See discussions, stats, and author profiles for this publication at: [Link]

net/publication/284340332

INHARMONICITY OF THE COPY OF ANTONIUS STRADIVARIUS VIOLIN

Article · January 2014

CITATIONS READS

0 372

4 authors, including:

Zoran N Milivojevic Zoran Velickovic

ACADEMY OF TECHNICAL-EDUCATIONAL PROFESSIONAL STUDIES Nis Nis College of Applied Technical Sciences - Visoka tehnička škola strukovnih studija, …
93 PUBLICATIONS   395 CITATIONS    25 PUBLICATIONS   53 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

User-Centric Analysis of the CAPTCHA View project

Video Watermarking View project

All content following this page was uploaded by Zoran Velickovic on 22 November 2015.

The user has requested enhancement of the downloaded file.

 INHARMONICITY OF THE COPY OF
ANTONIUS STRADIVARIUS VIOLIN

Zoran Milivojević, Zoran Veličković, Dragiša Balanesković

Keywords: Fundamental frequency, Harmonicity, Inharmonicity.

Abstract: This paper presents the algorithm for estimation of inharmonicity of the
string musical instruments. The second part of this paper presents the results of the
application of this algorithm in estimation of inharmonicity of the copy of Antonius
Stradivarious violin built in the transition period from the 19th to the 20th century. The
analysis was made for two ways of string arousal: a) by using a bow and b) by using
the piccicato technique. After that inharmonicityof the instrument was analysed in the
third octave when played by using the pizzicato technique. The results are given in
tables and graphics. In the end the comparative analysis of results was made.

1. INTRODUCTION
The string instruments produce a tone, i.e. they generate an acoustic wave by oscilating
the string. Oscilation of a string is well explained in physics and acoustics, substantiated by
a corresponding mechanical and acoustical model and described by adequate equations.
After the mechanical arousal (the string is put out of balance (or set into vibration) by
picking, striking or by a fiddle-bow) the string begins to oscillate tending to subside and
come back into the equilibrium. The string is vibrating with the basic or fundamental
frequency that depends on the dimensions of the string (length and diameter), the material it
is made of and the tensile strength. Along with the fundamental frequency, because of the
cimplex oscilating of the string (appearing of waves on 1/2, 1/4, 1/8, ...of the string length)
there are acoustic components of the waves on the frequencies that represent the integer
multipliers of the fundamental frequency. In the techniques these spectral components are
called charmonics. The theory of music describes the complex structure of the string
oscilating and the generated tone by means of aliquotes (lat. aliquoties - several times),
where the aliquotes represent charmonic components of the generated tone. The detailed
mathematical analysis shows that within one tone there are thanks to the aliquotes all tones


Zoran N. Milivojević, is with the College of Applied Technical Sciences Niš, Serbia (phone: 381-63-8609206;
e-mail: [Link]@[Link] )

Zoran Veličković, is with the College of Applied Technical Sciences Niš, Serbia
([Link]@[Link]).

