The acoustic guitar is a widely used musical instrument in various styles of classical and popular music. Some people think that digital synthesis of guitar music must be difficult, because the player can adjust the timbre in numerous ways: by changing the plucking style, the plucking position, or the force applied to the string, for example. Special playing techniques, such as legato and playing of harmonics, are also available. From the engineering viewpoint, all these are control problems, which are related to the playing of the instrument, but not on the sound production mechanism itself. In fact, we consider the acoustic guitar to be one of the easiest musical instruments for synthesis purposes. Still, it is far from being easy. We have improved our guitar synthesizer since early 1990s, and some of us think that it is still not perfect. For playing music with our guitar synthesizer, we need help from our colleagues at the Sibelius Academy (a university of music in Helsinki).
The physical modeling of the acoustic guitar is based on understanding the sound production chain, which we divide in three parts: (1) the plucking, (2) the string vibration, and (3) the soundbox. The soundbox is mainly an acoustic amplifier that is necessary for obtaining a sufficiently loud sound, but it also colors the string sound in an interesting way. Our physics-based guitar synthesizer is based on a simple algorithm, a so called waveguide model, that simulates the string vibration only. The plucking and the soundbox response are currently both included in the input signal of the synthesizer. This principle is called commuted waveguide synthesis. It allows the extraction of input signals for the synthesizer from recorded guitar tones. This process involves inverse comb filtering of the recorded tones so that the harmonics are suppressed.

The following table lists our acoustic guitar-related research in inverse chronological order. For some publications, the PDF-file and a companion web-page containing sound
examples are provided. These can be found in the leftmost column.
Publication | Short description |

 |
J. Pakarinen,
''Physical modeling of flageolet tones in string instruments.'' In the 13th European Signal Processing Conference (EUSIPCO 2005), Antalya, Turkey, September 4-8, 2005.
|
This paper introduces a physics-based method for synthesising flageolet tones, a.k.a. harmonics, produced by string
instruments. The synthesis model is a waveguide string, connected to a wave digital damper. Also, a simplified model and
a commuted version is presented. |

 |
J. Riionheimo and V. Välimäki, "Parameter estimation of a plucked string synthesis model using genetic algorithm with perceptual fitness calculation," EURASIP Journal on Applied Signal Processing, vol. 2003, no. 8, pp. 791-805, July 2003.
|
This paper describes a genetic-algorithm (GA) based calibration system on top of our earlier calibration routines for the acoustic guitar synthesizer. The GA-based method takes the waveguide string model parameter values from our previous routines and uses a GA to modify them and the excitation sequence in order to find a better fit to a given recorded guitar tone. It compares the original recorded tone and its own resynthesized tone in terms of an auditory error measure. Auditory masking is accounted for in a way similar to some audio codecs. The GA-based method is fully automatic and it slightly improves the sound quality, although synthesis is still based on our old waveguide string model.
|
N/A |
M. Laurson, V. Välimäki, and C. Erkut,
"Production of virtual acoustic guitar music," in Proc. AES 22nd Int. Conf.
on Virtual, Synthetic, and Entertainment Audio, pp. 249-255, Espoo, Finland, June 15-17, 2002.
|
In
addition to discussing the state of the
synthesis model and expressive notation
package at that moment, the virtual
24-string acoustic "super guitar" is
introduced. |
 |
P. Esquef, V. Välimäki, and M. Karjalainen, "Restoration and enhancement of solo guitar recordings based on sound source modeling,"
Journal of the Audio Engineering Society, vol. 50, no. 4, pp. 227-236, April 2002.
|
New
propositions to audio restoration and
enhancement based on sound source modeling
are presented. A case based on the
commuted waveguide synthesis algorithm for
plucked-string tones is described. The
main motivation is to take advantage of
prior information of generative models of
sound sources when restoring or enhancing
musical signals. |
 |
M. Laurson, C. Erkut, V. Välimäki, and M. Kuuskankare, "Methods for modeling realistic playing in acoustic guitar synthesis," Computer Music Journal, vol. 25, no. 3, pp. 38-49, Fall 2001.
|
This
article summarizes the achievements on
model-based sound synthesis of the
acoustic guitar with improved realism. It
discusses the calibration of the
synthesizer and the implementation with
the PWSynth in the ENP (Expressive
Notation Package) environment. |

