Sergei Muratov

The Art Of The Violin

Design
The Art of the Violin Design
By
Sergei Muratov
These two processes, being projected in the history of the development of the scientific
study of the musical instrument (the instrumentology) as well as in the history of
musicology on the whole, go stage by stage. The Renaissance can be considered as the
first stage of the development of an idea about the violin (the practical work of the Italian
luthiers) when the theory of the violin design was studied esoterically, i.e. it was explored
within a school and passed on to pupils verbally.
Proceeding from the proposition that the practical and theoretical creative work of the
luthiers was in syncretic unity with art, science, philosophy and religion, one can say that
the luthiers were sufficiently informed in the sphere of exact sciences. In the 17th century
the disintegration of the syncretic principle led to the decline of the luthiers' Golden Age,
the last specimens of which were gone in the 18th century (A.Stradivari, Guarneri del
Gesù and others).
Having no chance to be apprenticed to the great luthiers, makers simply copied their
instruments in the hope of repeating their sonority. On this tide of universal respect but
misunderstanding of the principles of the violin design (made by the great specimens of
this kind of art), the Academy of Arts and Sciences at Padua announced a competition for
the best work about the violin design. Antonio Bagatella (1726-1806), author of
Regulation for Constructing Stringed Instruments (1782), obtained a prize. This
publication created astonishment among all European connoisseurs and luthiers of his
day, and was quoted frequently through subsequent ages.
Of course every investigator did his bit to this methodology of the violin design, but all of
them kept the very main sign by which one can determine their idol: the use of compasses
for drawing, or rather copying, the outline of the instrument. Basically these luthiers
copied the Amatis' violins, not the Stradivaris' or the Guarneris', whose instruments were
copied from the 19th century. And one of the earliest works about violin acoustics,
Memoir on the Construction of Stringed Instruments (Paris, 1818) by Felix Savart
(1791-1841), is related to this time. He completely rejected the traditional violin form,
supposing the acoustic of the stringed instrument to be independent of its shape. Savart's
'trapezoid-violin' was tested and compared with Cremonas, eulogized as being superior,
etc., but subsequently found no endorsement from soloists. Practically all innovators who
cardinally changed the shape of the violin suffered the same fate.
So the second stage of the development of the idea about the violin can be called a period
of search for new methodological bases.
The third stage, taking place in our time, is that basic methodological schools bound up
with acoustics or with the search for the rational construction of the musical instrument,
or with physico-chemical processing, or with other practical and theoretical features of
the violin design, have already formed. And the subsequent study of the 'secrets' of the
violin as an acoustical phenomenon takes the path of the interaction of these methods, i.e.
it reverts to its 'syncretic past', but now on a new phase of the dialectical development.
If the luthiers had worked like a technical designer, i.e. firstly with an idea and a working
drawing, then the making of parts, their arrangement and, finally, the adjusting of the
finished articles, then their creative work would be characterized as technological. And
after analyzing the instruments themselves by different methods, determining the
physical data of their parts, even simply copying them, one could emulate their tone. The
same causes produce the same effects, don't they? But size-for-size copies do not work
because the frequencies are determined not only by the dimensions and construction of
the instrument but also by the wood's mechanical properties, which vary from sample to
sample.
There are also the elements of artistic work in the luthier's activity, concerning not only
the outward appearance of the instrument, but its tone, which are impossible to appraise
by any modern apparatuses, but only directly by man. So the quality of a violin's sound is
a result of both the luthier's sophisticated hearing and his scientific knowledge.
In the primary stage of a violin construction the technological problem is most important,
whereas in the process of work the main role gradually passes to the artistic outcome.
Proceeding from this the luthier's creative work must be considered from both points of
view: science and art.
The problem of the development of the violin in the historical plane as well as its design
must be the basis for the scientific research of the modern instrumentologist. As is
generally known this development involves the progressive changes in size, shape, and
function of an object during its historical existence. In the process of the development in
the intermediate stages the object's condition has a certain characteristic, which I name
the modus. In its application to the violin the modus is the configuration, design and
sound atmosphere of the instrument.
As to the development of the violin in the historical plane, this issue is only explored to a
certain extent, because of the limitations of historical and archaeological documents and
finds.
It is impossible to restore the methodology of the violin design of the Italian classical
schools in full, although the Civic Museum of Cremona contains a collection of moulds,
drawings, sketches, templates and original studies by A.Stradivari. But the working
drawing, displaying the way of thinking of the great luthiers, is absent. The whole of this
set, which was used for the design of the different parts of the instrument, comprises only
a number of the copies of a principal drawing which, if it only existed, would throw light
on the 'mystery' of the creative process of Stradivari and other luthiers of that time.
I shall not dispute whether such a drawing really ever existed, but too few investigators
have made an attempt to re-create it. In these works one can see the centuries-old interest
of people in the underlying mathematical regularities of the arts.
The question of mathematical prerequisites in the beautiful, and the role of mathematics
(specifically geometry) in the arts stirred the ancient Greeks and Babylonians. One can
even assume that mathematics and the arts came into existence almost at the same time in
view of the religious and philosophical searches of man, and that there are close and
varied connections between mathematics and the arts.
The role of mathematics in laying bare the secrets of the arts has been traced in the
creative work of such people as Pythagoras, Vitruvius, Albrecht Dürer, Leonardo da
Vinci and Thomas Hobbes. The enormous importance of geometry was not only relevant
to the above-named artists and architects, but to great luthiers too. Unfortunately, in
contrast to the former, the luthiers did not leave any theoretical propositions about their
work.
If we discuss the geometry of the violin, then the question is: What can we assume the
basis of its design - the aesthetic principle (beauty, elegance) or to be the physical one
(acoustics, mechanics)? The borderland between scientific and the artistic work turned
out to be a rather impassable obstacle for the mutual assimilation of two different worlds
lying on opposite sides: the world of scientific notions and the world of artistic images. In
the scholarly and scientific study of the musical instrument, employing geometry to build
a bridge between these two worlds proves to be difficult. Numerous popular methods of
the geometrical analysis of the string instruments, which were made by great luthiers,
have no acoustic substantiation and, what is more, the aesthetic advisability of such
methods gives rise to doubts. Various parts of the violin are drawn with compasses by the
mere selection of radii, which rather looks like copying, than a search for logical
regularity.
Of course at all times both the architects and the engineers used compasses and a ruler
when making the working drawing. And it is small wonder, as basically straight lines and
arcs are used in the constructions. But, for example, when designing aircraft, high-speed
cars, radar, etc., the circles of the compasses are not a great help. This kind of
construction can be designed only with some mathematical curve. Our task is finding
such a curve, one that would be up to the requirements of the violin design, i.e. it has to
be elegant and must allow for the curvatures of the instrument and needs to meet the
criterion of acoustic properties and their projection. As the character of the curvature of
the whole of the instrument is invariable, we must use only one kind of mathematical
curve, which we can increase or decrease, according to the given parts of the violin. In
other words, we must find such a standard module, which when scaled up or down can be
used to design any stringed and bowed instrument.
By analyzing different mathematical curves, I come to the conclusion that there exists
only one curve that is up to the requirements of the violin design. It is the Cornu spiral or
clothoid (C and S are the so-called Fresnel integrals) (Figure 1), important in optics and
engineering.
Clothoids have been used in engineering design for many years. In the past the spirals
have been found manually by draftsmen. This was a tedious process, which I did myself
twenty years ago in 1981 solving the problem of violin design for the first time. It is
much easier to use a computer to draw and calculate the position of the clothoids. The
design curve of a violin will be made up of segments of clothoids joined in such a way
that the curvature is continuous throughout.
The clothoid has the following parametric representation in Cartesian coordinates:

where the scaling factor a is positive, the parameter t is non-negative.

Figure 1: The Cornu spiral.
A curve parametrized by an arclength such that the radius curvature is inversely
proportional to the parameter at each point is a Cornu spiral. In contrast to another spirals
the clothoid has this very important property: the radius curvature starts from infinity and
aspires towards zero, continually approaching its asymptote (centre of volute), while the
curvature aspires to the ideal form - to the circle.
The curve of the violin outline is formed by joining segments of clothoids. In all cases it
is necessary to solve a nonlinear equation (I used the scale Tool of the computer
software) to find the scaling factor a. The angle of rotation of the tangent in each spiral
will be found empirically.
The other important factor in the violin's geometrical design is the use of proper
proportions. Throughout the ages, designers and architects have attempted to establish
ideal proportions. The numerically simple ratios 1:2; 2:3; 3:4; 4:5; 3:5, etc., were
considered the preferable proportions, but the most famous of all axioms about
proportion was the golden division (1.6180339...), established by the ancient Greeks.
According to this axiom, a line should be divided into two unequal parts, of which the
larger is to the smaller as the whole is to the larger. At a mathematical expression I shall
present it as μ
To construct two finite straight lines in the ratio of the golden division is very easy. Let
ABEF be a square (Figure 2).

Figure 2. The golden rectangle.

