Guitar Pickup: Measurement of its Two-Dimensional

Nonlinear Function
Antonin Novak, Bertrand Lihoreau, Emmanuel Brasseur, Pierrick Lotton,
Laurent Simon

To cite this version:

Antonin Novak, Bertrand Lihoreau, Emmanuel Brasseur, Pierrick Lotton, Laurent Simon. Guitar
Pickup: Measurement of its Two-Dimensional Nonlinear Function. Journal of the Audio Engineering
Society, 2020, 68 (3), pp.184-192. �10.17743/jaes.2020.0002�. �hal-02512148�

HAL Id: hal-02512148

https://hal.science/hal-02512148
Submitted on 19 Mar 2020

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
 Guitar Pickup: Measurement of its
Two-dimensional Nonlinear Function
Antonin Novak∗ , Bertrand Lihoreau, Emmanuel Brasseur,
Pierrick Lotton, Laurent Simon
Laboratoire d’Acoustique de l’Université du Mans, LAUM - UMR 6613 CNRS, Le Mans Université, Avenue Olivier
Messiaen, 72085 LE MANS CEDEX 9, France
∗
[email protected], ∗ https://ant-novak.com

Abstract
The electromagnetic pickup of an electric guitar is a nonlinear device that provides a pleasant distortion. Although
the pickup is a simple device consisting of a coil and a few magnets or pole pieces, the measurement of its nonlinear
function is a difficult task. This paper shows a measurement technique that can estimate the nonlinear function of a
pickup in both y− and z− directions of the vibrating string. The experimental results are provided for three different
types of pickups: a single-coil pickup with six staggered pole-pieces, a humbucker pickup with six equal height pole-
pieces, and a humbucker rail pickup. The measured non-linear functions of the three pick-ups are very different from
LAUM, CNRS UMR 6613

each other, leading to different distortions. These experimental results confirm that the pickup geometry plays an
important role in distortion.

The archived file is not the final published version of the article *********
The definitive publisher-authenticated version is available online at *****,
Readers must contact the publisher for reprint or permission to use the material in any form.

1
 Novak et al. Guitar Pickup Measurement — 2/13

1. INTRODUCTION
An electric guitar is one of many musical instruments whose beautiful sound is born with the help of a non-linear
mechanism. The heart of an electric guitar is a pickup, a non-linear sensor that picks up the mechanical vibrations
of the strings and converts them into an electrical signal [1, 2, 3, 4, 5, 6]. The most common type of pickup is the
magnetic one, which houses a coil wrapped around a core composed of either a row of six polar pieces (see Figure
1(a)) or a bar magnet (called rail) (see Figure 1(b)). A ferromagnetic string that vibrates near the pickup causes a
variation in the magnetic flux through the coil and, according to Faraday’s law, an electrical voltage is induced at
the coil’s terminals.
There are many pickups available on the market, each of them providing different outputs, some sound warmer,
some cleaner and others more distorted. All the electric-guitar players have probably wondered what distinguishes
one particular pickup from another. The answer to this puzzling question is essential not only for guitar players
but also for pickup manufactures and for digital audio effects engineers, especially those working with instrument
synthesis [7, 8, 9]. A few mathematical models of pickup available in the literature may help to find the answer to
this tricky question.
Horton and Moore provide a model of the magnetic field of pickups with pole pieces [3]. Guadagnin et al.
present an analytical model of the pickup and its influence on pickup distortion [10]. The work of Jungmann
[1] provides a very detailed overview of how pickups work and how their output signals may be affected by their
physical properties. Lastly, Paiva et al. provide a block-oriented model combining linear and nonlinear blocks to
describe the pickup behavior [11]. Based on these works one can conclude that the sound of a pickup is influenced by
LAUM, CNRS UMR 6613

