Professional Documents
Culture Documents
Development of Parametric Loudspeaker: IEEE Potentials November 2010
Development of Parametric Loudspeaker: IEEE Potentials November 2010
Development of Parametric Loudspeaker: IEEE Potentials November 2010
net/publication/224189070
CITATIONS READS
22 1,830
2 authors:
Some of the authors of this publication are also working on these related projects:
Effect of VR/AR Head Mounted Display (HMD) on Head Related Transfer function (HRTFs) View project
All content following this page was uploaded by Woon-Seng Gan on 28 May 2014.
S
ound is one of the most important
mediums used by people to
deliver information. Regarding
personal privacy in modern society, the
generation of directional sound is emerg-
ing significantly. For example, in the
library, a personal public announcement
system can help to communicate with a
group of listeners without disturbing
others. In a museum or art gallery, an
introduction of exhibitions can only be
heard by those standing in front of the
exhibit. There are also similar needs for
private messaging in vending and dis-
pensing machines, exhibition booths, © INGRAM PUBLISHING
NOVEMBER/DECEMBER 2010 21
Dimensionless Range (σ = 0) Dimensionless Range (σ = 1)
1 1
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
P/P0
P/P0
0 0
–0.2 –0.2
–0.4 –0.4
–0.6 –0.6
–0.8 –0.8
–1 –1
–50 –40 –30 –20 –10 0 10 20 30 40 50 –100 –80 –60 –40 –20 0 20 40 60 80 100
Dimensionless Retarded Time (τ ) Dimensionless Retarded Time (τ )
(a) (b)
Dimensionless Range (σ = 2) ×10–3 Dimensionless Range (σ = 3)
0.04 2
0.03
1
0.02
0
0.01
P/P0
P/P0
0 –1
–0.01
–2
–0.02
–3
–0.03
–0.04 –4
–100 –80 –60 –40 –20 0 20 40 60 80 100 –100 –80 –60 –40 –20 0 20 40 60 80 100
Dimensionless Retarded Time (τ ) Dimensionless Retarded Time (τ )
(c) (d)
×10–4 Dimensionless Range (σ = 4) ×10–5 Dimensionless Range (σ = 5)
4 4
3
2
2
1 0
0
P/P0
P/P0
–2
–1
–2 –4
–3
–6
–4
–5 –8
–100 –80 –60 –40 –20 0 20 40 60 80 100 –100 –80 –60 –40 –20 0 20 40 60 80 100
Dimensionless Retarded Time (τ ) Dimensionless Retarded Time (τ )
(e) (f)
Models of nonlinear acoustics rived from the Westervelt equation. The ' 2p c3 2 ' ' 3p b '2p2
5 =' p 1 3 31
.
Several model equations have been KZK equation assumes the sound to 'z't 2 2c 't 2rc3 't2
explored to describe the propagation of localize in the vicinity of the propagat-
finite-amplitude sound beams produced ing axis. It accurately describes the Here p is the sound pressure; z is the
by the parametric array. The Westervelt propagation of finite-amplitude sound coordinate along the beam propagation
equation (1963) is one of the fundamen- beams combining the effects of diffrac- direction; t 5 t 2 z/c is the retarded
tal equations of the parametric array, tion, absorption, and nonlinearity under time, and c is the small signal sound
and approximates the full second order a parabolic approximation, and ex - speed. Furthermore, p, d and b are the
wave equation. pressed by the first, second, and third density, dissipation factor corresponding
In 1971, the Khokhlov–Zabolotskaya– terms on the right hand side of KZK to thermoviscous absorption, and the
Kuznetsov (KZK) equation was de- equation, respectively: nonlinearity coefficient of the medium,
22 IEEE POTENTIALS
respectively. =2' is the Laplacian opera-
tor that operates on the X-Y plane per- Nonlinear
pendicular to the axis of the beam. Interaction
The numerical solution of the KZK Digital Class D Ultrasonic
Processor Amplifier Transducer
equation can show us the effect of wave-
form distortion (self-demodulation), Audio Audio
which is caused by the nonlinearity of
air. Fig. 4 shows the simulation result of Fig. 5 Block diagram of ABS.
axial waveforms at different dimension-
less ranges, whose original signal is a used to equalize the frequency response control, and our proposed modified
short tone burst with Gaussian envelope. before amplitude modulation. On the amplitude modulation techniques. The
As the wave propagates, the carrier wave right-hand side of the Berktay equation, special-designed class-D amplifier is
is being damped out, and the self- second-order derivation of the square of used to adjust the voltage gain to drive
demodulated waveform is formed simul- the envelop function is involved. To an array of ultrasound emitters, which
taneously. The relatively high absorption eliminate the distortion introduced by are made of lead zirconate titanate or
prevents higher harmonic spreading to these operations, Kite proposed a polyvinylidene fluoride materials. Due
farfield. If dissipation factor increases, square-root amplitude modulation of to the nonlinear interaction of the air,
the primary wave is more rapidly twice integrations of the modulating directional audible sound will be gener-
absorbed, and thus the fully demodu- wave. This preprocessing technique is ated at a targeted sweet spot.
lated waveform is generated closer to commonly known as the square-root The ABS was recently installed in a
the source. gaming booth in the Fusionopolis, a
Although the KZK equation is one of research and development complex
the most efficient models of parametric located at the One-North business park
array, it is too complex to use in prac- in Singapore. We used two ABS speakers
tice. Berktay provided a simple expres- (ultrasound emitters, see Fig. 6) to proj-
sion that can be used to predict the ect directional binaural sound from a
farfield array response on the propagat- gaining console to the gamer (see Fig. 7)
ing axis. It is stated that the demodulated standing in front of the gaming booth.
