Fractals and Music
Fractals and Music
Fractals and Music
Sudhakar Sahoo,
Silicon Institute of Technology, Silicon Hills, Patia, Bhubaneswar, 751024
Email: [email protected]
Soubhik Chakraborty
Department of Applied Mathematics
Birla Institute of Technology,
Mesra, Ranchi-835215, INDIA
Email: [email protected]
*Corresponding author
Abstract— In this paper we have defined one function that has been used to construct different
fractals having fractal dimensions between 1.58 and 2. Also, we tried to calculate the amount of
increment of fractal dimension in accordance with the base of the number systems. Further,
interestingly enough, these very fractals could be a frame of lyrics for the musicians, as we know that
the fractal dimension of music is around 1.65 and varies between a high of 1.68 and a low of 1.60.
Further, at the end we conjecture that the switching from one music fractal to another is nothing but
enhancing a constant amount fractal dimension which might be equivalent to a kind of different sets
of musical notes in various orientations.
The organization of the current paper is as follows. In section II some of the basic concepts on
fractals, fractal dimension and its relation in music as well as the concept of CVT in binary number
system are discussed. In section III, we have generalized the concept of formation of fractals using
CVT in any arbitrary bases of the number systems. Finally a conclusion is drawn in section IV.
1
II(A) REVIEW OF SOME FUNDAMENTALS OF FRACTALS
The scientific community was very much worried due to their inability to describe the shape of cloud,
a mountain, a coastline or a tree on using the traditional Mathematical tools. In Nature, clouds are not
really spherical, mountains are not conical, coastlines are not circular, even the lightning doesn’t
travel in a straight line. More generally, we would be able to conclude that many patterns of nature are
so irregular and fragmented, that, compared with Euclid Geometry –a term, can be used in this regard
to denote all of the standard geometry. Mathematicians have over the years disdained this challenge
and have increasingly chosen to flee nature by devising theories unrelated to natural objects we can
see or feel. After a long time, responding to this challenge, Benoit Mandelbrot developed a new
geometry of nature and implemented its use in a number of diverse arenas of science such as
Astronomy, Biology, Mathematics, Physics, and Geography and so on [1-4]. This new-born geometry
can describe many of the irregular and fragmented (chaotic) patterns around us, and leads to full-
fledged theories, by identifying a family of shapes, now-a-days which we people call ‘FRACTALS’.
Fractal dimension
Now let us try to define what fractal dimension (Similarity dimension) is. Given a self-similar
structure [figure 2], there is a relation between the reduction factor (scaling factor) ‘S’ and the number
of pieces ‘N’ into which the structure can be divided; and that relation is as follows…
N =1/SD, equivalently,
i.e. D =log (N)/log (1/S)
This ‘D’ is called the Fractal dimension (Self-similarity dimension).
Richard F. Voss and John Clarke [8] described the features of spectral density in recordings of
different musical genres and showed their parallels to the peculiarities of 1 f noise observing it to be a
good choice of stochastic composition. They also extended their one-dimensional model to a two-
voice structure which is partly correlated but whose rhythmic shape can be designed using 1 f noise.
Based on these works, Charles Dodge and Thomas A Jerse [9] described the generation of
1 f sequences and produced examples of musical mappings of these different noise forms.
Jeff Pressing [10] mapped the orbit (also called trajectory: the sequence of values that forms the result
for a particular variable of an iterative equation system; these values mostly approach an attractor of a
certain shape) of non-linear equation systems (sometimes called non-linear maps) on musical
parameters. The map output is used to control pitch, duration, envelope attack time, dynamics,
textural density and the time between notes of single events of synthesized sounds. Rick Bidlack [11]
also mapped the orbit of two, three and four dimensional equation systems on musical parameters.
Another interesting approach made by Jeremy Leach and John Fitch [12] derives a tree structure from
the orbit of a chaotic system, the design being inspired by the works of Lerdahl and Jackendoff (see
chap. 4 of [13]). It consists of a hierarchical arrangement of scales and note values and results from
the interpretation of the values of the orbit as hierarchic positions of nodes. Concrete note values are
2
produced by interpreting nodes of higher hierarchic order as pitches that structure a melodic
progression.