 |
M. Laurson, C. Erkut, and V. Välimäki, "Methods for modeling realistic playing in plucked-string synthesis:
analysis, control and synthesis," in Proc. COST-G6 Conf. Digital Audio Effects (DAFx'00), pp. 183-188, Verona, Italy, Dec. 7-9, 2000.
|
Discusses
developments in modeling realistic playing
in plucked-string synthesis namely,
transients between notes, pizzicato, and
improved dynamics. Improved dynamics here
means parameter estimation and modeling of
piano, mezzo forte, and fortissimo plucks
with the second-order filters. The
implementation with the PWSynth is also
discussed. |

 |
Penttinen, H., Härmä, A.,
and Karjalainen, M., ''Digital Guitar Body Mode Modulation with One Driving Parameter,''
in Proceedings of the COST-G6 Conference on Digital Audio Effects (DAFx00), pp. 31-36, Verona, Italy, December 7-9, 2000.
|
The resonant guitar body is modelled with a frequency warped filter structure. As a result the perceived size of the guitar body can be change by controlling only one parameter. Hence, creating a perceptually and physically inspired sound effect for the guitar. Can be used for acoustic and electric guitars. |
 |
T. Tolonen and H. Järveläinen, Perceptual study of decay parameters in plucked string synthesis, in
AES 109th Convention, Preprint 5205, (Los Angeles, USA), Sept. 2000.
|
This paper
reports the results of a listening
experiment on perception of variation of
decay parameters in plucked-string
instruments. The results provide
audibility thresholds for variation of the
overall and frequency-dependent decay. The
sounds used in the listening experiment
were created with a digital
commuted-waveguide-synthesis model.
|

 |
T. Tolonen, V. Välimäki, and M. Karjalainen, "Modeling of tension modulation nonlinearity in plucked strings,"
IEEE Transactions on Speech and Audio Processing, vol. 8, no. 3, pp. 300-310, May 2000.
|
This paper
presents a nonlinear discrete-time digital
waveguide model simulating a vibrating
string exhibiting tension
modulation. Furthermore, a technique of
obtaining the tension modulation parameter
from the recorded plucked string
instrument tones is presented. The paper
extends the results reported in
Välimäki99:ICASSP. |

 |
C. Erkut, V. Välimäki, M. Karjalainen and M. Laurson, "Extraction of Physical And Expressive Parameters
for Model-based Sound Synthesis of The Classical Guitar," presented at the 108th AES Int. Convention 2000,
preprint no. 5114, 52 p., Paris, France, Feb. 19-22, 2000.
|
This paper
revises previous parameter extraction
techniques for the classical guitar model
and introduces an iterative optimization
method. Furthermore, it develops
techniques for extracting expressive
performance characteristics, such as
damping, repeated plucks, vibrato,
different pluck styles, and dynamic
variations. Some of the techniques
developed in this paper are used in
Laurson01:CMJ for calibrating the
classical guitar model.
|
N/A |
Laurson, M., Hiipakka, J., Erkut, C., Karjalainen, M., Välimäki, V., and Kuuskankare, M., ''From Expressive Notation
to Model-Based Sound Synthesis: A Case Study of the Acoustic Guitar,'' in Proceedings of the International Computer
Music Conference (ICMC 1999), pp. 1-4, Beijing, China, October 22-28, 1999.
|
The article
discusses the Expressive Notation Package
(ENP). In addition, analysis and synthesis
of vibrato in acoustic guitar is
presented. |