three main factors: a) the pickup position which is closely related to the string vibration, b) the pickup impedance
together with the input impedance of the device to which the guitar is plugged which form a linear filter, and c) a
nonlinear behavior of the pickup. Furthermore, the main core of Paiva’s model [11] describing the magneto-electric
conversion is based on a static nonlinearity representing the nonlinear relation between the string displacement and
the magnetic flux, followed by a time differentiator.
Many of these studies focus on theoretical modeling and simulations. The work presented in this paper uses an
experimental approach and tries to answer the following questions from the experimental results. What distinguishes
one pickup from other from the nonlinear distortion point of view? Can the distortion contribution due to the
horizontal excitation be neglected? Does the distortion depend on the string rest position?
The work presented in this paper is a continuity of the recently published work [12], where a dedicated experi-
mental bench that provides a purely sinusoidal string excitation was used to show the harmonic distortion of several
pickups. In [12], the pickups are excited in z-direction (Fig. 1) and the static nonlinear function of the pickup is
measured using the dedicated bench. In this work, the same experimental bench as the one used in [12] is used to

y 6 permanent magnets permanent bar magnet

or pole pieces or pole piece
x
coil coil

base plate base plate

a) pickup with six pole pieces b) rail pickup

Figure 1. Schematic representation of (a) a pole-piece pickup and (b) a rail pickup.
 Novak et al. Guitar Pickup Measurement — 3/13

y(t)
static φ(t) d u(t)
NLfnc dt
z(t)

Figure 2. Block-oriented representation of the pickup nonlinear model with both horizontal y(t) and vertical z(t)
string displacement input and the voltage output u(t).

measure the static nonlinear function of three different pickups in both vertical (z-) and horizontal (y-) directions.
As the nonlinear behavior of the pickup is influenced by the vibration of the string in both directions (y and z),
Paiva’s static nonlinear model is extended to two inputs (Fig. 2). The experimental results verify some of earlier
theoretical results for both vertical [1, 4, 3] and horizontal [4, 3] directions.

2. EXPERIMENTAL SETUP
2.1 Pickups under test
Three types of pickups of the brand Seymour Duncan are used in this work to measure their differences from the
nonlinear point of view: (a) SSL-5 - a single-coil pickup, (b) SH-2N - a humbucker (double-coil) pickup, and (c)
STHR-1B - a humbucker rail pickup (Fig. 3).

2.2 Experimental bench

LAUM, CNRS UMR 6613

A guitar pickup is one of the non-linear systems that are not easy to measure. E.g., a precise measurement of the
harmonic distortion at the output of the pickup requires a pure sine motion of the string as the input signal. Since
the guitar’s string displacement is guided by the laws of vibrations, the input signal of the pickup is difficult to
control.
To impose the string displacement we use a dedicated measurement bench [12, 13] whose schematic representation
is depicted in Fig. 4 for vertical (z-axis) motion of the string. A piece of string (8 cm long and 1 mm in diameter)
is glued to a composite plate (3 x 8 cm) which is rigidly connected to an electrodynamic shaker (LDS V406). The
shaker, driven by a Devialet D-premier amplifier, RME Fireface 400 sound card, and a personal computer, is used
as a source of the string displacement. The pickup is then placed to the proximity of the string so that the string
rest position is at coordinates (y0 , z0 ). An electromagnetic shielding cage is placed around the shaker to avoid
a possible disturbance by the electromagnetic field of the shaker. An accelerometer PCB 352C22 is fixed to the
composite plate and is connected to a PCB sensor signal conditioner 482C series. The sound card RME Fireface
400 is then used to acquire both the accelerometer signal a(t) and the output voltage u(t) from the pickup (directly
connected to the sound card instrument input with an impedance of 470 kΩ).
To ensure that the string is vibrating in a pure sinusoidal way with no higher harmonic components due to
the nonlinear behavior of the shaker we use a recently developed adaptive technique of harmonic control that can

z z z

z=0 z=0 z=0

E A D G B e E A D G B e E A D G B e
(6th) (5th) (4th) (3rd) (2nd) (1st) (6th) (5th) (4th) (3rd) (2nd) (1st) (6th) (5th) (4th) (3rd) (2nd) (1st)

SSL-5 SH-2N STHR Hot Rails

y y y
y=0 y=0 y=0

a) SSL-5 pickup b) SH-2N pickup c) STHR pickup

Figure 3. Profiles of three pickups under test.