In an early work, Przemyslaw Prusinkiewicz [14] described the simple mapping of the generation of
note values from a turtle interpretation of Lindenmayer systems. Prusinkiewicz gives an example of a
Hilbert curve in which the notes are represented by successive horizontal line segments; the lengths of
the segments represents tone durations. The pitches result from the vertical position of the segments
and are mapped on the steps of a C major scale (chap. 6 of [13]).
John McCormack [15] compared stochastic processes, Markov models, different variants of
generative grammar and Lindenmayer systems in terms of their suitability for musical production.
Lindenmayer systems are rewriting systems like generative grammars and generate symbol strings by
applying production rules. Originally they were developed to simulate growth processes.
Conceptually, CVT in base-2 number system is same as performing the bit wise XOR operation of the
operands (ignoring the carry-in of each stage from the previous stage) and simultaneously the bit wise
ANDing of the operands to get a string of carry-bits, the latter string is padded with a ‘0’ on the right
to signify that there is no carry-in to the LSB (the overflow bit of this ANDing being always ‘0’ is
simply ignored).
Example:
Consider the CVT of the numbers (13)10 ≡ (1101)2 and (14)10 ≡ (1110)2. Both are 4-bit numbers. The
carry value is computed as follows:
Carry: 1 1 0 0 0
Augend: 1 1 0 1
Addend: 1 1 1 0
XOR: 0 0 1 1
Figure 1. Carry genereted in ith column saved in (i-1)th column
In the above example, bit wise XOR gives (0011)2 ≡ (3)10 and bit wise ANDing followed by zero-
padding gives (11000)2 ≡ (24)10. Thus CVT (1101,1110) = 11000 and equivalently in decimal notation
one can write CVT (13,14) = 24 . In the next section, CVT is used in different base of the number system
in formation of music fractals.
III(A) GENERATION OF SELF-SIMILAR FRACTALS USING CVT IN DIFFERENT BASES OF THE NUMBER
SYSTEMS
A table is constructed that contains only the carry values (or even terms) defined above between all
possible integers a’s and b’s arranged in an ascending order of x and y-axis respectively. We observe
some interesting patterns in the table. We would like to make it clear how the CV-table is constructed
in different bases of number system.
3
Step 1. Arrange all the integers 0 1 2 3 4 5 6 ... (as long as we want) in ascending order and place it in
both, uppermost row and leftmost column in a table.
Step 2. Compute CVT ( a, b) as mentioned in II(C) and store it in decimal form in the (a, b) position.
Then we look on the pattern of any integer, and we have made it color. This shows a very beautiful
consistent picture, which we see as a fractal as shown below.
Let us do find the fractals in different domain of number system with the help of CVT.
• Production of Fractal in Binary(2-nary) Number System
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2
2 0 0 4 4 0 0 4 4 0 0 4 4 0 0 4 4 0 0 4 4 0 0 4 4 0 0 4 4 0 0 4 4
3 0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6
4 0 0 0 0 8 8 8 8 0 0 0 0 8 8 8 8 0 0 0 0 8 8 8 8 0 0 0 0 8 8 8 8
5 0 2 0 2 8 10 8 10 0 2 0 2 8 10 8 10 0 2 0 2 8 10 8 10 0 2 0 2 8 10 8 10
6 0 0 4 4 8 8 12 12 0 0 4 4 8 8 12 12 0 0 4 4 8 8 12 12 0 0 4 4 8 8 12 12
7 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