Let the point D to divide AF in half. If AD = DF, then BD is the hypotenuse of the right-
angled triangle with the ratio of catheti (the other two sides) as 1:2. Therefore, by the
Pythagoras' Theorem, the length of the hypotenuse is √5. The ratios of the sides of this
triangle are very simple: AD/AB = 1/2, BD/AD = √5/1, BD/AB = √5/2. And therefore:
(AD+BD)/AB=(√5+1)/2= 1.6180339...
If μ = 1.6180339..., then 1/μ = (√5-1)/2 = 0.6180339... .
If BEM is an arc of a circle with centre D, then AM/AB = μ In that way we can construct
the finite lines which will be longer than the original line in the ratio of the golden
division.
The rectangle ABPM, having the side AM = μAB, is called a golden rectangle. ABEF is
a square, and we observe that the rectangle FEPM is also a golden rectangle, since EF =
μFM. If we now take this rectangle FEPM, and mark off a square EPTS from it, the
remaining rectangle FSTM also will be golden, and we can continue this process as long
as we want. In this way we can construct the lines which will be shorterthan the original
line in the ratio of the golden division.
If the proportion μ or 1/μ is found by the simple expedient of working out a problem of
the golden rectangle, the proportion 2/μ or μ/2 (very important in our work too) is
determined by the next method: with radius DA draw the arc of a circle centre D to cut
BD produced at N. Then BD is divided in the 2/μ ratio at N. To prove this, we note that
DN = 1 and NB = √5 - 1. If μ = (√5 + 1)/2 and 1/μ = (√5 - 1)/2 and 2/μ = (√5 - 1), then
NB/DN = 2/μ.
With radius BN draw the arc of a circle centre B to cut AB produced at K. Then AB is
divided in the golden section at K. AB and AM can be the sides of the golden triangle.
The golden triangles are constructed by the following method (Figure 3):
Figure 3. The golden triangles.
We see that golden triangle ABC is divided into three golden triangles AEC, ADE and
DBE, the sides of which are: AD = DE = EC = 1; DB = BE = AE = AC = μ; AB = BC =
1 + μ = μ². Another golden triangle, having the angles 90º and 54º and 36º with the ratio
of 5:3:2, is very interesting too. In this right-angled triangle the ratio of the big cathetus to
the hypotenuse is μ : 2 = cos 36º, hence the formula which binds the golden section and
π:
μ = (√5+1)/2 = 2 cos π
The geometry of the Great Pyramid of Khufu at Giza is a golden triangle too (Figure 4).

Figure 4. The Great Pyramid of Khufu at Giza.

If cos 51.82729º = 1/μ, then AD/AB = 0.6180339..., AB/AD = 1.6180339...
One can divide all A.Stradivari's creative work into a few periods:
1) From 1666 to 1688 Stradivari had worked after the Amati model. From 1689 he
experimented with the large model by N. Amati and enlarged it some more.
2) In 1692 Stradivari had created the 'elongated' model of a violin.
3) In 1698 he had returned to the Amati model, working on the model by Antonio and
Hieronymus Amati.
4) From 1705 to 1725 Stradivari worked with his own original model. It was his best
period of creative work.
5) From 1725 to 1737 - the last years in Stradivari's creative work - one can see his
declining powers, which can be attributed to his old age.
It would be true to suppose that the alterations of Stradivari violin forms, which are being
retraced in the span of his creative work, have an acoustic substantiation. If the curvature
of the internal mould of the Amati violin has a guitar-shaped form, the Stradivari has a
form as if its lines are affected at the middle C-bout. Also, Stradivari has altered the
arching and thickness of both the belly and the back and has revised the proportions
between different parts of the instrument.
But the process of the development of the violin, which started in the 16th century, was
completed only at the beginning of the 19th century, when the violin was modernized by
replacing the neck, the fingerboard, the bridge, the bass-bar and the soundpost. So the
Baroque violin is only an approximation of the ideal proportionality of its different parts
as well as of the whole of the instrument. And the sound of these instruments is being
defined by us now when they have the 'modernised' neck, fingerboard, etc., whereas in
the 18th century they sounded differently.
The most apparent modification in sound comes from the different strings of present
instruments. Here a new material for the production of strings has become stronger and
longer because of the new long neck, and the bridge has been raised, which increases the
string strain on the table. And, what is more, the eighteenth-century pitch, in general, may
be taken only as a¹= 422.5 Hz (according to the pitch of Handel's English tuning-fork,
which still exists). Therefore in the days of Tartini the strain on the strings was 29 kg,
whereas it is now 90. As resistance to this strain the original bass bar has been replaced
by one longer and stronger. The sum effect of these alterations was to develop the
optimum sonority of which the instrument was capable.
Hence one can say that the great Italians did not create the present violin sound, which
supposedly has passed ahead beyond its time (they could not even imagine how their
instruments could sound after 'modernization', you know), but they simply made the
violin body with the rich potentialities which did not reveal itself in 18th century in full.
Aforesaid changes in physical design started from about the beginning of the 19th
century. It was a second stage of the development of the violin. The paradox was that
luthiers, modernizing the old violins and making the new one, disregarded the traditional
methods of the violin design and did not study the instrument as well as the pupils of the
old time did it. And the 'secrets' of the Italian violins were gone.
We have some problems when we try to make an appraisal of the quality of the
instrument's sonority. Nowadays two methods exist: subjective, based on the hearing of
the investigator, and so-called objective, when special devises are used. The investigator
has a wide range of powerful analytical techniques at his command now. These
instruments take the frequency of sound and its intensity. The instruments, which
measure the sound pressure of the individual harmonics of the complex tone, analyze its
spectrum.
Because timbre is a quality of auditory sensations produced by the tone of a sound wave,
the timbre of the particular sound depends not only on its wave form, which varies with
the number of overtones, or harmonics that are present, their frequencies, and their
relative intensities, but on the some subjective peculiarity of our auditory sensation too.
The fact is that every harmonic, which is heard by a musician, is a compound tone,
consisting of an objective overtone, which can be fixed with the sound spectrograph, and
subjective resultant tones, which only occur in our consciousness because of interaction
between the objective overtones. One of them (difference tone) is a low one tallying with
the difference between the two vibration numbers, and the other of them (summation
tone) is a high one, but a very much fainter one, tallying with their sum. And so, it is hard
to describe the timbre of any instrument with objective and subjective components only
with analytical techniques; this process also needs a personality appraisal of the musician.
Let us examine two spectrograms. The spectrum of (Figure 5) the force exerted by a
bowed string at the bridge, and (Figure 6) the sound radiated by violin playing the same
note (open G-string). As we see (Figure 5), the amplitudes of harmonics gradually
decrease from first to last that is quite naturally. The spectrogram of the violin sound,
which is heard by man, would be like the first one, if only our brain could draw such
pictures. But really (Figure 6) the violin insufficiently radiates both the first (G3) and the
second (G4) harmonics.

Figure 5. The spectrum of the force exerted by a bowed G-string at the bridge.
Figure 6. The spectrum of the sound radiated by the violin playing the open G-string.
Strange as it may seem, but D5 sounds louder then other harmonics. The fact that we hear
the fundamental tone (the first harmonic) as loudest is a function of our brain which
creates and adds the amplitudes of the difference tones to the actually sounded
harmonics. Because the difference between frequencies of the adjacent harmonics always
is equal to the frequency of the fundamental tone, the insufficient amplitude of the first
harmonic is compensated by the difference tones of all adjacent pairs of the Harmonic
Series.
Now I want to dwell on the highly interesting moment connected with the note D5 and
length of a violin. Firstly I calculate the wavelength of the D5. In dry air (at 0 C and a
sea-level pressure of 1013.25 millibars) the speed of sound is 331.29 m/second. If the
frequency of A4 is 440 Hz, then open G string has 195.5 Hz. Hence the third harmonic
(D5) has 587.7 Hz. Dividing the speed of sound by 587.7 I find the wavelength of the D5:
33129 cm / 587.7 Hz = 56.3706 cm.
Getting ahead of my statement (details will be in the next chapter); I produce my
calculations of the length of a violin. As initial value I use πcm. (3.14159265...cm),
which is the first term of the progression. If the golden division is the common prime
factor, the seventh term of the progression will be 56.3735 cm, which is the length of the
whole instrument.
Because the spectrum of the sound radiated by the violin is inadequate to the timbre,
which is heard by us, it is naturally to ask, 'Is it possible to divine the sound of the
musical instrument, working at the acoustic of the parts?'
If the final results were dependent on the sum of the timbre of the different violin parts,
this problem would be worked out by merely tuning them up according to the certain
principle, copying some great instrument. But really all is far more intricate.
Sound is produced when a vibrating surface interacts with the surrounding air. As the
large, lightweight plates (the belly and back) moves forwards and backwards, the
surrounding air pressure is increased and decreased. These pressure variations speed
away from the source as sound waves at 331.29 m per second. The sound waves,
traveling from the outside and inside of the plates, differ in phase by 180º.
If the violin body was a single table, i.e. in free air, it would be like a fish out of water.
To see why a bare belly sounds bad, consider Figure 7.

Figure 7. Why a bare belly is inefficient at low frequencies.

The plus signs represent an increase in pressure as the belly moves against the air; the
minus sign, a decrease (a). When air from the high-pressure side of the belly mixes with
air from the low-pressure side, sound cancellation occurs. At high frequencies, the sound
is directional, so little mixing occurs; however, for frequencies at which the wavelength
is long compared to the size of the belly, the waves can curve back around the belly so
that the out-of-phase waves mix (b). One of the basic requirements of a violin body is to
block this unwanted mixing of out-of-phase waves (c).
As a violin body has small holes (the ff-holes), the air in the body retains its ability to act
like a spring, while the air in the f-holes acts like another diffuser. This air diffuser
vibrates in phase with some frequencies and out of phase at others (Figure 8). So sound is
created not only by the motion of two plates but also by the air being squeezed resonantly
in and out of the ff-holes. This system acts as a resonator, properly called a Helmholtz
resonator. The frequency of resonance for any Helmholtz resonator is determined by the
compliance of the air in the container and the mass of the air in the hole.