 Novak et al. Guitar Pickup Measurement — 4/13

y
electromagnetic shielding

z accelerometer shaker
x
pushing
rod
pickup

piece of string measured

acceleration
signal generation with
harmonic control

a) experimental setup for vertical excitation

pole piece of pickup

string

0 z0 z
LAUM, CNRS UMR 6613

b) string position (vertical excitation)

Figure 4. Schematic representation of the measurement for the vertical (z-direction) string motion.

pre-distort the input signal of the shaker to create a perfect periodical signal at the output of the shaker with
spectral purity up to 100 dB [14]. Using this technique, we can generate a pure harmonic displacement of the string
even for large amplitudes, completely getting rid of the higher harmonics created by the shaker.
To enable the measurement not only in vertical z-direction but also in horizontal y-direction the measurement
setup is modified as depicted in Fig. 5(a). The pickup is placed horizontally, and the y-axis is oriented in accordance
with Fig. 5(b). The distance between the string at rest and the pole piece is still denoted z0 .

2.3 Harmonic distortion

An example of harmonic distortion of each pickup is shown in Fig. 6 for both vertical (along z-axis) and horizontal
(along y-axis) string displacements. The rest position of the string is set at the distance of z0 = 3 mm in front of
the fifth string position (y0 = 9.5 mm). The string is then displaced with a pure sinusoidal excitation of frequency
110 Hz (corresponding to open A-string) and amplitude of ±1.5 mm in (a) vertical excitation and (b) horizontal
excitation. The spectra of the voltage signal at the output of each pickup are depicted in Fig. 6.
One can note that the vertical excitation leads to a voltage signal about 20 dB higher than the horizontal
excitation. On the one hand, the vertical excitation leads to a generation of similar spectra for all three tested
pickups; the STHR-1B rail pickup providing slightly stronger distortion compared to the other two pickups. On the
other hand, the horizontal excitation shows completely different spectra for all three pickups. The SSL-5 pickup
excited horizontally generates the first and the second harmonic at the level of -28 dB and -35 dB respectively. The
dominant harmonic of the SH-2N pickup is the second one (-29 dB); the first harmonic is attenuated to -53 dB.
The STHR-1B rail pickup has the first harmonic the strongest (-29 dB), the other harmonic being much smaller
(-50 dB for the second harmonic).
 Novak et al. Guitar Pickup Measurement — 5/13

z electromagnetic shielding

y accelerometer shaker
x
pushing
rod
pickup

piece of string measured

acceleration
signal generation with
harmonic control

a) experimental setup for horizontal excitation

pole piece of pickup

string

y y0 0
LAUM, CNRS UMR 6613

b) string position (horizontal excitation)

Figure 5. Schematic representation of the measurement for the horizontal (in y-direction) string motion.

The following experimental study is focused on the measurement of the static nonlinear function NLfnc and its
influence on the nonlinear distortion of each pickup under test.

3. STATIC NONLINEAR FUNCTION

The block-oriented model of the pickup proposed by Paiva [11] (static nonlinearity NLfnc followed by the time
derivative) has been experimentally verified in several works using the vertical excitation [15, 16, 12]. In [12] it has
been shown that the static nonlinear function NLfnc can be obtained experimentally using the following procedure.
First, we can deduce the variable φ(t) from Paiva’s model using Faraday’s law of induction

dφ(t)
u(t) = − , (1)
dt
that defines the voltage u(t) generated at the output of a coil. Next, we obtain φ(t) as an integral of the measured
voltage u(t)
Z t
φ(t) = − u(t0 )dt0 + C, (2)
−∞