Dragiša Balanesković, is with Faculty of Art Niš, Kneginje Ljubice 10, Niš, Serbia.
Z. Milivojević, Z. Veličković, D. Balanesković: Inharmoniciti of the copy of 95
Antonius Stradivarius violin
included (the aliwuotes of various tomes overlap mutually). Different numbers of the
present aliquotes and their different relative intensity within the total sounding determine
the timbre of the sound [1]. Aliquotes are also called partial tones, i.e. the partials.
The detailed analysis of the oscilation of the string shows that because of the parameters
of the string and the tensile strength the frequencies of the harmonics are not with the
integer multiplier, i.e. there is inharmonicity. Inharmonicity of the oscilation of the string
can be described by the coefficient of inharmonicity. Appearing of inharmonicity inevitably
leads to the disturbing of aliquocy. Since there is inharmonicity in some instruments, the
partials can be with frequencies that are: a) integer multipliers of the fundamental frequency
(harmonics) and b) non-integer multipliers (inharmonics) [2]. A tone produced on an
inharmonic instrument is not necessarily unpleasant. In [3] there is a statement that softly
inharmonic tone possesses certain warmth
The quality of the string musical instruments can be, in addition to other parameters,
expressed by means of the degree of inharmonicity, too. The strings are trightened by the
great force so that their elasticity is decreased. The result of this is that the frequecy
positions of the partials are on the positions of non-integer multipliers of the fundamental
frequency. Consequently, the instrument with such strings is not harmonical. Along with
the rigidity of the string, another factor for increasing inharmonicity is the character of the
acoustic impedance of the resonating plate of the piano or the resonating body of the guitar,
violine etc [3]. Inharmonics are outstanding at tones with lower frequencies. On higher
frequencies it is harder to detect inharmonicity. The reason for a good detecting at tones
with lower frequency lies in the fact that a great number of partials is within the range in
which man can hear well. At tones with higher frequencies only a few first partials which
do not express great inharmonicity are found in this region. On the base of the analysis of
inharmonicity it is possible to make conclusions about the kind of the instrument that
produces a tone. In [4] an algorithm for the automatic classigication of tones of the pioano
and the guitar is described. The guitar as well as some other instruments may produce the
same tone on three different strings. By the analysis of the degree of inharmonicity it is
possible to determine the played tone, the string and the field where the string was pressed
[5].
This paper presents the results of the analysis of inharmonicity of the violin, actually a
copy of the Antonius Stradivarius violin, built in the Czeck Republik in the transition
period from the 19th to the 20th century. The first part of the paper deals with the algorithm
proposed by the authors of the paper [6]. That algorithm consists of two entities: a)
estimation of the frequency position of partials and b) estimation of inharmonicity.
Estimation of the frequency position is based on the peaks of the maximum within the
spectrum and on the application of PCC of the convulsive kernel. Parameters of this part of
the algorithm represent a part of the results found by the authors and published in [7] and
[8]. Out of a great number of published formulas for calculating the inharmonicity the
authors have applied the formula from [9]. After that the algorithm for estimation of the
inharmonicity of the double stop published in [12] was described. The second part of the
paper deals with the analysis of inharmonicity of the violin when the strings are aroused to
vibrate by: a) afiddle-bow and b) by picking (pizzicato technique). Then the analysis of
inharmonicity of the double stops from the third octave played on the first and third strings
by the pizzicato technique was done. The obtained results are compared to the results of
 96 ETF Journal of Electrical Engineering, Vol. 20, No. 1, September 2014.

inharmonicity of the concert piano Steinway D [10], pianino August Förster (built in the
Czeck Republic in 1971) [11] and the electric guitar Fender Stratokaster [6].
The organisation of the work is as follows. In the section 2 the algorithm for the
estimation of inharmonicity is presented. In the section 3 the experimental results and the
comparative analysis are presented. The section 4 represents a conclusion.

2. INHARMONICITY AT THE STRING MUSICAL INSTRUMENTS

A. The inharmonicity coefficient of the vibrating string
The theory of music implies harmonicity in defining the frequency composition of a tone,
i.e. that the harmonics (partials) are the integer multipliers of the fundamental frequency,
which mathematically can be presented as:

f k  k  f 0 , k  1,2,... , (1)

where f0 is the fundamental frequency, k. the ordinal number of a partial and fk the
frequency of the partial. The frequency shifting of the partial from the frequency position of
the harmonic represents the inharmonicity of a tone. Inharmonicity is defined by the
coefficient of inharmonicity :

f k  k  f 0 1    k 2 , k  1,2,... (2)

The coefficient of inharmonicity  depends on the material the string is made of and can
be calculated on the base of:

 3 Q  d 4
 , (3)
64  l 2  F

where Q is Jung’s modul of elasticity of the material the string was made of, d the diameter
of the string, l the length of the string and F the tension force.
B. The algorithm for estimation of the partials
The algorithm for estimation of partials is based on the spectrum analysis of the signal x.
First of all the the spectrum X is calculated by using of the discrete Fourier transformation.
After that the position of the maximal spectral component that represents the fundamental
frequency is determined by the method of spectrum peaking [7]. In order to increase the
estimation precision the parametric cubic convolution is applied and the fundamental
frequency f0 is calculated [7, 8]. Estimation of the partials is performed by the algorithm
shown in Fig.1. That algorithm consists of the following steps:

Input: Spectrum Xb of the frame xb of the signal x, the estimated fundamental frequency
f0, number of partials NP.
Outout: partials fP.
Z. Milivojević, Z. Veličković, D. Balanesković: Inharmoniciti of the copy of 97
Antonius Stradivarius violin
FOR p=1:NP
Step 1: peaking of the p-th spectral component
Step 2: Counting of the p-th partial by using PCC interpolation.
END FOR p

Start

f0 , Frame Xb

FOR p=1:NP

Peak picking

PCC kernel PCC Interpolation

fp

f1,...,fNP

End

Fig. 1. Algorithm for estimation of the partials.