Figure 8. How air moves at different frequencies. At some frequency, the f-hole air
moves in phase with the belly (a). At another frequency, the f-hole air moves out of phase
with that of the belly (b).
The violin radiates the wide spectrum of the sound. Owing to the shape of the violin
body, the phases of the harmonics are altered, when they go out of the ff-holes. It is
conducive to the subtraction and addition of its amplitudes. One of the peculiarities of a
Helmholtz resonator is that the sound that is radiated from a hole does not vary with the
size of the hole if it has the round form (like in a guitar). When the hole has the form of a
slit (like in a violin) the radiated sound varies appreciably through the ratio of length and
width of the hole. So, the narrow hole is used by the luthiers for adjusting the sound
quality.
Since the ffs with the internal volume of air in the violin body form the resonance system,
it is very important for a luthier to check the correlation between these two volumes of
air. The balance is achieved by the increase or reduction of the volume of air into the
instrument's body so thereby changing the parameters of the ffs. The configuration of the
body has no small importance.
Certainly, I remember about the nature of arcs of the belly and the back, their thicknesses
and adjustment; both the whole boards and their separate areas. However it is not
possible to define exhaustively what work is necessary to be conducted with all the
details of an instrument to get the Italian timbre. Any attempt to limit the class of
considered phenomena by a type of an equation or an enumeration of some physical
characteristics usually brings about failure, an example will always be found that will not
go into the accepted scheme.
The use of probabilistic-statistical methods of study in the field of violins (study of the
Chladni patterns, the laser interferograms, thicknesses and tones of separate areas of the
belly and the back and a great deal of other concerns including holograms and
voiceprints) reveals the effect of a total action of unambiguous dynamic laws.
The wave processes, occurring in the system of body-ffs-outside air, have a complex
nature and must be described by different systems of equations. However, for the
understanding of the most important phenomena, occurring in the given system
(interference, diffraction, reflection and refraction, dissipation and etc.) there is no need
to analyze the source, generally speaking, complex systems of equations. The simple
effects, as a rule, are described by simple and universal mathematical models.
The violin body is a closed space for the sound field (while for this explanation the ff-
holes have no importance). In the closed space the sound waves, repeatedly reflecting
from borders, form the complex field of the air's oscillatory moving, which is defined not
only by the characteristics of the sound source (in the violin body the belly and the back
are these sources), but also by the geometric form and sizes of the space, and the ability
of the borders of the space to reflect, miss and absorb the acoustic energy. The picture of
the wave processes, occurring in the violin body, gets complicated by the presence of the
ff-holes.
Because of its small volume the body of a violin cannot be diffusive, so the sound waves
of its field are coherent and there are the stable phenomena of interference in it. As a
result of that the secondary sources of the sound waves, which are located between the
actual sources of the waves (the belly and the back), appear in a certain point of space in
the violin body (the Huygens-Fresnel' principle). Due to its contours the body of a
violin forms this secondary source in the region of the ffs.
On the output from the body through the ffs the sound wave is changed into the wave
pencil. Sometimes this pencil can be considered as a ray, whose behavior is described by
the laws of the geometric optics. However the spread of the real wave pencils is different
from the behavior of the rays. The reason for this difference is due to the phenomena of
diffraction.
We cannot get the exact and mathematically correct decision of the diffraction of the
sound wave when it passes through the ffs, since this will entail greater difficulties: the
very complex form of the screen (the belly) and not less complex form of the slot (the
ffs). So the good ear for music of the luthiers is very important for the determination of
the quality of the sound, passing through the ffs. But if we take into consideration only
the good ear of the luthiers, we must finish cutting the ffs after the instrument was
assembled? It was done by A.Stradivari whose ffs never agree with the intended drawing
on the inner face of the belly.
It is hardly probable that great masters worried about the external aesthetics of the ffs
more than about the acoustics of the instrument. Many masters, including Guarnerius del
Jesu, cut the ffs crudely enough in general, then stopped to consider if contented with the
sound knowing that a drastic 'correction' could harm the sound quality. Precisely such a
work method is substantiated with the ffs by modern theoretical physics (the Kirchhoff's
method), which proves that when the wave passes through a screen with a hole, its
spectrum is enlarged.
The width of the angular spectrum is defined by the attitude of a wavelength to sizes of
the hole and dependent upon the direction of the spreading wave, falling on the screen.
The last remark pertains to distance between the ffs. To tell the truth, the wider the ffs are
located on the belly, the clearer the lower harmonics stand out and the violin speaks in a
bass voice. If for a violin such an effect can be considered as a defect, then for a viola and
a cello a deeper sound with shortened model is possible only, when the ffs are located
wider than on the big model. In other words, the low timbre of an instrument depends on
the wide location of the ffs more than on the size of the instrument's body. This principle
was understood by all luthiers of the old time and sons of A.Stradivari had well
assimilated this rule, which their father conceived and carried out his own instruments.
The Museum of Cremona contains A.Stradivari's drawing of the central part of a cello
with the scheme for the location of the ffs. On the back of this sheet of paper his sons,
Francesco and Omobono, have repeated the same design with a modification of the
measurement and the placing of the ffs for the shortened model of the cello. In their
variant the distance between ffs is increased in contrast with the variant of their father by
approximately 15 mm. Shortening the model, Francesco and Omobono tried to maintain
the depth of sound of Antonio's cello.
Chapter two
The Geometric Designing of the Violin
The design phase is largely theoretical. Drawing upon the general fund of violinmaking
knowledge and my own research, I produce a mathematical model of a violin that I think
will meet all of the specifications to study the violin design. My simpler simulation
performed by personal computer consists of geometric models. More advanced
simulation, such as that that emulates the dynamic behavior of this acoustical system, is
usually performed on powerful workstations or on mainframe computers. This simulation
can be useful in enabling observers to measure and predict how the functioning of an
entire system may be affected by altering individual components within that system.
I used patterns of the clothoid to draw the outline of the violin. The clothoid was drawn in
Adobe Illustrator with Spiral tool by the co-ordinates referred to below (Table 1). This
Table is made up at the relative dimensions (a = 1). It is necessary to multiply these
dimensions by the clothoid's scaling factor to draw any given clothoid.

s X Y R
0.00 0.0000 0.0000
10 1000 0005 3.1831
20 1999 0042 1.5915
30 2994 0141 1.0610
40 3975 0334 0.7958
0.50 4923 0647 6366
60 5811 1105 5305
70 6597 1721 4547
80 7228 2493 3978
90 7648 3398 3537
1.00 7799 4383 3183
10 7638 5365 2894
20 7154 6234 2653
30 6386 6863 2449
40 5431 7135 2274
1.50 4453 6975 2122
60 3655 6389 1989
70 3238 5492 1872
80 3336 4509 1768
90 3945 3733 1675
2.00 4883 3434 1592

Table 1. The co-ordinates of the clothoid.

The scroll
I have already mentioned sketches by A.Stradivari and emphasised that they were only a
number of the copies of that principal drawing by which one can retrace the way of his
thinking. And his drawing of the violin scroll is not an exception.
Side by side with this sketch I will analyse scrolls by both A.Stradivari and other Italian
luthiers.
By analysing the outline of a scroll for the violin I can conclude that it was drawn with
two curves: The logarithmic spiral or Bernoulli spiral (Figure 9) and The Cornu spiral
or clothoid.

Figure 9. The Bernoulli spiral.

I draw this spiral in Illustrator with Spiral tool by next parameters:
radius = 16.5 mm, decay = 86%, segments =11, rotate = 90º.
Each segment is a quadrant.
In the process of analysis of different scrolls I will use different initial radii and their
decay.
An algorithm of the geometric reconstruction of Stradivari's sketch of a violin scroll is
shown in Figure 10. I begin with two parallel lines AB and ED, which are the tangents to
the scroll; the first extended meets the surface of the neck. The Bernoulli spiral for
development of the scroll from B to O has parameters: radius = 16 mm, decay = 85%,
segments = 11. The clothoid's scaling factors for the development of other parts of the
scroll are 106, 58 and 51. OC/AD = 82.25 mm/50.83 mm = μ.
Figure 10. The geometric reconstruction of Stradivari's sketch of a violin scroll.

Figure 11. The geometric reconstruction of A.Stradivari's scroll of the violin, 1715.
In Figure 11 one can see that the disposition and sizes of the clothoids are identical with
the previous reconstruction. Here and further I have added one more clothoid a-50, which
shapes a tail of the scroll. OC/AD = μ. The Bernoulli spiral is slightly different from the
previous one, and its parameters are visible in Figure.
Although the scroll of the 'Emperor' violin was made with the same pattern to the
previous one, its outline is slightly different.
The methods of violin analysis, which are chosen by me and which demand
superimposing drawings and photos of the whole instrument as well as its different parts,
have one shortcoming: a photo cannot reproduce a geometrically accurate outline of an
instrument without some distortion. It is clearly visible in the next example (Figure 13),
where I analyze A.Stradivari's scroll of the violin photographed from both sides. Here
one can see not only the difference between the sides made by Stradivari, but the optical
distortion too.
In Figure 14 we can see that the outline of the pegbox has a different shape. Now the
clothoid a-110 for the development of the upper part of the pegbox begins its movement
from the line AB to the volute, repeating the curvature of the box. The back of the
pegbox is drawn with clothoid a-65, which touches with the clothoid a-102, whereas in
the previous example such a junction was impossible.
Figure 12: The geometric reconstruction of A.Stradivari's scroll of the 'Emperor' violin,
1715.
Figure 13. The geometric reconstruction of A.Stradivari's scroll of the violin drawn from
both sides.
Figure 14. The geometric reconstruction of A.Stradivari's scroll of the violin, 1689.
Figure 15. The geometric reconstruction of Guarnerius del Jesu's scroll of the violin,
1725.
In my opinion Guarnerius del Jesu's scroll of the violin (Figure 15) is nearly ideal. I like,
for instance, that all the clothoids begin from lines, which are a framework of my
construction where I use the additional line GH, which is parallel to AB and CD.
Moreover, the distance between all three lines is equal, i.e. AC = CG = 50.8 mm, and
their sizes are in simple proportional relations. The main clothoid a-100 describes the
outer face of the scroll. Just the same clothoid, which begins from line HG, determines
the building of the rear sides of the pegbox. For drawing the higher part of the pegbox I
use the clothoid a-113. Here it begins directly with one of the lines of the framework. The
violin scrolls of Guarnerius with their configuration are closer to the Bernoulli spiral,
than the scrolls of Stradivari.
In general, the configuration of the violin scroll of Guarnerius del Jesu noticeably differs
from Stradivari's one. It reads easily compared to the mismatched scroll outlines with
clothoids which are had with the violins of Stradivari, and Guarnerius.
I emphasize that the clothoid only helps me to describe the configuration of different
parts of a violin and do not confirm that the luthiers of the past used the clothoid as well
as I do.