where an unknown constant of integration C is inherent in the construction of anti-derivatives. Finally, the static
nonlinear function NLfnc is plotted directly from measured displacement z(t) (deduced from the acceleration) and
the time integral of the voltage as plot(z, phi) in Matlab language. This procedure is graphically represented
in Fig. 7 with measured data provided from 60 Hz1 excitation with a sinusoidal z−axis displacement of ±3 mm
around y0 = 9.5 mm and z0 = 3 mm.
1 Note that 60 Hz is the resonant frequency of the shaker at which it can reach high displacement (±3 mm).
 Novak et al. Guitar Pickup Measurement — 6/13

Seymour Duncan Seymour Duncan Seymour Duncan

SSL-5 SH-2N STHR-1B Hot Rails

VERTICAL EXCITATION (z-axis)

0 0 0
Voltage [dBV]

Voltage [dBV]

Voltage [dBV]
-20 -20 -20
-40 -40 -40
-60 -60 -60
-80 -80 -80
0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000
Frequency [Hz] Frequency [Hz] Frequency [Hz]

HORIZONTAL EXCITATION (y-axis)

0 0 0
Voltage [dBV]

Voltage [dBV]

Voltage [dBV]
-20 -20 -20
-40 -40 -40
-60 -60 -60
-80 -80 -80
0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000
LAUM, CNRS UMR 6613

Frequency [Hz] Frequency [Hz] Frequency [Hz]

Figure 6. Spectra of the output voltage of three pickups under test obtained with the excitation frequency of 110
Hz: a single-coil pickup with six staggered pole-pieces on the left, a humbucker pickup with six equal height
pole-pieces in the middle, and a humbucker rail pickup on the right. Results depicted for both vertical and
horizontal string motions. The string is placed at the distance of z0 = 3 mm in front of the fifth pole piece
position (y0 = 9.5 mm).

The results from [12] show that the whole nonlinear function NLfnc can be built from several measurements for
different rest positions of the string z0 by adjusting the unknown constant C for each measurement as depicted in
Fig. 8. It is also shown in [12] that the results are independent of the excitation frequency, which is in accordance
with Paiva’s hypothesis of static nonlinear function (static meaning independent of frequency).
The procedure used for the measurement with the vertical (z− axis) excitation is next used for horizontal (y−
axis) excitation. The pickup is placed as depicted in Fig. 5. The string is placed at its rest position (y0 , z0 ),
displaced horizontally along the y−axis with the shaker, and the static nonlinear function NLfnc is estimated using
the same procedure as for the vertical excitation. An example with a 60 Hz excitation with a sinusoidal y−axis
displacement of ±3 mm around y0 = 9.5 mm and z0 = 3 mm is shown in Fig. 9 for the SH-2N pickup. The same
measurement is next made for several values of y0 , as shown in Fig. 10.
The static nonlinear function NLfnc in the horizontal direction (z0 = 3 mm) for the SH-2N pickup depicted
in Fig. 10 shows an interesting pattern. The nonlinear function NLfnc has two local maxima at y0 = 0 mm and
y0 = 9.5 mm corresponding to the 6th and 5th pole pieces of the pickup. The NLfnc also has two local minima
in-between the positions of pole pieces. Finally, the NLfnc drops for negative values of y (outside the pickup). The
relation between the shape of the NLfnc and the nonlinear distortion is detailed in section 5.
 Novak et al. Guitar Pickup Measurement — 7/13

y(t) = y0
static φ(t) d u(t)
NLfnc dt
z(t)

a) block diagram of the pickup model

6 4

time integral of voltage [mV s]

4 2
z [mm]

2 0

0 -2

b) displacement in z-axis c) time integral of

(obtained from measured acceleration) measured voltage

4
time integral of voltage [mV s]

3 I/O curve of static NLfnc

plot(z,phi)
2

1
LAUM, CNRS UMR 6613

-1

-2
0 2 4 6
z [mm]

d) time integral of voltage vs. displacement

Figure 7. (a) A block diagram of the pickup nonlinear model, (b) string displacement in vertical direction (obtained
from the measured acceleration), (c) time integral of the measured voltage, and (d) a plot of time integral of
voltage vs. string displacement in an I/O graph. Measurements performed on a SH-2N pickup with a string rest
position (y0 = 9.5, z0 = 3) mm and 60 Hz harmonic string oscillation in z-direction with amplitude of ±3 mm.