8 0 0 0 0 0 0 0 0 16 16 16 16 16 16 16 16 0 0 0 0 0 0 0 0 16 16 16 16 16 16 16 16
9 0 2 0 2 0 2 0 2 16 18 16 18 16 18 16 18 0 2 0 2 0 2 0 2 16 18 16 18 16 18 16 18
10 0 0 4 4 0 0 4 4 16 16 20 20 16 16 20 20 0 0 4 4 0 0 4 4 16 16 20 20 16 16 20 20
11 0 2 4 6 0 2 4 6 16 18 20 26 16 18 20 26 0 2 4 6 0 2 4 6 16 18 20 26 16 18 20 26
12 0 0 0 0 8 8 8 8 16 16 16 16 24 24 24 24 0 0 0 0 8 8 8 8 16 16 16 16 24 24 24 24
13 0 2 0 2 8 10 8 10 16 18 16 18 24 26 24 26 0 2 0 2 8 10 8 10 16 18 16 18 24 26 24 26
14 0 0 4 4 8 8 12 12 16 16 20 20 24 24 28 28 0 0 4 4 8 8 12 12 16 16 20 20 24 24 28 28
15 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32
17 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 32 34 32 34 32 34 32 34 32 34 32 34 32 34 32 34
18 0 0 4 4 0 0 4 4 0 0 4 4 0 0 4 4 32 32 36 36 32 32 36 36 32 32 36 36 32 32 36 36
19 0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6 32 34 36 38 32 34 36 38 32 34 36 38 32 34 36 38
20 0 0 0 0 8 8 8 8 0 0 0 0 8 8 8 8 32 32 32 32 40 40 40 40 32 32 32 32 40 40 40 40
21 0 2 0 2 8 10 8 10 0 2 0 2 8 10 8 10 32 34 32 34 40 42 40 42 32 34 32 34 40 42 40 42
22 0 0 4 4 8 8 12 12 0 0 4 4 8 8 12 12 32 32 36 36 40 40 44 44 32 32 36 36 40 40 44 44
23 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 32 34 36 38 40 42 44 46 32 34 36 38 40 42 44 46
24 0 0 0 0 0 0 0 0 16 16 16 16 16 16 16 16 32 32 32 32 32 32 32 32 48 48 48 48 48 48 48 48
25 0 2 0 2 0 2 0 2 16 18 16 18 16 18 16 18 32 34 32 34 32 34 32 34 48 50 48 50 48 50 48 50
26 0 0 4 4 0 0 4 4 16 16 20 20 16 16 20 20 32 32 36 36 32 32 36 36 48 48 52 52 48 48 52 52
27 0 2 4 6 0 2 4 6 16 18 20 26 16 18 20 26 32 34 36 38 32 34 36 38 48 50 52 54 48 50 52 54
28 0 0 0 0 8 8 8 8 16 16 16 16 24 24 24 24 32 32 32 32 40 40 40 40 48 48 48 48 56 56 56 56
29 0 2 0 2 8 10 8 10 16 18 16 18 24 26 24 26 32 34 32 34 40 42 40 42 48 50 48 50 56 58 56 58
30 0 0 4 4 8 8 12 12 16 16 20 20 24 24 28 28 32 32 36 36 40 40 44 44 48 48 52 52 56 56 60 60
31 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62
Figure 2. A Fractal Structure on using CVT of Different Integer Values in Binay Number System
Carry: 0 0 1 0
Augend: 1 1 1
Addend: 1 1 2
Addition process in ternary: 220
Figure 3. A Fractal Structure on using CVT of Different Integer Values in Ternary Number System
4
Dimension of this fractal
For this fractal [see figure 3], number of self-similar copies N=3 and scaling factor S=1/2, where l is
the initial length.
So, referring the discussion in section II, Fractal dimension D is given by
6=1/(1 / 3)D
Or D= log6/log3 ≈ 1.630929
Figure 4. A Fractal Structure on using CVT of Different Integer Values in 4-nary Number System
Figure 5. A Fractal Structure on using CVT of Different Integer Values in 5-nary Number System
5
music is nearly 1.65. And undoubtedly there are a lot of fractals, which are of fractal dimension 1.68,
and possibly these fractal-frames make new lyrics for music, which needs further investigation.
Similarly, it could be possible to generate fractals in any base of the number system.
Theorem 1: The fractal dimension SD converges to the topological dimension (Euclidian dimension) 2
as the base ‘n’ of the number system diverges to infinity.
Proof:
So, starting from the binary number system the fractal dimension of the generated fractal will go on
increasing with the increase of the base of the number system and finally it converges to the
topological dimension 2.