C. Algorithm for estimation of inharmonicity
The algorithm for estimation of inharmonicity consists of the following steps [6,9]:

Input: The fundamental frequency f0, partials fP, number of partials NP.
Output: The coefficient of inharmonicity β.
Step 1: Determination of frequencies of the harmonic components:

f k  k  f 0 , k  1,2,..., N P . (4)

Step 2: Counting of the coefficient of inharmonicity according to the algorithm from [9]:

2
 m 2
 fk   fm
 k 
 2
, (5)
2 2 m
k f m2  m  fk 
 k

where m and k are partials and . fm and fk the corresponding frequencies of partials.
 98 ETF Journal of Electrical Engineering, Vol. 20, No. 1, September 2014.

D. Algorithm of estimation of two-tone aliquote distortion

Two-tones represent simultaneous sounding of two tones. Theory of aliquotes [1] sais
that spectral content of one tone consists of harmonics which are, in the same time,
harmonics of other tones. Simultaneous sounding of more tones means spectral overlapping
of corresponding aliquotes. But, due to existence of some tones inharmonicity, aliquotes get
untuned. Discrepancy of one tone’s aliquotes related to other tone’s corresponding
aliquotes, inevitably leads to distortion of reproduced two-tone.
Algorithm for calculation of aliquote two-sound distortion consists of following steps
[12]:

Input: aliquotes of two tones f1,k and f2,k where k=1:Np, and Np is the number of aliquotes
analyzed.
Output: mean aliquote frequency error (MAFE), mean aliquote cent error (MACE),
instrument’s inharmonicity  P .
Step 1: detection of tones which form two-tone,

f1,1  ton_1; f 2,1  ton_2 , (6)

Step 2: Calculation of mutual aliquotes:

f1, k  k  f1,1 ; f 2,l  l  f 2,1

 k  f1,1  l  f 2,1 , (7)

 k , l g , k  1,..., N p , l  1,..., N p , g  1,..., G 
where G is number of pairs (k,l) which fulfill condition of equality of aliquotes.
Step 3: Mean aliquote frequency error:

G
1
MAFE 
G f
g 1
1, k  f 2 ,l
g
. (8)

Mean aliquote cent error:

 ftocent  f   ftocent  f 2,l , f1,1  g ,

1
MACE  1, k , f1,1 (9)
G g 1

where ftocent is transformation function of frequency axis into cent axis with normalization
to f1,1 frequency.
Step 4: Inharmonicity calculation of all two-tones in contra octave:
Z. Milivojević, Z. Veličković, D. Balanesković: Inharmoniciti of the copy of 99
Antonius Stradivarius violin
ND
1
MAFE 
ND  MAFE
d 1 _ 1
d , (10)

ND
1
MACE 
ND  MACE
d 1 _ 1
d , (11)

where ND is a number of two-tones analyzed.

Step 5: Calculation of instrument’s inharmonicity as a mean value of some tone’s
inharmonicity coefficients:
ton _ N S
1
P 
NS 
S ton _ 1
S , (12)

where S is a continuum of tones analyzed, NS number of continuum S elements, and S

inharmonicity factor of corresponding tone.

3. EXPERIMENTAL RESULTS AND ANALYSIS

Further in the paper the coefficients of inharmonicity of the copy of Antonius
Stradivarius violin, built in the Czeck Republick (according to the expertise performed in
Germany in 1993) (Fig. 2). The strings set on it are of the type Dominant Set 135 B, the
worldwide famous producer Thomastic Infeld Vienna (the string e2: chrome steel, ball end,
the string a1: synthetic core, aluminium wound, the string d1: synthetic core, silver wound,
the string g: synthetic core, silver wound). The analysis will imply the estimation of
inharmonicity of all strings and determination of the mean value that will represent
inharmonicity of the instrument for two ways of string arousing: a) playing with the fiddle-
bow and b) playing by the pizzicato technique (by picking the string). In addition to that the
analysis of iharmonicity for the double stops from the third octave with the pizzicato
technique will be done. The double stops were performed by the simultaneous playing of
tones on the first and second strings.
For the purpose of the experiment a base of signals was formed. On the signals the
algorithms described in section 2 was applied. The parameters of the algorithm are T=0.66s,
NFFT=10*218, m=6, k=10.