4. TWO-DIMENSIONAL NONLINEAR FUNCTION

Both vertical (along z−axis) and horizontal (along y−axis) measurements for different (y0 , z0 ) can be joined together
to form a two-dimensional function NLfnc (y, z) whose value depends on the instantaneous position of the string.
Fig. 11 shows the function NLfnc (y, z) of the SH-2N pickup. The function is represented as a two-dimensional graph
composed of 44 measured curves (each plotted in a different color). The vertical measurements are provided at a
grid of 6x4 positions (y0 , z0 ) with y0 from -4.75 to 19 mm with a step of 4.75 mm, and z0 ∈ (3, 5, 7, and 10 mm).
The horizontal measurements are provided at a grid of 2x10 points (y0 , z0 ) with y0 from -8.5 to 18.5 mm with a
step of 3 mm and z0 at 3 and 5 mm.
The flux φ(t) is given as


φ(t) = NLfnc y(t), z(t) , (3)
whose time derivative is equal to the phase inverted (minus sign) output voltage of the pickup (see Eq. (1))


dNLfnc y(t), z(t)
u(t) = − . (4)
dt
 Novak et al. Guitar Pickup Measurement — 8/13

Figure 8. I/O graphs (time integral of measured voltage vs. string displacement in vertical direction) for four
different pickup/string distances z0 . All the four I/O graphs are superposed and shifted by the unknown constant
of integration (Eq. (2)). Measurements performed on a SH-2N pickup with a string rest position (y0 = 9.5 mm,z0 )
and 60 Hz harmonic string oscillation in z-direction with amplitude of ±3 mm.

Applying the chain rule for the derivative of functions of two variables we can express the output voltage as
LAUM, CNRS UMR 6613

∂NLfnc dy(t) ∂NLfnc dz(t)

u(t) = − − . (5)
∂y dt ∂z dt

Denoting the velocity vector ~v (t)  

dy(t) dz(t)
~v (t) = , , (6)
dt dt
and the gradient  
~ ∂NLfnc ∂NLfnc
∇NL fnc = , , (7)
∂y ∂z
we can express the voltage at the output of the pickup as the following dot product

~
u(t) = −∇NLfnc · ~
v (t). (8)

Consequently, the distortion of each pickup thus depends on the local gradient of the nonlinear function NLfnc .

5. DISCUSSION
Eq. (8) shows that the pickup output voltage is proportional both to the velocity of the string and to the local
gradient of the static nonlinear function NLfnc . It is thus straightforward to guess from the shape of the function
NLfnc how the string position (y0 , z0 ) influences the signal level and the nonlinear distortion of the output voltage.
Fig. 12 shows the results of several vertical (along z−axis) and horizontal (along y−axis) measurements. The
vertical (along z−axis) results show that the NLfnc decreases with the distance z for all measured pickups but that
the gradient (slope) of each function varies with distance z. Based on the previous analysis we can conclude that
for string positions close to the pickup (e.g., z0 = 3 mm), in the case of the SSL-5 pickup the NLfnc shows lower
gradient variation compared to the STHR-1B pickup. This is correlated with the distortion results presented in
Fig. 6 where the SSL-5 pickup exhibits lower distortion than the STHR-1B pickup.
The horizontal (along y−axis) results presented in Fig. 12 are shown for two different string distances: z0 = 3 mm
and z0 = 5 mm. Let us first discuss the results for string distances of z0 = 3 mm. There is an obvious difference
 Novak et al. Guitar Pickup Measurement — 9/13

y(t)
static φ(t) d u(t)
NLfnc dt
z(t) = z0

a) block diagram of the pickup model

0.2

time integral of voltage [mV s]

