot, tail, neck, and head, but without back plate. The ribs of
the structure had a width of approximately 1 in. to impose
fixed boundary conditions on the contour of the top plate
when it was forced to vibrate. The top plate was trapped
between the ribs of the supporting structure and a piece of
wood, which had the same shape and width of the upper face
of the ribs. This piece of wood, half an inch thick, was
attached to the ribs by means of 48 screws evenly spaced.
Three different configurations of the same top plate
were investigated: (a) The top plate without a fingerboard;
(b) the top plate with a section of a finished fingerboard
glued to it at only four points (see Fig. 1), leaving a small
gap of approximately 1 mm between the top plate and the

J. Acoust. Soc. Am. 129 (3), March 2011 0001-4966/2011/129(3)/1153/4/$30.00 C 2011 Acoustical Society of America
V 1153

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 132.174.255.116 On: Fri, 28 Nov 2014 09:35:45
 piece of fingerboard; and (c) the top plate with the finger-
board glued together as it is traditionally constructed in a
common guitar.
The effect of the coupling between the top plate and the
fingerboard on the radiated sound power in a commercial
guitar was studied in the second part of the investigation.
The sound power radiated by the complete guitar was first
determined. The experiments were repeated with the same
guitar after the section of its fingerboard from the 12th fret
and above was removed.
A. Determination of the radiated sound power
The sound intensity was measured using two closely
spaced pressure microphones.10,12 A real time signal ana-
lyzer of the type B&K 2133 was used combined with a
sound intensity probe of the type B&K 3545 and a remote
control of the analyzer of the type B&K ZH0354. In all the
measurements a spacer of 50 mm was used between the two
microphones of the type B&K 4181. That spacer allows
measuring the sound intensity from the one-third octave
band of 20 Hz up to the one-third octave band of 1250 Hz.
To measure sound intensity at higher frequencies, a smaller
spacer is required. A windscreen of the type B&K UA0781
was mounted on the intensity probe to minimize any error
due to air flow.
The measurements were carried out inside a concrete
chamber of approximately 70 m3 with irregular shape; none
of the surfaces of the chamber was parallel to another. The
interior of the chamber was also well isolated from external
noise.
In measurement of sound intensity with two closely
spaced pressure microphones, phase mismatch between the
two measurement channels is the most serious source of
error.10 Since reverberation increases the error due to phase
mismatch, a significant part of walls of the chamber was
covered with acoustic absorbent material.
The error due to phase mismatch can be estimated com-
bining the pressure-intensity index of the actual measure-
ment and the pressure-residual intensity index of the
measurement system.11,12 The pressure-intensity index is
FIG. 1. (Color online) Diagram of a part of the studied top plate indicating
by means of arrows the four points where a finished section of a fingerboard
was glued for the second configuration of the top plate.
1154 J. Acoust. Soc. Am., Vol. 129, No. 3, March 2011
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 132.174.255.116 On: Fri, 28 Nov 2014 09:35:45
defined as 10 log10(P2=|I|), where P2 is the mean square pres-
sure and I is the sound intensity. If the intensity probe is
placed and oriented in a sound field where both microphones
are exposed to the same sound pressure (thus the intensity is
zero), the corresponding pressure-intensity index is known
as the pressure-residual intensity index. The pressure-inten-
sity index was determined for each measurement with the
B&K 2133 signal analyzer. By using a sound intensity cali-
brator of the type B&K 3541, the pressure-residual intensity
index of the measurement system was obtained before
the first series of measurements, and it was checked
occasionally.
The system formed by the top plate and the guitar-
shaped wooden structure in which it was mounted were dis-
posed vertically as shown in Fig. 2. By using an electrome-
chanical shaker, the sound plate was excited at different
positions on the saddle. The vibration of the shaker was per-
pendicular to the surface of the top plate, and pink noise gen-
erated by the signal analyzer was used as the driving signal.
To provide an identical level of excitation for all the differ-
ent measurements of the sound intensity, the gain in the am-
plifier connected to the shaker was set constant.