Fig. 2. The tested violin.

 100 ETF Journal of Electrical Engineering, Vol. 20, No. 1, September 2014.

A. Base
The base consists of all the tones (one octave) on the strings E5 (e2), A4 (a1), D4 (d1) and
G3 (g). The recording was done with the measuring frequency fs=44.1 kHz and 16
bits/sample.
B. Results
In Fig. 3 and Fig.4 there are diagrams related to the tones obtained by playing with the
fiddle-bow (Fig. 3 (G3), Fig. 4 (D4)) while in Fig. 5 and Fig. 6 there are time diagrams of
the tones obtained by the pizzicato technique (Fig. 5 (G3), Fig. 6 (D4)). In every figure a)
represents the time form of the complete signal, b) the time form of the frame lasting 32 ms,
c) the difference between the frequencies of harmonic and inharmonic components and d)
the part of amplitude characteristic to the 10th harmonic. The vertical line denotes the
position of the harmonic component. In Tab. I there are values of inharmonicity calculated
by using of the algorithm described in sectionb 2. Fig. 7 shows frequency positions of
harmonics and inharmonics aliquotes for playing E5-G#5, while Fig 8 shows positions of
notes. Positions of inharmonic aliquotes are presented at fig. 9.a (aliquote G#7), fig. 9.b
(aliquote G#8), fig. 9.c (aliquote G#9). Values of semitones inharmonicity coefficients are
displayed in table II. Frequency and cent values of differences of harmonious and
inharmonious aliquotas are given in table III.

0.5 0.1

0.05
x (t)

0
x(t)

−0.05

−0.1
0 0.01 0.02 0.03 0.04
−0.5 t (s)
0 0.5 1 1.5 2
t (s) b)
a)
1.4

1.2

0.8
e (Hz)

0.6

0.4

0.2

−0.2
0 2 4 6 8 10
Partial k

c) d)

Fig. 3. Signal of the tone G3 produced by playing with the fiddle-bow.

Z. Milivojević, Z. Veličković, D. Balanesković: Inharmoniciti of the copy of 101
Antonius Stradivarius violin
0.6 0.1

0.4
0.05
0.2

x (t)
x(t)

0
−0.2

−0.4 −0.05

−0.6

−0.8 −0.1
0 0.5 1 1.5 2 0 0.01 0.02 0.03 0.04
t (s) t (s)

a) b)
0.6

0.4

0.2

0
e (Hz)

−0.2

−0.4

−0.6

−0.8
0 2 4 6 8 10
Partial k

c) d)

Fig. 4. Signal of the tone D4 produced by playing with the fiddle-bow.

0.3 0.15

0.2 0.1

0.1 0.05

0 0
x (t)
x(t)

−0.1 −0.05

−0.2 −0.1

−0.3 −0.15

−0.4 −0.2
0 0.5 1 1.5 0 0.01 0.02 0.03 0.04
t (s) t (s)

a) b)
14

12

10

8
e (Hz)

0
0 2 4 6 8 10
Partial k

c) d)

Fig. 5. Signal of the tone G3 produced by picking the string..

102 ETF Journal of Electrical Engineering, Vol. 20, No. 1, September 2014.

0.6 0.6

0.4 0.4
0.2
0.2
0
0

x (t)
x(t)

−0.2
−0.2
−0.4
−0.4
−0.6

−0.8 −0.6

−1 −0.8
0 0.5 1 1.5 2 0 0.01 0.02 0.03 0.04
t (s) t (s)

a) b)
40

30

20
e (Hz)

10

−10
0 2 4 6 8 10
Partial k

c) d)

Fig. 6. Signal of the tone D4 produced by picking the string.

4
x 10
2.5

E5 harmonic
2 E5 partial
G#5 harmonic
G#5 partial
1.5
f (Hz)

0.5

0
0 5 10 15 20
Partial k

Fig. 7. Frequential position of aliquote components (harmonics and inharmonious) when

playing (E5,G#5).
Z. Milivojević, Z. Veličković, D. Balanesković: Inharmoniciti of the copy of 103
Antonius Stradivarius violin

B9
G9#
E9
C9
G8#
E8
C8
Note

G7# E5 harmonic
E7 E5 partial
G#5 harmonic
C7
G#5 partial
G6#
E6
C6
G5#
E5
0 5 10 15 20
Partial k

Fig. 8. Note position of aliquote components (harmonics and inharmonious) when

playing (E5,G#5).