12
0.1

y [mm]
10
0

8 -0.1

6 -0.2

b) displacement in y-axis c) time integral of

(obtained from measured acceleration) measured voltage

0.2
time integral of voltage [mV s]

0.1

0
LAUM, CNRS UMR 6613

I/O curve
of static NLfnc
-0.1
plot(y,phi)

-0.2
6 8 10 12
y [mm]
d) time integral of voltage vs. displacement

Figure 9. (a) A block diagram of the pickup nonlinear model, (b) string displacement in horizontal direction
(obtained from the measured acceleration), (c) time integral of the measured voltage, and (d) a plot of time
integral of voltage vs. string displacement in an I/O graph. Measurements performed on a SH-2N pickup with a
string rest position (y0 = 9.5, z0 = 3) mm and 60 Hz harmonic string oscillation y-direction with amplitude of
±3 mm.

in behavior between the rail pickup (STHR-1B, Fig. 12 right column) and the pickups with pole pieces (SSL-5 and
SH-2N, Fig. 12 left and middle column respectively). While there is an obvious influence of pole pieces in the case
of SSL-5 and SH-2N resulting in local maxima and minima along the y−axis (Fig. 12 left and middle column),
the NLfnc of the STHR-1B pickup is smooth with no oscillations (Fig. 12, right column). The local maxima and
minima of the NLfnc of the pole-piece pickups (SSL-5 and SH-2N) are located at positions y0 = 0, y0 = 9.5 mm, and
y0 = 19 mm corresponding to positions of 6th, 5th, and 4th pole piece respectively. Measurements in the horizontal
direction at different string positions z0 = 3 mm and z0 = 5 mm (Fig. 12) show that for higher z0 (string positions
further away from pickup) the variations of function NLfnc are much smaller.
The shape of the function NLfnc in the horizontal direction (Fig. 12) also has an influence on the nonlinear
distortion shown in Fig. 6. In Fig. 6, the harmonic distortion is evaluated for the position (y0 = 9.5 mm, z0 = 3 mm).
Fig. 12 shows that at this position for horizontal excitation there is a local maximum in the case of SSL-5 and SH-2N.
In the close proximity of this maximum the function NLfnc behaves rather symmetrically. The symmetry is almost
 Novak et al. Guitar Pickup Measurement — 10/13

time integral of voltage [mV s]

2.5

1.5 y = -8.5 mm
0
y = -2.5 mm
0
1 y = 3.5 mm
0
y = 9.5 mm
0.5 0
y = 15.5 mm
0
0
-10 -5 0 5 10 15
y [mm]

Figure 10. I/O graphs (time integral of measured voltage vs. string displacement in horizontal direction) for five
different pickup/string distances y0 . All the five I/O graphs are superposed and shifted by the unknown constant
of integration (Eq. (2)). Measurements performed on a SH-2N pickup with a string rest position (y0 ,z0 = 3 mm)
and 60 Hz harmonic string oscillation in y-direction with amplitude of almost ±3 mm.

perfect for the SH-2N pickup which is well correlated with the dominating 2nd harmonic (see Fig. 6). For the
SSL-5 pickup the symmetry is not perfect resulting in a presence of both 1st and 2nd harmonic components (see
LAUM, CNRS UMR 6613