 |
V. Välimäki, T. Tolonen, and M. Karjalainen, "Plucked-string synthesis algorithms with tension modulation nonlinearity,"
in Proc. IEEE Int. Conf. Acoustics, Speech, and Signal Processing (ICASSP'99), vol. 2, pp. 977-980, Phoenix, Arizona, March 15-19, 1999.
|
This paper
describes a digital waveguide model of a
nonlinear vibrating string when the
nonlinearity is essentially caused by
tension modulation. In addition,
computationally efficient simplified
algorithm presented in Karjalainen98:JMC
is devised. |

 |
M. Karjalainen, V. Välimäki, and T. Tolonen, "Plucked-string models: from the Karplus-Strong algorithm to digital
waveguides and beyond," Computer Music Journal, vol. 22, no. 3, 17-32, Fall 1998.
|
This paper describes how a waveguide model of a plucked string can be reduced
to an extended form of Karplus-Strong model. Furthermore, the paper discusses
further extensions to the derived model. The model introduced in this paper is
widely used and extended in physical modeling of plucked-string instruments. |
N/A |
V. Välimäki and T. Tolonen, "Development and calibration of a guitar synthesizer," Journal of the Audio
Engineering Society, vol. 46, no. 9, pp. 766-778, Sept. 1998.
|
Discusses
the development and the calibration
routine used for the digital waveguide
models amply applied by the Acoustics Lab
@ TKK. Also, includes a separate and
parametric model for the lowest guitar
body modes, running at a slower sampling
rate. |
 |
Tolonen, T., Välimäki, V., and Karjalainen, M., ''A New Sound Synthesis Structure for Modeling the Coupling of Guitar Strings,'' in Proceedings of the 3rd IEEE Nordic Signal Processing Symposium (NORSIG-98), pp. 205-208, Vigsø, Denmark, June 8-11, 1998.
|
This paper
presents a new synthesis structure for
physical modeling of the guitar. The model
is capable of simulating the coupling of
vibratory motion in guitar strings. The
model extends the single-delay-loop model
presented in Karjalainen98:CMJ.
|
 |
Tolonen, T., and Välimäki, V., ''Automated Parameter Extraction for Plucked String Synthesis,'' in Proceedings of
the International Symposium on Musical Acoustics (ISMA'97), vol. 1, pp. 245-250, Edinburgh, Scotland, August 19-22, 1997.
|
An improved
parameter estimation algorithm is
introduced for acoustic guitar synthesis:
The sinusoidal modeling technique replaces
the inverse filtering method in the
extraction of the input signal. A few
low-frequency resonances of the body are
extracted and modeled using digital
resonant filters. |

|
Tolonen, T., and Välimäki, V., ''Analysis and Synthesis of Guitar Tones Using Digital Signal Processing Methods,''
in Proceedings of the 1997 Finnish Signal Processing Symposium (FINSIG'97), pp. 1-5, Pori, Finland, May 22, 1997.
This paper received the Excellent Paper Award.
|
The
second-generation commuted waveguide model
is described in this paper that was
presented at a national
symposium. Improvements are introduced in
the parameter estimation procedure and in
the efficient implementation of the guitar
synthesizer. |
N/A |
Karjalainen, M., and Smith, J. O., ''Body Modeling Techniques for String Instrument Synthesis,'' in Proceedings of
the International Computer Music Conference (ICMC96), pp. 232-239, Hong Kong, August 19-24, 1996.
|
This paper
continues the theme of body modeling that
was started five years earlier in the
WASPAA'91 paper. The warped FIR filter is
suggested for obtaining a perceptually
good approximation of an impulse
response. |
N/A |
V. Välimäki, J. Huopaniemi, M. Karjalainen, and Z. Jánosy, "Physical modeling of plucked string instruments with application
to real-time sound synthesis," Journal of the Audio Engineering Society, vol. 44, no. 5, pp. 331-353, May 1996.
|
Our first
journal paper on physical modeling
synthesis describes all aspects of the
early commuted waveguide synthesizer. The
main application is high-quality synthesis
of acoustic guitar tones, but also several
others string instruments are mentioned,
such as the electric guitar and the
kantele. Many methods explained here, for
example the one-pole loop filter (see
Eq. (6)) and the inverse filtering
techniques, have been used for many years
in almost the same form. |
N/A |
Jánosy, Z., Karjalainen, M., and Välimäki, V., ''Intelligent Synthesis Control with Applications to a Physical Model
of the Acoustic Guitar,'' in Proceedings of the International Computer Music Conference (ICMC'94), pp. 402-406,
Aarhus, Denmark, September 12-17, 1994.
|
Rules-based
methods to assist the player to control a
physical modeling synthesizer are
introduced. Related real-time demos
included triggering of strumming using a
single key on a MIDI keyboard, for
example. |