Figure 16. The geometric reconstruction of Guarnerius del Jesu's scroll of the violin,
1730-30.
I had already noticed that the curvature of the higher part of the pegbox is described by
the clothoid, disposed as toward the volute, and in inverse direction. And though the
rotation of the tangent in the main clothoid a-100, which describes the volute, has another
angle, the second clothoid a-100, beginning from the line HG, is tangential to the first.

Figure 17. The geometric reconstruction of Guarnerius del Jesu's scroll of the violin,
1733.
Figure 18. The geometric reconstruction of Guarnerius del Jesu's scroll of the violin,
1733.
Figure 19. The geometric reconstruction of Guarnerius del Jesu's scroll of the violin.
Figure 20. The geometric reconstruction of Guarnerius del Jesu's scroll of the violin,
1735.
Figure 21. The geometric reconstruction of Guarnerius del Jesu's scroll of the violin,
1740-41.
Figure 22. The geometric reconstruction of Diuseppe Guadagnini's scroll of the violin.
The violin scroll (Figure 22) of Diuseppe Guadagnini (1736-1805) is very interesting
geometrically. The volute has only 9 segments. But one can see how beautifully the
clothoids are disposed on the drawing.
Figure 23. The geometric reconstruction of Stradivari's pattern of viola scroll.
Figure 24. The geometric reconstruction of the scroll of the 'Medicea viola' by
A.Stradivari.
Figure 25. The geometric reconstruction of the scroll of the 'Paganini' viola by
A.Stradivari', 1731.
Figure 26. The geometric reconstruction of the scroll of the 'Gore Booth' cello by
A.Stradivari', 1710.
In spite of the greater difference of the scrolls, which we see in instruments of the Italian
masters, all of these share a similar nature of construction and, to greater or smaller
degrees, repeat the sizes and proportions which I have adduced above.
Clearly it is very easy to make a drawing of a violin scroll with the help of the clothoid
and the Bernoulli spiral. With the desire to create our own original form of the scroll we
need to use these two spirals, varying the rotation of clothoids with different angles and
drawing the Bernoulli spiral with a different degree of reduction of the radius. The main
condition for the production of new variants always must be the logic in the algorithmic
building of the whole drawing; but an aesthetic value of spirals and proportions will help
the artist in this difficult function.
Because the scroll is three-dimensional, we need to analyze the successive widths of the
back of it. A.Stradivari has left a drawing of the rear sides of the scroll, showing the
geometric proportions of the successive widths of the pegbox (Figure 27).

Figure.27. Reconstruction of the A.Stradivari's sketch for the back of the violin scroll.
This geometry is not difficult. The tail of the pegbox is outlined with the radius of the
compasses approximately at 12-13 mm, and the width of the finest place of the scroll is
approximately 11mm. The greater difficulty is presented in the successive widths of the
volute (Figure 28) from the narrowest place (the point M) to the broadest one in the
centre of the volute (the point K), which correspond to the similar points in Figure 15.
Here I present the successive widths of the volute in their radii.

Figure 28. The successive widths of the volute.

0 0 5.5
________________________________
1 5.9 5.53
2 11.8 5.65
3 17.7 5.84
4 23.6 6.10
5 29.5 6.44
6 35.4 6.85
7 41.3 7.32
8 47.2 7.85
9 53.1 8.42
10 59.0 9.03
11 65.9 9.66
12 71.8 10.29
13 77.7 10.90
14 83.6 11.46
15 89.5 11.94
16 95.4 12.30
17 101.3 12.53
18 107.2 12.61
19 112.1 12.75
20 118.0 13.17
21 123.9 13.87
22 130.8 14.85
23 136.7 16.09
 24 142.6 17.53
25 148.5 18.99
26 154.4 20.17
27 160.3 20.63
______________________________
Table 2.
As can be seen from the table, the narrowest place of the scroll is 11 mm, but the broadest
one is 41.26 mm. Of course these precise measurements are not axiomatic; one can use
other sizes, leaving the nature of the successive widths of the volute similar with my
drawing.

Figure 29. A Stradivari: the violin, 1702.

As already seen in the example of the violin scroll, for its construction I use the golden
division, which I defined by the relation between the height of the scroll (50.83 mm) and
the distance between the upper nut of the neck and the centre of the volute (82.25 mm).
The size 50.83 mm is deduced by multiplying the number π (3.14159...) by 1.6180339
cm (golden division - μ).
Thereby, for the design of the violin I will use two moduses: the number π for revealing
the sizes of the main parts of an instrument and the clothoid for the drawing of its outline.
As the modulor I will use the golden division, its derivatives and relations 1/2; 2/3; 3/4;
4/5; 3/5; 5/8; etc.
Henceforth, I will define the main sizes of the violin as a geometric progression of the
number π in the proportional attitude of the golden division. In Figure 29 this progression
is defined by the following lengths: AB = 82.25mm; BC = 133.08mm; AC = 215.33mm;
CD = 348.41mm; AD = 563.74mm. 31.4 mm is the height of the bridge and the height of
the ribs.
A note should be taken that CD (348.41 mm) not the length of the body of the instrument,
but the length of the mould which helps to assemble the ribs. The point E (the internal V
notches) divides CD into two lengths in the following ratio:
2ED : CE = μ(1.6180339...) .
One can find these lengths as follows:
CE = CD: (μ : 2 + 1) = 192.6 mm;
ED = CD: (2μ - 1) = 155.81 mm.
Usually, the length of a neck is measured from the upper nut to the edge of the belly, so I
will define it as the length AB, solving the right-angled triangle ABC, where AC is the
hypotenuse. In Stradivari's time the neck was applied with a slight backward inclination.
In the given picture of the modified A.Stradivari's violin (Figure 30) this angle CAB is
7.5º.
AB = AD (133.08 mm) - BD (3.5 mm - the distance between the edge of the belly and the
block) = 129.58 mm, then AC = AB/cos 7.5º = 130.7 mm

Figure 30. The neck of the modified violin of A.Stradivari.