Fig. 6). Finally, the STHR-1B pickup shows a smooth nonlinear function NLfnc (Fig. 12) which is due to the rail
magnet (single bar pole piece). Consequently, the distortion of the STHR-1B pickup in the horizontal direction
generates mainly the 1st harmonic (see Fig. 6).
The shape of the function NLfnc in horizontal direction (Fig. 12) also indicates how the y0 position influences
the level and the distortion of the pickup voltage signal. In the case of the STHR-1B rail pickup (Fig. 12, right
column, horizontal excitation), the function is almost flat around 4th string position (y0 = 19 mm). Its gradient is
almost zero leading to almost no voltage output for horizontal excitation. On the other hand, the gradient around
6th position (y0 = 0) is higher, leading to a higher voltage output.
Similar behavior can be observed in the case of SSL-5 and SH-2N pickups (Fig. 12, left and middle column for
horizontal excitation): the voltage level and distortion change with the position y0 . We can also observe a different
behavior in the horizontal direction between SSL-5 and SH-2N pickups. Indeed, the SH-2N pickup pole pieces
have all the same height while the SSL-5 pickup have the pole pieces staggered (the height of each pole piece in
z−direction is different), see Fig. 3. The staggered pickup thus breaks the symmetry of the function NLfnc (Fig. 12,
left column, horizontal excitation).
Although horizontal excitation generates lower output voltage than vertical excitation, it seems important to
take this into account for modeling higher harmonics. Distortion measurement of SH-2N pickup (Fig. 6) shows that
horizontal excitation can generate a second harmonic whose amplitude is higher than the third harmonic caused by
vertical excitation and which is 25 dB below the first harmonic. A more detailed psycho-acoustic study is needed
to draw conclusions about the audibility of these non-linear effects of guitar pick-ups.

6. CONCLUSION
Why do some guitar pickups distort more than others? This paper provides a part of the answer to this puzzling
question by showing how to measure the static nonlinear function of guitar pickup as a function of the vertical and
horizontal displacements of the string. The estimation of the non-linear function can be very useful for the design
of pickups.
 Novak et al. Guitar Pickup Measurement — 11/13

ΦNL (y,z) [mV s]

z [mm]
y [mm]

Figure 11. Nonlinear function of the SH-2N pickup obtained as a superposition of all the measured I/O curves and
depicted in a single three-dimensional graph.

Three different pickups, a single-coil pickup with six staggered pole-pieces, a humbucker pickup with six equal
height pole-pieces, and a humbucker rail pickup, are measured. It is shown that the gradient of the nonlinear static
LAUM, CNRS UMR 6613

function is directly related to pickup distortion and that vertical and horizontal string vibrations contribute to
pickup distortion.
The results show that all three studied pickups have a similar shape of the nonlinear function for the vertical
displacement of the string. However, the gradient of this function and its amplitude, both responsible for output
level and distortion, differ from one pickup to another. The results also show that the nonlinear function for
horizontal string displacement is highly dependent on the geometry of the polar pieces of the pickup. For the rail
pickup (single pole pieces), the nonlinear function is flat and smooth, resulting in very low distortion due to the
horizontal displacement of the strings. For pickups with six pole pieces, the nonlinear function consists of small
local minima and maxima, which leads to a much higher nonlinear distortion for horizontal excitation. While the
horizontal excitation provides smaller output voltage than the vertical one, its nonlinear contribution to the pickup
distortion can not be neglected, especially for pickups with six pole pieces.

7. ACKNOWLEDGMENT
The measurements, discussions, and redaction of this paper have been conducted mainly in a free time of all the
authors, motivated by their passion for guitars and nonlinear systems. We would very much like to thank our wives
and families for their understanding.
 Novak et al. Guitar Pickup Measurement — 12/13

Seymour Duncan Seymour Duncan Seymour Duncan

SSL-5 SH-2N STHR-1B Hot Rails

VERTICAL EXCITATION (z-axis)

y0 = 9.5 mm y0 = 9.5 mm y0 = 9.5 mm

LAUM, CNRS UMR 6613

HORIZONTAL EXCITATION (y-axis)

z0 = 3 mm z0 = 3 mm z0 = 3 mm

z0 = 5 mm z0 = 5 mm z0 = 5 mm

Figure 12. Comparison of the measured nonlinear functions for the three tested pickups depicted for one vertical
excitation around y0 = 9.5 mm and two horizontal excitations with z0 = 3 mm and y0 = 5 mm.
 Novak et al. Guitar Pickup Measurement — 13/13

References
[1] T. Jungmann, Theoretical and practical studies on the behavior of electric guitar pick-ups, Master’s thesis,
Helsinki Univ. of Tech., Espoo, Finland (1994).