A cubic measurement surface with a side length of 1 m,
defined by a frame constructed of 12 1-m long polyvinyl
chloride (PVC) tubes, was used (see Fig. 2). The sound in-
tensity was measured following the scanning method,10,12
which involves moving the sound intensity probe continu-
ously along a path that covers the measurement surface uni-
formly. The cubic frame was placed on the rigid floor of the
chamber. Thus the integral of the normal component of the
sound intensity over the measurement surface, which gives
the sound power, was estimated by scanning the other five
faces of the frame. Five parallel lines were used for the scan-
ning on each face.
The same experimental setup and procedure were used
to determine the sound power radiated by the commercial
guitar.
FIG. 2. (Color online) Picture of the top plate of a guitar mounted on the
rigid support structure placed in vertical position and the frame defining the
cubic surface for the measurements of the radiated sound intensity.
H. Garcı́a-Mayén and A. Santillan: Letters to the Editor

III. RESULTS AND ANALYSIS
Preliminary measurements of the sound intensity radi-
ated by the top plate were carried out for different positions
of the tip of the shaker along the saddle. Since the results of
the determined sound power varied only slightly for different
positions of the driving point, it was decided to excite the
top plate at the center of the saddle for simplicity.
On every measurement of the sound intensity reported
below, the pressure-residual intensity index was more than
10 dB higher than the pressure-intensity index of the actual
measurements. Therefore, the error due to phase mismatch
can be expected to be less than 0.5 dB.11,12
A. Top plate with fixed boundary conditions
Six different measurements of the sound intensity were
made for each of the three configurations of the top plate in
different days. The average values of the sound power and
the corresponding standard deviations are shown in Fig. 3.
For each of the different measurements, the shaker and the
guitar-shaped wooden structure were removed and reas-
sembled to reduce bias errors. The sound power radiated at
frequencies below the one-third octave band of 100 Hz are
not included in Fig. 3, because it was practically indistin-
guishable from the floor noise of the measured system.
It can be observed in Fig. 3 that the sound power radi-
ated at each frequency band was very similar among the
J. Acoust. Soc. Am., Vol. 129, No. 3, March 2011
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 132.174.255.116 On: Fri, 28 Nov 2014 09:35:45
FIG. 3. (Color online) Comparison of the radi-
ated sound power in one-third octave bands for
the three studied configurations of the top plate.
three studied configurations of the top plate. Moreover, the
values of the standard deviation were significantly small.
The radiated acoustic power with the fingerboard coupled to
the top plate by four points was slightly higher only in the
bands with central frequency of 100 Hz and 1 kHz in com-
parison with the fingerboard coupled in the traditional way.
The very small differences in the radiated acoustic power
between these two ways of gluing could be related to the im-
pedance of the thin layer of air trapped between the top plate
and the fingerboard. However, the restriction of motion
imposed on the top plate by the trapped layer of air seemed
to have a small effect on the radiated acoustic power since
no relevant differences were observed in the sound power
radiated by the top plate without the fingerboard.
The highest difference in the radiated acoustic power
between the top plate without the fingerboard and the config-
uration of the fingerboard glued to the top plate in the tradi-
tional way appeared in the band with central frequency of
800 Hz. The difference there was slightly less than 3 dB.
The sum of the contributions in the one-third octave
bands of Fig. 3 for each configuration gives the following
results: 11 lW without the fingerboard, 9 lW with the fin-
gerboard glued by four points, and 11 lW with the finger-
board traditionally glued. Thus, the presence of the
fingerboard and the way it was coupled to the top plate had
practically no effect on the total radiated power in the stud-
ied frequency range.
FIG. 4. (Color online) Sound power radiated in
one-third octave bands by a complete commer-
cial guitar compared with the sound power
radiated after the section of the fingerboard on
its top plate was removed.
H. Garcı́a-Mayén and A. Santillan: Letters to the Editor 1155
 B. Commercial guitar
The results for the determined sound power radiated by
the studied commercial guitar are shown in Fig. 4. The effect
of the air cavity on the radiated spectrum can be observed:
the power output at low frequencies is significantly larger
compared with the results in Fig. 3; there is a peak at the
band of 100 Hz, and the sound power in the band of 80 Hz
was clearly distinguishable. The standard deviations were
again very small.