THE BODY OF THE VIOLIN
For the geometric analysis of the violin body I begin with the design of the ff-holes. I
have chosen this starting point because it determines the size and position of the centre
bout, the so-called waist of the violin.
As was shown above (Figure 29), we find the location of the internal V notches of the ffs
on the line E. Then the measure of the instrument (the distance between the internal
notches of the ffs and the upper edge of the belly) is 192.6mm + 3.5mm = 196.1mm
when the length of the body is 355.41mm (348.41mm + 3.5mm + 3.5mm). Below we
adduce the summary table of the length and the measure of the instruments, made by
different masters of Italy (the table 3).
v
____________________________________________________
Master Date Length,mm Measure,mm
____________________________________________________
A.Stradivari 1686 354 195
1688 358 193
1688 357 198
1700 358 197
1705 351 190
1706 356 196
1707 356 196
1708 353.5 195
1708 362 196
1710 358 197
1711 359 195
1716 354.5 192
1718 350 188
no date 354 196
after 1727 357.5 195
after 1727 356.5 195
 1736 358 198
N.Amati 1658 356 195
1663 354.5 197
no date 353 196
1678 351 190
A.&J.Amati 1628 352 196
1629 351 195
Guarnerius del Jesu "Ysayë" 357.5 196
1742 350 188
A.Guarnerius 1665 353 193
_______________________________________________
Table 3.
THE ff-HOLES
I will begin the analysis of sizes, configuration and locations of the ffs with drawings
which were made by A.Stradivari (Figure 31). The lines and arcs of a circle, which we
see in the drawing, are only an orientation for carrying the ffs from the paper on to the
belly. We don't know how Stradivari drew the ffs, though he has left the wooden patterns
of ffs for all models of his instruments. And they are only the copies of a drawing which,
if it only existed, would throw light on the 'mystery' of the creative process not only of
Stradivari, but also of other luthiers of the past.
I shall demonstrate my study of the ffs design on the several drawings "step by step". I
think it is of very suitable to primary familiarization with a train of my thought. The
further examples, describing both the ffs of A.Stradivari and other masters, I shall already
give only on one drawing.
Figure 31. Reconstruction of the A.Stradivari's sketch for the position of the ff holes in
violins (the "G" model).
Figure 32. The geometric reconstruction of the ff holes in violins (the "G" model).
First, I have elaborated the diameters of the eyes (10 mm for lower eyes and 7.5 mm for
upper one) and radius of arc (with the centre in the lower eye), which gets through the
centre of the upper eye and point E (61mm). Hereinafter I have drawn two parallel lines
A¹A² and F¹F², which are coincident to the lines of the bridge and the narrowest place of
the violin's centre bouts. The lines AB and A¹B¹ , which are also parallel to each other
and perpendicular to the previous lines, are the tangent to the centre bouts and internal
surfaces of the lower eyes.
With radii 38mm and 61mm draw the arcs of a circle centre E, going through the centers
of the upper and lower eyes accordingly. The smaller circumference also goes through
the centre of the ffs and touches with the line F¹F². The ratio between both
circumferences is: 61mm/38mm = 1.61, very close to the golden section.
Since the ratio between both circumferences is much close to the golden division, I try to
place the isosceles golden triangle (the angle at base is 72º) on my drawing. Two corners
are rested in the centers upper and lower eyes, but the third one - on the point E.
Certainly, one can see that the corners of the triangle do not touch the correspondent
points exactly, but this may be explained by the discrepancies of Stradivari's drawing at
compasses, whereas I used the computer.
I have added three lines: HJ, LM and PN. The rectangle GHJK has the sides in ratio 3/2,
since 103.7mm/69.1mm = 1.5 and LMNP is a square with side 38.8mm.
I draw the ff-holes with three clothoids: a-22; a-29.3; a-39.1. Since the sizes of the eyes
have a ratio 3/4, the sizes of the clothoids have the same ratio:
22/29.3 = 3/4; 29.3/39.1 = 3/4; 22/39.1 = 9/16, i.e. 3²/4².
The positioning of the clothoids is clearly visible on the drawing. The clothoids a-22 (the
inner face of the upper connecting arm of ffs) and a-29.3 (the outer face of the lower
connecting arm of ffs) are inscribed in the circle of the eyes; a-39.1 begins from the
internal V notch, touches the lines HJ and is inscribed in the a-29.3.
Stradivari connected the upper eye to the lower one by means of paper templates, which
reproduced the ffs. Often in the cutting he did not follow the outline completely. It is
interpreted by both the acoustic task (when the master works at the ffs, correcting the
sound quality) and a peculiarity of the paper template. It consists of three parts: a long
body and two small tails. The template has a rather broad crosspiece between these parts,
but in the cutting of the ffs a cut remains very narrow in this place. In Figure 33 I have
shown this difference by unbroken and dotted lines.
Below I adduce the analysis of the Stradivari's ffs, made from photographs of his violins.
I have the same problems with optical distortion, as before when analysing the violin
scroll, but the main geometric ideas, it seems to me, will be made obvious.
Figure 33. The ff-holes.
Figure 34. A.Stradivari: detail of the belly of the violin, 1702.
In Figure 34 I use only two clothoids: a-25 for the upper eye and a-37 for the lower one.
It is noticeable that the line of the waist (F) is a tangent to the small circle (r = 36.5mm)
and crosses the greater circle (r¹ = 62.8 mm) at the point of the narrowest place of the
violin's centre bouts. The diameters of the eyes are 10 mm and 6 mm.
In Figure 35 I analyze the central part of the "Emperor" violin. As in the previous violin,
the line F¹F² is a tangent to the small circle (r = 38 mm) and intersects the greater circle
(r¹ = 62.8 mm) at the narrowest place of the violin's centre bouts. The small circle (r = 38
mm) gets through the upper eyes, and left f-hole, but it only touches the right f-hole. This
can be explained by asymmetrical ffs, rather than optical distortion, since both lower eyes
are symmetrical. The golden triangle rests one of its own angles to the point E, but other
two slightly do not comply with the centers of eyes. I drew the ffs with two clothoids: a-
28 and a-34.7. And an though analysis of the patterns with corners will be adduced later,
I have taken this opportunity to show the location of the clothoids in the centre bout (C-
bout), which are absolutely symmetrical and described by the clothoids a-47.7 and a-
71.7.
Figure 35. The geometric reconstruction of A.Stradivari's ffs of the 'Emperor' violin.
Figure 36. The geometric reconstruction of A.Stradivari's ffs of the violin.
Figure 36 shows the analysis of another Stradivari violin. Here, as we see, the centre
bouts are asymmetrical. The lines of both bouts so different that I drew them with
different clothoids, whose sizes are seen on the drawing.
The geometry of these ffs is completely subordinated to a law of the golden division, for
instance: two circles with radii 65.38 mm and 40.4 mm (65.38/40.41 = 1.618...), the
golden triangle PES with sides 65.38 mm and 40.41 mm and the golden rectangle PNMV
with sides 31mm and 50mm. The ffs were drawn with two clothoids: a-31 and a-37.
Figure 37. Guarnerius del Jesu: detail of the belly of the violin, 1733.
Figure 38. The geometric reconstruction of A.Stardivari's ffs for a viola.
In Figure 37 I have analyzed the violin of Guarnerius del Jesu by the same method. I used
the golden rectangle with sides 31.4mm and 50.8 mm and the golden triangle, but as can
be seen from drawing, not all of their angles have complied with the centres of the eyes. I
have tried to dispose the clothoids on the right f-hole unsuccessfully - they obviously do
not comply with the lines of the f-hole. On the left f-hole, instead of a clothoid, I have
drawn the circles (D = 17.6 mm and 22.5 mm), which comply with the f-hole enough
well.
Figure 39. The geometric reconstruction of A.Stardivari's ffs for a cello.
THE VIOLIN PATTERN
The term violin pattern refers the internal part of the body of the instrument, comprising
the ribs and all six blocks.
Because the composite position of the corners of the pattern is always co-coordinated
with the scroll in the violins of A.Stradivari, in my first drawing (Figure 41), I have
shown the whole instrument. Additionally this is only a theoretical scheme. Its use is for
the purposes of geometric analysis.
There are many versions about the role of the violin waist, the majority of these
concerning its practical functions, allowing a performer to play the violin with a bow
comfortably. I hold the opinion, that the waisted musical instruments must be considered
as a combination of two resonators united together. In all times masters made musical
instruments both with one, and with two resonators, and it did not matter whether it was
plucked musical instrument or a bowed one.
Below I produce the drawings of three plucked instruments with two resonators (Figure
40).
Figure 40. The plucked musical instruments. 1. The South Indian vina, which has two
resonators made from a pumpkin. 2. The sitar, which has two resonators made from a
pumpkin. 3. The waisted tar of Turkey.
As a rule the resonators are made from different sizes: the lower one is greater, but the
upper one is smaller. Hereby, I will consider the violin body as a waisted double
resonator. Moreover, the upper and lower bouts are of two bulbs united together.
When analyzing the violin moulds of A.Stradivari I will use clothoids of different scales
proceeding from the sizes of the moulds themselves. The sizes and location of the widest
places of the lower and upper bouts are defined by both the scales of the clothoids, and by
the angle of their rotation. For the lower bout this angle is KMP, where MP is the tangent
(X) of the clothoid, but for the upper bout this angle is JST, where ST is also the tangent
of the clothoid.
The location of the lower corners of the violins (to put it more exactly, corner blocks,
rather then the belly itself) is found at the intersection of three lines: G¹A¹, which connect
the widest place of the lower bout with the eye of the scroll; MD, being the tangent (X) of
the clothoid; GH is some straight line, having some angle with CD. For the upper corners
these lines are G¹A¹, ST and FH accordingly.
I will draw these corners with small clothoids, which must touch the proper bulb and get
through points T or P.
Figure 41. The scheme of analysis of geometry of a violin.
The geometric reconstruction of Nicolo Amati's mould 'MB' for a violin (Figure 42). The
length of the pattern (CD) is 343 mm, the width of the upper bout (F¹F¹) is 155 mm, the
width of the lower bout (G¹G¹) is 193.2 mm, and the waist is 101.5 mm. The two widths
are in the ratio of 5 to 4 (G¹G¹/F¹F¹ = 5/4).
The point M, a root of the lower bulb, divides the length of the pattern into two lengths in
the ratio of 2/μ(1.2360678...), then MD = 189.6 mm and CM = 153.4 mm.
Having constructed the lower bulb I used the clothoid a-200. The root of the clothoid lies
at the point M, touches the C-bout and is inserted in the lower bout of the mould up to the
end block. Moreover, the angle EMP = 58º.
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-130 so that it is a tangent to the horizontal line D, and
segments of both clothoids a-130 and a-200 are joined in such a way that the curvature is
continuous throughout.
As it can be seen from the drawing, the lower bout was composed symmetrically.
The upper bout is drawn with the clothoid a-185. Here I do not find any mathematical
regularity in the position of the point S, but will only indicate that CS = 176 mm, and SD
= 167 mm. The angle of the rotation of JST = 51.83º. I have already indicated that this
angle liesthe clothoid, i.e. in the base of the Great Pyramid and is used when building
the golden triangle.
I finish the design of the upper bout with the clothoid a-96.5
As can be seen from the drawing, the upper bout is not symmetrical at all. If the left
clothoid lies in the point S by its root, the right one is shifted to the right a little and has