|
M. Karjalainen, V. Välimäki, and Z. Jánosy, "Towards high-quality synthesis of the guitar and string instruments," in Proc.
Int. Computer Music Conf. (ICMC'93), pp. 56-63, Tokyo, Japan, Sept. 10-15, 1993.
|
This is our
second paper on commuted waveguide
synthesis. Parameter estimation methods
using the short-time Fourier transform and
others techniques are described. Julius
O. Smith presented the commuted waveguide
synthesis in the same
conference. |
N/A |
M. Karjalainen and V. Välimäki, "Model-based analysis/synthesis of the acoustic guitar," in Proc. Stockholm Music Acoustics Conf.
(SMAC'93), pp. 443-447, Stockholm, Sweden, July 28-Aug. 1, 1993.
|
The
commuted waveguide synthesis principle was
first published in this paper. The sound
quality of acoustic guitar synthesis was
much improved by using a sampled body
response as the input signal to a
waveguide string model. Julius O. Smith at
Stanford University independently invented
the same idea. |
N/A |
Karjalainen, M. and Välimäki, V.
''Implementation and Real-Time Synthesis of String and Wind Instruments on a Floating-Point Signal Processor,'' in Proceedings of the 6th European Signal Processing Conference (EUSIPCO'92), vol. 3, pp. 1717-1720, Brussels, Belgium, August 24-27, 1992.
|
The early
real-time implementations of the TKK
Acoustics Lab physical modeling
synthesizers are described. Those days we
used a Texas Instruments TMS320C30 signal
processor. |
N/A |
Karjalainen, M., Laine, U. K., Laakso, T. I.,and Välimäki, V.,
''Transmission-Line Modeling and Real-Time Synthesis of String and Wind Instruments,'' in Proceedings of the International Computer Music Conference (ICMC'91), pp. 293-296, Montreal, Quebec, Canada, October 16-20, 1991.
|
Our early
work on physical modeling of the guitar
and the flute is summarized. The
methodology used is currently called
digital waveguide modeling. |
N/A |
M. Karjalainen, U. K. Laine, and V. Välimäki,
"Aspects in Modeling and Real-Time Synthesis of the Acoustic Guitar," in Proceedings of the IEEE Workshop on Applications
of Signal Processing to Audio and Acoustics (WASPAA'91), New Paltz, NY, USA, October 20-23, 1991.
|
Synthesis of acoustic guitar tones using waveguide modeling techniques are described. Experiments on modeling the impulse response of the guitar body using high-order FIR and IIR filters are discussed.
|
N/A |
M. Karjalainen and U. K. Laine, "A Model for Real-Time Sound Synthesis of Guitar
on a Floating-Point Signal Processor," in Proceedings of the IEEE International Conference on Acoustics, Speech and Signal
Processing (ICASSP'91), vol. 5, pp. 3653-3656, Toronto, Canada, May 14-17, 1991.
|
This was
the first publication of the TKK Acoustics
Lab on physical modeling synthesis. A
waveguide model was used for simulating
the string vibration. The Lagrange
interpolation filter was proposed for
implementing the fractional delay required
for fine-tuning the pitch and for
continuously changing the string length,
such as in glissando playing. |
|
|