waveform along the axis of propagation This setup confines sound to a predeter-
is proportional to the second-time deriv- mined sound zone. A laser or LED device
ative of the square of the envelope of ABS Speaker fitted to the ABS speakers can project a
the primary signal: (Ultrasound Emitters) light beam with the same direction as the
sound beam to indicate the sweet spot
bP20 a2 d2 2 on the floor. The gamer within the zone
p2 5 E 1t2,
16rc4zd dt2 can enjoy playing the game without
Fig. 6 ABS and its ultrasound emitters. disturbing the other visitors beyond the
where a, P0 are source radius and pres- defined “tune-in” zone.
sure amplitude at source, respectively;
and is the modulation envelop. The amplitude modulation. Since there is no
other notations are defined identically as theoretical models that can provide
in the KZK equation. accurate descriptions of the entire non-
Berktay derived his equation from linear self-demodulation process, all
the Westervelt equation. Since the KZK these preprocessing algorithms are pro-
equation and the Westervelt equation posed under certain limitations. The
are identical on the axis of farfield, it is trade-off of high fidelity and computa-
not surprising that the Berktay equa- tion is still a challenge to the design of
tion can also be derived from the KZK parametric loudspeakers.
equation. By quasilinear assumption,
the Berktay equation is an analytical Prototyping of
solution to the KZK equation on the parametric loudspeaker
axis of propagation. Based on the above theoretical analy-
sis, we designed and implemented a
Distortion and directional sound projection (or para-
preprocessing technique metric loudspeaker) system, known as
These theoretical models reveal many the audio beam system (ABS). The pro-
essential characteristics of a parametric totyping of the parametric loudspeaker
array. According to Berktay’s equation, a consists of three main components (see
12-dB/octave slope in the frequency Fig. 5): a digital signal processor (DSP),
response of a parametric array is pre- amplifier, and ultrasound emitters. The
dicted. This has been verified by both programmable DSP is the main process-
numerical simulations and experiments. ing block of the parametric array that
Therefore, it is recommended that a low- performs preprocessing, such as equal-
pass filter with 12-dB/octave transition is ization, dynamic range control, carrier Fig. 7 ABS installed in a gaming booth.
NOVEMBER/DECEMBER 2010 23
Conclusion Read more about it Processing, vol. 14, no. 3, pp. 1018–1025,
The development of theory on the • K. Brown, “Acoustic imaging sound 2006.
parametric array in the air provides an dome,” World Intellectual Property Orga- • E. L. Tan, P. Ji, W. S. Gan, and J.
attractive and challenging approach to nization, WO 95/12960, May 1995. Yang, “Distortion analysis and reduction
generate directional sound. A paramet- • H. Mizoguchi, T. Kanamori, S. for the parametric array,” in Proc. Audio
ric loudspeaker has many useful charac- Kagami, K. Hiraoka, M. Tanaka, T. Shige- Engineering Society (AES) 124th Con-
teristics that allow a high directivity, hara, and T. Mishima, “Implementation vention, May 2008, Paper 7426.
controllable beam, and reasonable of invisible messenger system to whis- • E. L. Tan, W. S. Gan, and J. Yang,
implementation size, unlike the conven- per in a person’s ear remotely by inte- “Preprocessing techniques for paramet-
tional loudspeaker, which radiates grating visual face tracking and speaker ric loudspeakers,” in Proc. Int. Conf.
sound waves in omnidirection and array,” in Proc. IEEE Systems Man and Acoustics, Speech and Signal Process-
requires a large array to generate a Cybernetics Conf. (SMC’02), 2002, pp. ing (ICALIP’08), July 2008, pp. 1204–
focused sound. Even though the bass WA2N4(1)– (5). 1208.
quality of a parametric loudspeaker is • P. J. Westervelt, “Parametric acous- • M. F. Hamilton and D. T. Black-
not satisfied to reproduce music, it is tic array,” J. Acoust. Soc. Amer., vol. 35, stock, Nonlinear Acoustics. San Diego,
anticipated that this problem can be no. 4, pp. 535–537, Apr. 1963. CA: Academic, 1998.
solved by psychoacoustics technology. • M. B. Bennett and D. T. Blackstock, • NTU Digital Signal Processing Labo-
One the other hand, as updated advance- “Parametric array in air,” J. Acoust. Soc. Amer., ratory (2010) [Online]. Available: http://
ments in digital processor and ultra- vol. 57, no. 3, pp. 562–568, Mar. 1975. eeeweba.ntu.edu.sg/DSPLab/AudioBeam/
sound emitters are becoming available, • V. P. Kuznetsov, “Equations of AudioBeam.htm
more preprocessing algorithms can be nonlinear acoustics,” Sov. Phys. Acoust.,
conducted in real-time, and more appli- vol. 16, pp. 467–470, 1970. About the authors
cations can benefit in the deployment of • T. D. Kite, J. T. Post, and M. F. Chuang Shi (shic0002@ntu.edu.sg) is
parametric loudspeakers. Hamilton, “Parametric array in air: Dis- currently pursuing his Ph.D.’s degree in
tortion reduction by preprocessing,” J. electrical and electronic engineering at
Acknowledgment Acoust. Soc. Amer., vol. 103, pp. 2871– Nanyang Technological University.
This work was supported by the Sin- 2871, May 1998. Woon-Seng Gan (ewsgan@ntu.edu.
gapore National Research Foundation • W. S. Gan, J. Yang, S. E. Tan, and sg) is an associate professor of electrical
Interactive Digital Media R&D Program M. H. Er, “A digital beamsteerer for dif- and electronic engineering at Nanyang
under research grant NRF2007IDM- ference frequency in a parametric ar- Technological University. He is a Senior
IDM002-086. ray,” IEEE Trans. Audio, Speech Lang. Member of IEEE.
24 IEEE POTENTIALS