[2] D. Hunter, The Guitar Pickup Handbook: The Start of Your Sound (Hal Léonard Corporation) (2008).

[3] N. G. Horton, T. R. Moore, “Modeling the magnetic pickup of an electric guitar,” American Journal of Physics,
vol. 77, no. 2, pp. 144–150 (2009).

[4] M. Zollner, Physik der Elektrogitarre (im Selbstverlag, [email protected]) (2014).

[5] M. Mustonen, D. Kartofelev, A. Stulov, V. Välimäki, “Experimental verification of pickup nonlinearity,”

presented at the Proceedings International Symposium on Musical Acoustics (ISMA 2014), Le Mans, France,
vol. 1 (2014).

[6] A. Falaize, T. Hélie, “Guaranteed-passive simulation of an electro-mechanical piano: A port-Hamiltonian

approach,” presented at the Proc. of the 18 th Int. Conference on Digital Audio Effects (DAFx-15) (2015).

[7] V. Välimäki, J. Huopaniemi, M. Karjalainen, Z. Jánosy, “Physical Modeling of Plucked String Instruments
with Application to Real-Time Sound Synthesis,” J. Audio Eng. Soc, vol. 44, no. 5, pp. 331–353 (1996).

[8] M. Karjalainen, H. Penttinen, V. Välimäki, “Acoustic sound from the electric guitar using DSP techniques,”
LAUM, CNRS UMR 6613

presented at the Acoustics, Speech, and Signal Processing, 2000. ICASSP’00. Proceedings. 2000 IEEE Inter-
national Conference on, vol. 2, pp. II773–II776 (2000).

[9] M. Karjalainen, T. Mäki-Patola, A. Kanerva, A. Huovilainen, “Virtual air guitar,” Journal of the Audio
Engineering Society, vol. 54, no. 10, pp. 964–980 (2006).

[10] L. Guadagnin, B. Lihoreau, P. Lotton, E. Brasseur, “Analytical Modeling and Experimental Characterization
of a Magnetic Pickup for Electric Guitar,” Journal of the Audio Engineering Society, vol. 65, no. 9, pp. 711–721
(2017).

[11] R. C. Paiva, J. Pakarinen, V. Välimäki, “Acoustics and modeling of pickups,” Journal of the Audio Engineering
Society, vol. 60, no. 10, pp. 768–782 (2012).

[12] A. Novak, B. Lihoreau, P. Lotton, E. Brasseur, L. Simon, “Experimental Study Of Guitar Pickup Nonlinearity,”
presented at the Proc. of the 18th Int. Conference on Digital Audio Effects (DAFx-18) (2018).

[13] P. Lotton, B. Lihoreau, E. Brasseur, “Experimental Study of a Guitar Pickup,” presented at the Int. Symp.
Music. Acoust., pp. 355–360 (2014).

[14] A. Novak, L. Simon, P. Lotton, “A simple predistortion technique for suppression of nonlinear effects in periodic
signals generated by nonlinear transducers,” Journal of Sound and Vibration, vol. 420, pp. 104–113 (2018).

[15] A. Novak, L. Guadagnin, B. Lihoreau, P. Lotton, E. Brasseur, L. Simon, “Non-linear identification of an

electric guitar pickup,” presented at the Proceedings of the 19th International Conference on Digital Audio
Effects (DAFx-16), Brno, Czech Republic, pp. 5–9 (2016).

[16] A. Novak, L. Guadagnin, B. Lihoreau, P. Lotton, E. Brasseur, L. Simon, “Measurements and Modeling of the
Nonlinear Behavior of a Guitar Pickup at Low Frequencies,” Applied Sciences, vol. 7, no. 1, p. 50 (2017).