C8 C9

B8
B7
A8#
A7#
A8
A7
Note
Note

G8#
G7#
G8
G7

E5 partial F8# E5 partial

F7#
G#5 partial G#5 partial
F8
F7
E8
E7 7.5 8 8.5 9 9.5 10 10.5
3 3.5 4 4.5 5 5.5 6
Partial k Partial k

a) b)

B9
A9#
A9
G9#
G9
F9#
F9
Note

E9
D9#
D9 E5 partial
C9# G#5 partial
C9
B8
A8#
A8
G8#
G8
15 16 17 18 19 20
Partial k

c)
Fig. 9. Inharmonious aliquotes position details with E5-G#5: a) G#7, b) G#8, and c)
G#7.
 104 ETF Journal of Electrical Engineering, Vol. 20, No. 1, September 2014.

Table I
Coefficient of inharmonicity of tones of the empty string of the violin.
Fiddle-bow Pizzicato
F0 [Hz]
 (x10-6)  (x10-4)
G3 196 6.05 1.31
D4 293.66 1.38 2.86
A4 440 3.55 3.82
E5 659.25 2.06 1.13
βsr=3.26 βsr=2.28

Table II
Coefficient of inharmonicity of tones of the first and second string of the violin.
Tone S (x10-4)
E5 1.13
F5 6.9341
Fis5 6.8746
G5 2.5016
Gis5 6.1697
A5 18.2261
P 6.9726

Table III
Coefficient of inharmonicity of tones of the empty string of the violin.
Tone MAFE [Hz] MACE [cent]
E5-F5 413.0169 63.6348
E5-Fis5 654.7776 101.5170
E5-G5 93.9050 12.8624
E5-Gis5 1067.6 150.2831
E5-A5 276.6344 64.3648
MAFE  501.1867 MACE  78.5324

C. Analysis of the results

On the base of the results shown in Fig. 3 through Fig. 9 and in Tab. I-III it can be
concluded that:
a) the coefficient of inharmonicity at the pizzicato technique is 2.28x10-4/3.26x10-6=69.9
times greater than the one obtained when the string was aroused by playing with the fiddle-
bow,
b) values of the coefficient of inharmonicity of the sound produced by the fiddle-bow is
of order 10-6 and can be considered that thare is no inharmonicity
c) when the pizzicato techhnique is used, inharmonicity of the violin is aproximately in
the same limits as at the pianino August Förster (1÷2)x10-4 [11] and the electric guitar
Z. Milivojević, Z. Veličković, D. Balanesković: Inharmoniciti of the copy of 105
Antonius Stradivarius violin
Stratokaster (0.8÷2.5)x10-4 [6], while it is greater for the size order at the concert piano
Steinway D (0.6÷0.8)x10-4 [10].
d) analysis of the double stops produced by the pizzicato technique on the first and
second strings showed that the mean inharmonicity error was MAFE  501.1867 Hz, while
the mean cent error was MACE  78.5324 cent. In comparison to the results of
inharmonicity of the double stops of the pianino “August Förster” in the counter octave
( MACE  31.8357 cent) [12] the analysed violin obviously had 2.46 times greater cent
error.
The mentioned results point to the fact that at instruments where the sound is produced
by picking, their inharmonicity is considerable because of irregular oscillating of the string
after the arousal when it is let to oscillate freely. On the other hand, when aroused by a
fiddle-bow, it is not oscillating freely and inharmonicity is negligible
Considering the fact that on the cent scale the distance between any two adjacent
halftones is 100 cent, it can be concluded that the mean cent error is 78% of a halftone. In
Fig. 4 it can be seen that considerable aberrations of the inharmonic components in relation
to the harmonic ones for the analysed double stops occur above 10 Hz. In this range man
has lesser sensitivity. Reviewing all the presented results it can be said that the analysed
violin has good acoustic parameters, which is confirmed by its price on the market.
Further investigations will be directed toward testing of greater number of instruments
from the group of string instruments (that can be played by picking or by using the fiddle-
bow) and according to the results hopefully a global conclusion will be made