So far we have discussed we are in position to achieve the fractals having fractal dimension lying in
between [1.58, 2), with the help of CVT.
First of all we paste one base (n-1) CVT fractal into base n CVT fractal. Next, the overflowed portion
is extracted which can be seen as a self-similar figure. These self-similar pieces derived from
overflowed portion would lead to another fractal with some fractal dimension. What we observe, this
very measure (fractal dimension) is the actual increment of fractal dimension in switching from one
base to another base of the number system.
6
(a) (b) (c)
Figure 6. Shows Binary CVT Fractal generetors in Binary, Ternary, Four-nary form left to right.
Figure 7. Binary CVT generetor over Ternary generetor lead to the another fractal.
This generator leads to another fractal; let us calculate the fractal dimension of the generated fractal
as follows:
7
Figure 8. Ternary CVT generetor over Four-nary generetor leading to another fractal.
This generator also leads to another fractal whose similarity dimension is also ≈ 1.585. Thus, CVT
fractal as obtained in this process can be regarded as a relative to Sierpinski triangle. Here also we are
having the same increment as above. In fact, in general we could make a conjecture that if we like to
paste an k-nary CVT generator over an (k+1)-nary CVT generator, then we will be able to have
another generator which lead to the fractal of fractal dimension 1.58, i.e. the attractor fractal is always
homeomorphic to Sierpinski Gasket.
Let us consider two music of fractal dimension 1.66 and 1.68 respectively. And music is nothing but a
collection of notes in different orientation. And we know that, the fractal dimension is a measure of
amount of chaos or amount of information. So in this regard we can say the music of fractal
dimension 1.68 contains more information about musical notes than the music of fractal dimension
1.66. It is noticeable that, these two music are not too much different according to fractal dimension.
Now question is how much more information is there in the music of fractal dimension of 1.68 than
1.66. We are claiming this amount of information is 1.58. It is the firm conviction of the authors that
the music fractals thus generated will have this kind of incremental fractal dimensions with regards to
certain base or bases. So immediately we will take up this investigation as our future research effort.
REFERENCES
[1] B.B. Mandelbrot, The fractal geometry of nature. New York, 1982.
[3] H. O. Pietgen, H. Jurgeens, D. Saupe, Chaos and Fractals New Frontiers of Science, ISBN 3-540-
97903-4, Springer Verlag, 1992
[4] A.K. Ghosh, P. P. Choudhury, R. Choudhury, Production of fractals by various means and
measuring their dimensions with probable explanation, Laser Horizon, Journal of Laser Science
and Technology Centre (LASTEC), vol. 6/No. 2, pp.50-59, 2003
8
[5] A. Bogomolny Cut The Knot! An interactive column using Java applets http://www.cut-the-
knot.org/ctk/Sierpinski.shtml
[6] P. P. Choudhury, S. Sahoo, B. K Nayak, and Sk. S. Hassan, Carry Value Transformation: It’s
Application in Fractal Formation 2009 IEEE International Advanced Computing Conference
(IACC 2009), Patiala, India, 6-7 March, pp 2613-2618, 2009
[8] R. F. Voss and J. Clarke, “1/f noise” in music: Music from 1/f noise, Journal of the Acoustical
Society of America, 63(1), 1978
[9] C. Dodge and T. A. Jerse, Computer music: synthesis, composition and performance, 2nd edn,
Schirmer Books, New York, 1997
[10] J. Pressing, Nonlinear maps as generators of musical sign, Computer Music Journal, 12(2), 1988
[11] R. Bidlack, Chaotic systems as simple (but complex) compositional algorithms,Computer Music
Journal, 16(3), 1992
[12] J. Leach and J. Fitch, Nature, music and algorithmic composition, Computer Music Journal,
19(2), 1995
[14] P. Prusinkiewicz, Score generation with L-systems, Proc. of the International Computer Music
Conference, International Computer Music Association, San Francisco, 1986
[15] J. McCormack, Grammar based music composition, In: R. Stocker, H. Jelinek, B. Durnota and
T. Bossomaier (eds), Complex systems 96:From local interactions to global phenomenon, ISO
Press, Amsterdam, 1996.