the smaller angle of rotation (51º).
Before analyzing the geometric position of the corners, we must obtain the points F and
G, which correspond to the widest places of the upper and lower bouts accordingly. I do
it by a simple measurement and define that CF = 64 mm and GD = 67 mm. The length
FG (212 mm) and the length of the pattern (343 mm) are in the ratio of the golden
division.
The line, connecting the widest place of the lower bout with the eye of the scroll, goes
through the lower and upper corners of a violin pattern (G¹A¹). I have only to draw
another two lines to indicate the exact location of these corners.
Draw the line from the point F through the upper corner and get FH, which is at an angle
of 54º to CD. Remember this angle participates in building of the golden triangle. The
third line, which is the tangent of the clothoid a-185, passes very close to point T.
The lower corner is found by a similar way, where the tangent of the clothoid a-200 gets
through the point P precisely and the line GH makes an angle of 51.83º with CD.
The configuration of the corners and C-bout is drawn with the clothoids a-66 and a-44,
which are in the ratio of 3/2.
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the next ratios: CJ(117.5 mm)/JK(90.2 mm) = 4/3;
JK/KD(135.3mm) = 2/3. The centre of circle, getting through all corners lies in point S;
diameter = 185 mm.
Figure 42. The geometric reconstruction of Nicolo Amati's pattern 'MB' for a violin.
The geometric reconstruction of A.Stradivari's pattern "PG"-1689 for a violin (Figure
43). The length of the pattern (CD) is 348 mm, the width of the upper bout (F¹F²) = 161
mm, the width of the lower bout (G¹G²) is 200 mm, and the waist is 103 mm. Moreover
G¹G²/ F¹F² = 1.24 (that is close to the ratio of 5/4 = 1.25).
The point M, a root of the lower bulb, divides the length of the pattern into two lengths in
the ratio of 2/μ (1.2360678...), then MD = 192.4 mm and CM = 155.6 mm.
Having constructed the lower bulb I used the clothoid a-200. The root of the clothoid lies
in the point M, touches the C-bout and is inserted in the lower bout of the mould up to the
end block. Moreover, the angle KMP = 60º. Remember that the angle 60º occurs in the
right-angled triangle with angles of 90º, 60º and 30º in the ratios of 3:2:1.
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-128.5 so that it is a tangent to the horizontal line D, and
segments of both clothoids a-128.5 and a-200 are joined in such a way that the curvature
is continuous throughout.
As can be seen from the drawing, the lower bout was composed symmetrically.
The upper bout is drawn with the clothoid a-185. CS = 176 mm (corresponding to the
previous mould of N. Amati) and SD = 172 mm. Since in this pattern the widest place of
the upper bout is more, than that of Amati's, the angle of the rotation of the clothoid is
increased, i.e. the angle JST = 54º. I finish the design of the upper bout with the clothoid
a-94.3.
As can be seen from the drawing, the upper bout is symmetrical.
I obtain the points F and G by simple measurement and define that CF = 66.1 mm and
GD = 75.3 mm. As it can be seen from this construction, Stradivari has enlarged the
length of the pattern by increasing the lower bout in size. The length of the pattern (348
mm) and the length FG (206.6 mm) are in the ratio of 1.68.
The line, connecting the widest place of the lower bout with the eye of the scroll, goes
through the lower and upper corners of the violin pattern (G¹A¹).
Draw the line from the point F through the upper corner and get FH, which makes an
angle of 54º with CD. The third line, which is a tangent of the clothoid a-185, goes
through the point T exactly.
The lower corner is found by a similar way, where the tangent of the clothoid a-200 goes
through the point P exactly, and the line GH makes an angle of 54º with CD.
The configuration of the corners and C-bout are drawn with the clothoids a-66 and a-44,
which are in ratio of 3/2.
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the next ratios: CJ(121.3 mm)/JK(86.5 mm) = 7/5;
KD(140mm)/JK = μ (golden division). The centre of circle, getting through all corners
lies in point W; diameter = 188 mm.
Figure 43. The geometric reconstruction of A.Stradivari's pattern 'PG' for a violin.
Figure 44. Superimposing the drawing of Nicolo Amati's pattern 'MB' for a violin
(dashed lines) and the drawing of A.Stradivari's mould 'PG' for a violin.
The geometric reconstruction of A.Stradivari's pattern "SL"-1691 for a violin (Figure 45).
The length of the pattern (CD) is 350 mm, the width of the upper bout (F¹F²) = 154.5
mm, the width of the lower bout (G¹G²) is 195.5 mm, and the waist is 100 mm. Moreover
G¹G²/ F¹F² = 1.27 (that is close to the ratio of 5/4 = 1.25).
The point M, a root of the lower bulb, divides the length of the pattern into two lengths in
the ratio of 2/μ (1.2360678...), then MD = 193.5 mm and CM = 156.5 mm.
Having constructed the lower bulb I used the clothoid a-202. The root of the clothoid lies
at the point M, touches the C-bout and is inserted in the lower bout of the mould up to the
end block. Moreover, the angle KMP = 58.7º (that is slightly less than 60º).
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-128.5 so that it is a tangent to the horizontal line D, and
segments of both clothoids a-128.5 and a-200 are joined in such a way that the curvature
is continuous throughout.
As can be seen from the drawing, the lower bout is composed symmetrically.
The upper bout is drawn with the clothoid a-185.4. CS = 178 mm, and SD = 172 mm.
Since in this pattern the upper bout is a little more extended than that of the previous
pattern, the angle of rotation of the clothoid is decreased, i.e. the angle JST = 51.83º. I
finish the design of the upper bout with the clothoid a-100. As it can be seen from the
drawing, the upper bout is composed symmetrically.
I obtain the points F and G by simple measurement and define that CF = 65 mm and GD
= 73.4 mm. The length of the pattern (350 mm) and the length FG (211.6 mm) are in the
ratio of 1.68.
The line, connecting the widest place of the lower bout with the eye of the scroll, goes
through the lower and upper corners of the violin pattern (G¹A¹).
Draw a line from the point F through the upper corner and get FH, which makes an angle
of 54º with CD. The third line, which is the tangent of the clothoid a-185, goes through
the upper corner of the pattern exactly.
The lower corner is found by a similar way, where the tangent of the clothoid a-202 goes
through the point P exactly, and the line GH makes an angle of 51.83º with CD.
The configuration of the corners and C-bout are drawn with the clothoids a-66 and a-44
(which are in ratio of 3/2).
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the following ratios:
CJ(118.3 mm)/JK(90 mm) = 4/3; JK/KD(141.7mm) = 0.635.
If we compare this pattern with the previous one, it is obvious that Stradivari has enlarged
the length of the pattern by increasing the centre part. The centre of circle, getting
through all corners lies in point S; diameter = 188 mm.
Figure 45. The geometric reconstruction of A.Stradivari's pattern "SL"-1691 for a violin.
Figure 46. Superimposing the drawing of A.Stradivari's pattern 'PG' for a violin (dashed
lines) and the drawing of A.Stradivari's mould 'SL' for a violin.
The geometric reconstruction of A.Stradivari's pattern "B"-3/6/1692 for a violin (Figure
47). The length of the pattern (CD) is 353.5 mm, the width of the upper bout (F¹F²) is
154.5 mm, the width of the lower bout (G¹G²) is 194.8 mm, and the waist is 102 mm.
Moreover G¹G²/ F¹F² = 1.26 (that is close to the ratio of 5/4 = 1.25).
The point M, a root of the lower bulb, divides the length of the pattern into two lengths in
the ratio of 1.36, then MD = 203.8 mm and CM = 149.7 mm.
Stradivari again enlarges the previous model by increasing the lower bout. Having
constructed the lower bulb I used the clothoid a-214. Moreover, the angle KMP = 55.6º
much less than 60º, since the bout is extended downwards.
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-130 so that it is a tangent to the horizontal line D, and
segments of both clothoids a-130 and a-214 are joined in such a way that the curvature is
continuous throughout.
As can be seen from the drawing, the lower bout was composed a little asymmetrically.
The upper bout is drawn with the clothoid a-184.5. CS = 181.3 mm, and SD = 172.2 mm.
Since in this pattern the upper bout is a little extended in contrast with the previous
pattern, the angle of the rotation of the clothoid is decreased, i.e. the angle JST = 50.5º. I
finish the design of the upper bout with the clothoid a-94. As can be seen from the
drawing, the upper bout was composed asymmetrically.
I obtain the points F and G by simple measurement and define that CF = 66.2 mm and
GD = 76.5mm. The length of the pattern (353.5 mm) and the length FG (210.8 mm) are
in the ratio of 1.68.
The line, connecting the widest place of the lower bout with the eye of the scroll, goes
through the lower and upper corners of the violin pattern (G¹A¹).
Draw the line from the point F through upper corner and get FH, which makes an angle
of 54º with CD. The third line, which is the tangent of the clothoid a-194.5, goes through
the upper corner of the pattern exactly.
The lower corner is found by a similar way, where the tangent of the clothoid a-214 goes
through the point P exactly, and the line GH makes an angle of 51.83º with CD.
The configuration of the corners and C-bout are drawn with the clothoids a-60 and a-43
(which are in ratio of 7/5).
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the next ratios: CJ(119 mm)/JK(89.6 mm) = 4/3;
KD(144.9mm)/JK = μ.
If we compare this pattern with previous one, it is seen that Stradivari has enlarged the
length of the pattern by increasing the lower bout. The centre of circle, getting through all
corners lies in point W; diameter = 185 mm.
Figure 47.The geometric reconstruction of A.Stradivari's pattern "B"-3/6/1692 for a
violin.
Figure 48. Superimposing the drawing of A.Stradivari's pattern 'SL' for a violin (dashed
lines) and the drawing of A.Stradivari's mould "B"-3/6/1692 for a violin.
The geometric reconstruction of A.Stradivari's pattern "B"-6/12/1692 for a violin (Figure
49). The length of the pattern (CD) is 347.5 mm, the width of the upper bout (F¹F²) is 154
mm, the width of the lower bout (G¹G²) is 195 mm, and the waist is 102 mm. Moreover
G¹G²/ F¹F² = 1.26 (that is close to the ratio of 5/4 = 1.25). In this mould Stradivari retains
the general width of the previous model, but makes it shorter, approaching in size to the
pattern 'PG'.
The point M, a root of the lower bulb, divides the length of the pattern into two lengths in
the ratio of 1.32, then MD = 197.9 mm and CM = 150.2 mm.
Having constructed the lower bulb I used the clothoid a-207.5. Moreover, the angle KMP
is increased, in contrast with the previous pattern, up to 56.7º
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-129 so that it is a tangent to the horizontal line D, and
segments of both clothoids a-129 and a-207.5 are joined in such a way that the curvature
is continuous throughout.
As can be seen from the drawing, the lower bout was composed a little asymmetrically.
The upper bout is drawn with the clothoid a-185. CS = 176.3 mm, and SD = 171.8 mm.
Since in this pattern the upper bout is not extended as much as in the previous pattern, the
angle of the rotation of the clothoid is increased, i.e. the angle JST = 51.83º. I finish the
design of the upper bout with the clothoid a-94.8. As can be seen from the drawing, the
upper bout was composed asymmetrically too.
I find the points F and G by simple measurement and define that CF = 64.7 mm and GD
= 73.6mm. The length of the pattern (347.5 mm) and the length FG (209.8 mm) are in the
ratio of 1.66.
The line, connecting the widest place of the lower bout with the eye of the scroll, goes
through the lower and upper corners of the violin pattern (G¹A¹).
Draw the line from the point F through the upper corner and get FH, which makes an
angle of 54º with CD. The third line, which is the tangent of the clothoid a-185, goes
through the upper corner of the pattern.
The lower corner is found by a similar way, where the tangent of the clothoid a-