4. CONCLUSION
This paper has presented the results of testing the copy of Antonius Stradivarius violin
built in the Czeck Republic in the transition period from the 19th to the 20th century,
regarding inharmonicity. The testing was performed for two ways of producing the sound:
by picking the string (the pizzicato technique) or by playing with a fiddle-bow. The results
show that inharmonicity is negligible if the fiddle-bow is used (the coefficient of
inharmonicity of order 10-6), while if it the pizzicato technique is used the coefficient of
inharmonicityis of order 10-4. The mean value of inharmonicity of the double stops of the
violin in the second octave is MACE  78.5324 cent, which is 2.46 times higher in
comparison to inharmonicity of the double sounds of “August Förster” for the double stops
from the counter octave. Considering the fact that the spectral components of tones from
the second octave are in higher parts of the spectrum, where man has lesser sensitivity, it
can be said that inharmonicity of the violin played with the pizzicato technique is of the
same order of magnitude as of the pianino “August Förster” and the electric guitar “Fender
Stratokaster”.

REFERENCES
[1] M. Tajčević, Osnovna teorija muzike, Prosveta, Beograd, 1962.
 106 ETF Journal of Electrical Engineering, Vol. 20, No. 1, September 2014.

[2] S. Lixin, Z. Junxing, and P. Lingbin, "Piano fundamental frequency estimation algorithm based
on weighted least square method", in Proc. Int. Forum Inf. Technol. Applicat., IFITA ’09, Vol. 2,
pp. 155–158, May 2009.
[3] L. Ortiz-Berenguer, J. Casajus-Quiros, M. Torres-Guijarro, J. Beracoechea and J. Aranda,
"Simple modeling of piano inharmonicity due to soundboard impedance", In Proc. 120th AES
Convention, Preprint, Vol. 66, pp. 1–6, 2006.
[4] D. Fragoulis, C. Papaodysseus, M. Exarhos, G. Roussopoulos, T. Panagopoulos and Dimitrios
Kamarotos, "Automated Classification of Piano–Guitar Notes", IEEE Transactions On Audio,
Speech, And Language Processing, Vol. 14, No. 3, pp. 1040-1050, May 2006.
[5] I. Barbancho, L. Tardon, S. Sammartino, A. Barbancho, "Inharmonicity-Based Method for the
Automatic Generation of Guitar Tablature", IEEE Transactions On Audio, Speech, And
Language Processing, Vol. 20, No. 6, pp. 1875-1868, August 2012.
[6] Z. Milivojević, D. Brodić, "Inharmoničnost bas sekcije Fender Stratocaster električne gitare",
Informacione Tehnologije IT’13, Zbornik str. 89-92, Žabljak, Crna Gora, 2013.
[7] Z. Milivojević, M. Mirković, "Estimation of the fundamental frequency of the speech signal
modeled by the SYMPES method", International Journal of Electronics and Communications
(AEU), vol. 63, no. 3, pp. 200-208, 2009.
[8] Z. Milivojević, D. Brodić, "Estimation of the fundamental frequency of the speech signal
compressed by G.723.1 algorithm applying PCC interpolation", Journal of Electrical
Engineering, vol. 62, no. 4, pp. 181-189, 2011.
[9] M. Hodgkinson, J. Wang, J. Timoney and V. Lazzarini, "Handling inharmonic series with
median-adjustive trajectories", 12th Int. Conference on Digital Audio Effects DAFx-09, Como,
Italy, Sept. 1-4, 2009.
[10] A. Galembo and A. Askenfelt, "Signal Representation and Estimation of Spectral Parameters by
Inharmonic Comb Filters with Application to the Piano", IEEE Transactions on Speeach and
Audio Processing, Vol. 7, No. 2, pp. 197-203, March 1999.
[11] M. Rajković, D. Milosavljević, Z. Milivojević, "Inharmoničnost bas sekcije August Forster
pijanina", INFOTEH’13, Sekcija Računarski sistemi i softver – informacioni sistemi, str. 980-
986, Jahorina, 2013.
[12] Z. Milivojević, M. Rajković, D. Milosavljević, Inharmonicity Of Two-Tones In Contra Octave
Of Upright Piano, XLVIII International Scientific Conference On Information, Communication
And Energy Systems And Technologies ICEST 2013, Section: Signal Processing, Ohrid, 2013.

View publication stats