It can be seen in Fig. 4 that the removal of the section of
the fingerboard had a very small effect on the radiated sound
power, which is consistent with the results obtained for the
top plate attached to the rigid frame even though the radiated
spectrum of the top plate is quite different from the spectrum
of the commercial guitar. In the bands with central frequen-
cies of 125 and 160 Hz, the acoustic power was slightly
higher for the case of the guitar with complete fingerboard;
however, the radiated power increased in the rest of the meas-
ured bands after the removal of the section of the fingerboard.
The largest difference was 4 dB at the band of 200 Hz.
The total sound power in the studied frequency range
radiated by the guitar before and after the removal of the sec-
tion of the fingerboard was 7.1 and 8.9 lW, respectively.
The guitar radiated slightly more acoustic power after the
section of the fingerboard was removed, but the difference
was also practically negligible.
IV. CONCLUSIONS
The results presented in the previous sections provide
evidence that (i) the way in which the fingerboard is coupled
to the top plate of a classical guitar is barely related to the
total emitted acoustic power up to 1122 Hz, and (ii) the cou-
pling between those elements of the guitar is not significant
in the radiated power spectrum measured in one-third octave
bands up to the band with central frequency of 1 kHz.
The effect of the coupling between the top plate and the
fingerboard on the acoustic power radiated by the guitar
1156 J. Acoust. Soc. Am., Vol. 129, No. 3, March 2011
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 132.174.255.116 On: Fri, 28 Nov 2014 09:35:45
could be relevant at higher frequencies and needs to be
studied.
Subjective evaluations should be conducted to investi-
gate if the measured small variations in the radiated power
spectrum produced by different ways of coupling between
the top plate and the fingerboard can be perceived, and in
that case, the effects produced in the listener.
ACKNOWLEDGMENTS
The authors are very grateful to Ricardo Rubio, an
expert craftsman in guitar construction, for his help and his
careful work. We also like to thank the associate editor and
referees for their constructive comments and suggestions.
1
N. H. Fletcher, and T. D. Rossing, The Physics of Musical Instruments
(Springer-Verlag, New York, 1991), Chap. 9, pp. 208–234.
2
A. Ezcurra, M. J. Elejabarrieta, and C. Santamaria, “Fluid-structure cou-
pling in the guitar box, numerical and experimental comparative study,”
Appl. Acoust. 66, 411–425 (2005).
3
M. J. Elejabarrieta, A. Ezcurra, and C. Santamaria, “Coupled modes of
the resonance box of the guitar,” J. Acoust. Soc. Am. 111, 2283–2292
(2002).
4
M. J. Elejabarrieta, C. Santamaria, and A. Ezcurra, “Air cavity modes in
the resonance box of the guitar: The effect of the sound hole,” J. Sound
Vib. 252, 584–590 (2002).
5
J. C. S. Lai and M. A. Burgess, “Radiation efficiency of acoustics guitars,”
J. Acoust. Soc. Am. 88, 1222–1227 (1990).
6
A. Le Pichon, S. Berge, and A. Chaigne, “Comparison between experi-
mental and predicted radiation of a guitar,” Acustica 84, 136–145 (1998).
7
A. Chaigne and M. Rosen, “Analysis of guitar tones for various structural
configurations of the instrument,” Catgut. Acoust. Soc. J. 3, 24–31 (1999).
8
O. Christensen, “An oscillator model for analysis of guitar sound pressure
response,” Acustica 54, 289–295 (1984).
9
T. J. W. Hill, B. E. Richardson, and S. J. Richardson, “Acoustical parame-
ters for the characterisation of the classical guitar,” Acta Acust. 90, 335–
348 (2004).
10
F. Jacobsen, “Sound intensity,” in Handbook of Acoustics, edited by
T. D. Rossing (Springer, NewYork, 2007), Chap. 25, pp. 1053–1075.
11
F. J. Fahy, Sound Intensity, 2nd ed. (E & FN Spon, London, 1995), Chap.
7, pp. 143–152.
12
Acoustics—Determination of sound power levels of noise sources using
sound intensity – Part 2: Measurement by scanning, International Standard
ISO 9614-2: 1996 (International Organization for Standardization,
Geneva, Switzerland, 1996), pp. 1–18.
H. Garcı́a-Mayén and A. Santillan: Letters to the Editor