207.5goes through the point P, and the line GH makes an angle of 51.83º with CD.
The configuration of the corners and C-bout are drawn with the clothoids a-66 and a-44
(which are in ratio of 3/2). The centre of circle, getting through all corners lies in point S;
diameter = 185 mm.
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the following ratios:
CJ(117.1 mm)/JK(88.2 mm) = 4/3; KD(142.7mm)/JK = μ
Figure 49.The geometric reconstruction of A.Stradivari's pattern "B"-6/12/1692 for a
violin.
Figure 50. Superimposing the drawing of A.Stradivari's pattern "B"-3/6/1692 for a violin
(dashed lines) and the drawing of A.Stradivari's mould "B"-6/12/1692 for a violin.
The geometric reconstruction of A.Stradivari's pattern "S"-1703 for a violin (Figure 51).
The length of the pattern (CD) is 345 mm, the width of the upper bout (F¹F²) is 155 mm,
the width of the lower bout (G¹G²) is 195 mm, and the waist is 100 mm. Moreover G¹G²/
F¹F² = 1.26.
The point M, a root of the lower bulb, divides the length of the pattern into two lengths in
the ratio of 1.22, then MD = 189.7 mm and CM = 155.3 mm.
To construct the lower bulb I use the clothoid a-199.3. Moreover, the angle KMP, in
contrast with the previous pattern, is increased up to 59º.
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-128.5 so that it is a tangent to the horizontal line D, and
segments of both clothoids a-128.5 and a-199.3 are joined in such a way that the
curvature is continuous throughout.
As can be seen from the drawing, the lower bout was composed symmetrically.
The upper bout is drawn with the clothoid a-185. CS = 175.9 mm, and SD = 169.1 mm.
The angle of the rotation of the clothoid, i.e. the angle JST = 51.83º. I finish the design of
the upper bout with the clothoid a-94.3.
As can be seen from the drawing, the upper bout was composed little asymmetrical.
I obtain the points F and G by simple measurement and define that CF = 62 mm and GD
= 70 mm. The length of the pattern (345 mm) and the length FG (213.2 mm) are in the
ratio of μ.
The line, connecting the widest place of the lower bout with the eye of the scroll, goes
through the lower and upper corners of a violin pattern (G¹A¹).
Draw a line from the point F through the upper corner and get FH, which makes an angle
of 51.83º with CD. The third line, which is a tangent of the clothoid a-185, gets through
the upper corner of the pattern.
The lower corner is found by a similar way, where the tangent of the clothoid a-199.3
goes through the point P, and the line GH makes an angle of 51.83º with CD. Thereby,
for the first time the triangle FGH is seen to be isosceles and it completely repeats the
geometry of the Great Pyramid.
The configuration of the corners and C-bout are drawn with the clothoids a-60 and a-42.
The centre of circle, getting through all corners lies in point S; diameter = 185 mm.
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the following ratios:
CJ(119.5 mm)/JK(88 mm) = 1.36; KD(137.5mm)/JK = 1.56.
Figure 51. The geometric reconstruction of A.Stradivari's pattern "S"-1703 for a violin.
Figure 52. Superimposing the drawing of A.Stradivari's pattern "B"-6/12/1692 for a
violin (dashed lines) and the drawing of A.Stradivari's mould "S"-1703 for a violin.
The geometric reconstruction of A.Stradivari's pattern "P"-1705 for a violin (Figure 53).
The length of the pattern (CD) is 348 mm, the width of the upper bout (F¹F²) is 161 mm,
the width of the lower bout (G¹G²) is 200 mm, and the waist is 102 mm. Moreover G¹G²/
F¹F² = 1.24.
The point M, a root of the lower bulb, divides the length of the pattern into two lengths in
the ratio of 1.22, then MD = 190 mm and CM = 158 mm.
In the construction of the lower bulb I use the clothoid a-200. Moreover, the angle KMP,
in contrast with the previous pattern, is increased up to 60º.
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-128.4 so that it touches the horizontal line D, and segments
of clothoids a-128.4 and a-200 are joined in such a way that the curvature is continuous
throughout.
As can be seen from the drawing, the lower bout was composed symmetrically.
The upper bout is drawn with the clothoid a-182. The point S divides the length of the
pattern in half so CS = SD = 174 mm. The angle of the rotation of the clothoid, i.e. the
angle JST = 55º. I finish the design of the upper bout with the clothoid /b>a-106.
As can be seen from the drawing, the upper bout was composed a little asymmetrically.
I obtain the points F and G by simple measurements and define that CF = 65mm and GD
= 73 mm. The length of the pattern (348 mm) and the length FG (210mm) are in the ratio
of 1.657.
The line, connecting the widest place of the lower bout with the eye of the scroll, goes
through the lower and upper corners of the violin pattern (G¹A¹).
Draw the line from the point F through the upper corner and get FH, which makes an
angle of 54º with CD. The third line, which is a tangent of the clothoid a-182, goes
through the upper corner of the pattern.
The lower corner is found by a similar way, where the tangent of the clothoid a-200 goes
through the point P, and the line GH makes an angle of 54º with CD. Thereby, the
triangle FGH can be seen as isosceles again, but with other angles that correspond to the
golden triangle.
The configuration of the corners and C-bout are drawn with the clothoids a-66 and a-44
(which are in ratio of 3/2). The centre of circle, getting through all corners lies in point
W; diameter = 190 mm.
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the following ratios:
CJ(120 mm)/JK(90 mm) = 1.36; KD(138mm)/JK = 1.53.
Figure 53. The geometric reconstruction of A.Stradivari's pattern "P"-1705 for a violin.
Figure 54. Superimposing the drawing of A.Stradivari's pattern "S"-1703 for a violin
(dashed lines) and the drawing of A.Stradivari's mould "P"-1705 for a violin.
The geometric reconstruction of A.Stradivari's pattern "G"-1715 for a violin (Figure 55).
The length of the pattern (CD) is 354 mm, the width of the upper bout (F¹F²) is 161.5
mm, the width of the lower bout (G¹G²) is 201 mm, and the waist is 103 mm. Moreover
G¹G²/ F¹F² = 1.245.
The point M, a root of the lower bulb, divides the length of the pattern into two lengths in
the ratio of 1.27, then MD = 198 mm and CM = 155.9 mm.
Having constructed the lower bulb I used the clothoid a-207.7. Moreover, the angle KMP
is 59º.
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-124.5 so that it is tangential to the horizontal line D, and
segments of both clothoids a-124.5 and a-207.7 are joined in such a way that the
curvature is continuous throughout.
As can be seen from the drawing, the lower bout was composed symmetrically.
The upper bout is drawn with the clothoid /b>a-185. The point S divides the length of the
pattern in half so CS = SD = 177 mm. The angle of the rotation of the clothoid, i.e. the
angle JST = 54º. I finish the design of the upper bout with the clothoid a-124.5.
As can be seen from the drawing, the upper bout was composed a little asymmetrically.
I obtain the points F and G by simple measurement and define that CF = 65.6 mm and
GD = 73.2 mm. The length of the pattern (354 mm) and the length FG (215.5 mm) are in
the ratio of 1.64.
The line, connecting the widest place of the lower bout with the eye of the scroll, goes
through the lower and upper corners of the violin pattern (G¹A¹).
Draw the line from the point F through the upper corner and get FH, which makes an
angle of 54º with CD. The third line, which is a tangent of the clothoid a-185, goes
through the upper corner of the pattern.
The lower corner is found by a similar way, where the tangent of the clothoid a-207.7
goes through the point P, and the line GH makes an angle of 51.83º with CD.
The configuration of the corners and C-bout are drawn with the clothoids a-66 and a-44
(which are in ratio of 3/2). The centre of circle, getting through all corners lies in point S;
diameter = 190 mm.
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the following ratios:
CJ(121.4 mm)/JK(89.2 mm) = 1.36; KD(144.3mm)/JK = μ
Figure 55. The geometric reconstruction of A.Stradivari's pattern "G"-1715 for a violin.
Figure 56. Superimposing the drawing of A.Stradivari's pattern "P"-1705 for a violin
(dashed lines) and the drawing of A.Stradivari's mould "G"-1715 for a violin.
The summary table of the main sizes of the violin patterns
of N. Amati and A.Stradivari.
Table 4.
Above I introduce the summery table of the main sizes of the violin patterns of N.Amati
and A.Stradivari (the table 4). The upper row of letters represents the marks made by
them on their own moulds. "MB" is the mould of N.Amati; "B¹" is the mould of
Stradivari made on 3/6/1692; "B²" is the mould of Stradivari made on 6/12/1692.
When I spoke of the proportional relations between parts of a violin (Figure 29), I pointed
to the size of the pattern equal to 48.4 mm. A.Stradivari has this size in his two moulds
"PG" and "P". Moreover the mould "P" was made in 1705, when Stradivari begins to
create his best instruments.
In these moulds the ratios between the width of the lower bout and the width of the upper
bout are also equal and approach proportions of 1.2360678 = 2/μ. If we connect the
centers of the bouts with the corners, we will get an isosceles triangle with angles at its
base = 54º. This is only in the patterns "PG" and "P", but in the rest the angle in the lower
bouts is 51.83º. Only one pattern "S" has an isosceles triangle with an angle of 51.83º.
The proportional relations between the sizes of upper, centre and lower parts of the violin
patterns (CJ, JK, KD) are basically repeated from pattern to pattern. In four patterns
"PG", "B¹", "B²" and "G" the ratio of the lower part (KD) to the central one (JK) is the
golden division; although in the rest this proportion is less, but close to μ.
In five patterns 'MB ", "SL", "B¹", "B²" and "P" the ratio of the upper part (CJ) to the
central is 4/3, in the rest this proportion is, or is close to 7/5.
Only in two patterns "MB" and "S" the base (FG) of a large triangle, whose sides go
through the corners of the violin pattern, have the ratio of the golden division with the
general length (CD), in the rest this proportion is either a little more or less.
The lower bouts of all patterns irrespective of their sizes are drawn with the clothoid a-
200 or close to it, but the upper bouts are drawn with the clothoid a-185 or close to it.
The angle of the rotation of the lower clothoid is, or is close to 60º. In four patterns
"MB", "SL", "B²" and "S" the upper clothoid is rotated at an angle of 51.83º, in the
pattern "B¹" this angle is a little less, but in the rest this angle is, or is close to 54º. In the
first three patterns the point M divides CD into two lengths in the ratio of 2/μ =
1.2360678, in the rest this ratio is either a little more or less.
***
But now, using the method, which I have used when analysing the violins of Italian
luthiers, I will build my own model of the instrument, which will differ from instruments
already known by us, but will use the main principles of violin design. It will give me the
possibility to show the creative method of approach of a luthier in the art of the violin
design. Whilst maintaining the cardinal principles of design, the master is free enough in
his creative activity to make a unique and inimitable musical instrument in best traditions
of the great masters of Italy.
Designing the upper and lower ovals of violin (Figure 57) I begin from the vertical line
CD = 348.4 mm, which is calculated from the geometric progression of number π in the
ratio of the golden division. The point M is obtained by dividing CD into two lengths
MD = 192.6 mm and CM = 155.8 mm in the ratio of 2/μ.
The tangent of the clothoid a-200 of the lower bout makes an angle of 60º with CD (i.e.
the angle KMP = 60º) and rests its root to point M. The widest place of the lower bout is
200 mm and removed from the lower edge at a distance of 75.3 mm (GD).
The upper bout is built with the clothoid a-185, whose tangent makes an angle of 54º
with CD (i.e. the angle JST = 54º) and rests its root to point S, dividing CD in half (i.e.
CS = SD = 174.2 mm). The widest place of the upper bout is 160 mm and removed from
the upper edge at a distance of 64.1mm (CF).

Figure 57. Designing the upper and lower ovals of violin.

The ratio between the two widths is 5/4, i.e. 200/160 = 5/4. The areas around the end
blocks are completed by the clothoids a-91 and a-130 accordingly.
Hereinafter I connect the edges of widest place of the lower bout with the eyes of the
scroll, as I did earlier, and mark the cross points of these lines and the tangents of the
clothoid with the letters N, L, P and T. Practically I already have marked the positions of
the upper and lower corners of the violin pattern by these points. Let us connect these
points with the centres of the upper (F) and lower (G) bouts accordingly: the angles FGH
and GFH = 54º.
The length FG (209 mm) and the length of the pattern (348.4 mm) are in the ratio of 3/5.
Let us recall that in the past the proportions connected with the Fibonacci series: 1, 2, 3,
5, 8, 13, 21, 34, 55, 89 and so on, were used often. This set of numbers is formed by
adding two previous numbers, the ratio between which approaches the golden division.
So if 3/5 = 0.6 and 55/34 = 1.6176, then 144/89 = 1.618.
Before trying to draw the C-bouts, it is necessary to obtain the size and position of the
waistlines of the violin, which are defined by the sizes and position of the ffs. So we go to
Figure 58.
In Figure 58 I have shown the terminating stage of the modeling of the violin pattern. The
positioning of the point E, through which the line of the bridge and internal V notches of
the ffs are passed, is determined by finding the point of balance of the finished belly.
With same length of belly, but with the different ratio between the widths of the upper
and lower bouts, this point is displaced upward or downward according to this ratio. At
the length of a belly of 35.5 mm and the ratio between the widths of the bouts of 5/4 the
measure of instrument (the distance between the internal notches of the ffs and the upper
edge of the belly) is approximately 196 mm. If an upper bout is slightly wider, and
therefore heavier, this point is positioned upward and the measure can be 193-194 mm.
On the contrary, at a narrow upper bout (or a wide lower bout that is same) the measure
can get to 198 mm.
If the length of pattern is 348.4 mm, the Point E must be at a distance of 192.6 mm from
the upper end-block. I find this point by dividing the length of pattern into two lengths in
the ratio of 2/μ = 1.2360678. Then CE = 192.6 mm and ED = 155.8mm.
The following stage concerns the selection of the ffs' model and size. Each master is free
to choose what he prefers: to copy the ffs of the great masters or to design one's own. I
draw your attention to my own model of the ffs, completely built with the golden
division, the main ideas of which I have demonstrated already when analyzing the ffs of
the great Italian masters. On the horizontal line through the point E I build the golden
rectangle, were the smaller side is perpendicular to the horizontal line and 10π mm
(31.415 mm) in length, while the greater side is 50.83 mm and bisected at the point E. On
the drawing this rectangle is marked by the letters a, b, c and d.
If we connect the points E and b, the length Eb can become the base of an isosceles
triangle with an angle at the base of 72º, which is identified as a golden triangle. This
triangle rests its own angles at the point E, in the centre of the upper eye (the point b) and
in the centre of the lower eye (the point e) respectively.
The length Eb, as a hypotenuse of the right-angled triangle aEb, is 40.41mm. Then Ee =
Eb x μ = 65.38 mm.
Figure 58. Terminating stage of modeling of the violin pattern.
Draw a circle centre E and radius Eb (r - the small radius), and another circle, centre E
and radius Ee (R - the big radius). Let the small circle and CD intersect at f. Draw the
horizontal line through f. Let this line and big circle intersect at g and h. I define the
length of gh, which is the waistline of violin pattern, as follows:
gh = 2√(R²- r²) = 2√(65.38 mm²- 40.41 mm²) = 102.8mm.
Draw the circles centre b (radius of 3.75 mm for the upper eye) and centre e (radius 5 mm
for the lower eye).
The further work of building the ffs and the C-bout with clothoids is similar to my
analyses of the instruments of the Italian masters and is easy to read from the drawing.
Here I have used the clothoids a-40 for the lower eyes and a-30 for the upper eyes, but
the C-bout was drawn with the clothoids a-71 and a-44.
THE VIOLAS AND CELLOS PATTERNS
A.Stradivari made his first mould of a viola contralto in 1672. There was only one
instrument made from this form. All the following instruments came from the mould of
1690 made for the viola contralto Medicea. On that day, October 4th 1690, Stradivari
made two forms: the first mould was for the contralto, and the second one - for the tenor.
The geometric proportions of violas are noticeably differ from the violins' one. To trace
the logic of the master, in changing the proportional relations between the parts of each
instrument, I will do similar collations to those I did with his violins. When comparing
the drawing of the viola moulds with drawings of violin patterns I will reduce them to the
comparable sizes.
In Figure 60 I have superimposed the drawing of the viola mould of 1672 with the
drawing of the violin pattern "MB". Here one can see that Stradivari has enlarged the
volume of the lower bout, leaving a diminished C-bout. If we compare this pattern with
those looked at previously, it can be seen that Stradivari has enlarged the length of the
pattern by increasing the lower bout.
The geometric reconstruction of A.Stradivari's pattern for a viola contralto, 1672 (Figure
59). The length of the pattern (CD) is 403 mm, the width of the upper bout (F¹F²) is 184
mm, the width of the lower bout (G¹G²) is 241 mm, and the waist is 124 mm. Moreover
G¹G²/ F¹F² = 1.31. The point M, a root of the lower bulb, divides the length of the pattern
into two lengths in the ratio of 1.3, then MD = 228 mm and CM = 175 mm. Having
constructed the lower bulb I used the clothoid a-238.5. The root of the clothoid lies at the
point M, touches the C-bout and is inserted in the lower bout of the mould up to the end
block. Moreover, the angle KMP = 60º.
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-148 so that it is a tangent to the horizontal line D, and
segments of both clothoids a-238.5 and a-148 are joined in such a way that the curvature
is continuous throughout. As can be seen from the drawing, the lower bout was composed
a little asymmetrically.
The upper bout is drawn with the clothoid a-213.6, the root of which does not lie on the
line CD. Since in this pattern the widest place of the upper bout to the waist is less than
that of a violin, the angle of rotation of the clothoid is decreased, i.e. the angle JUS¹ =
54º. I finish the design of the upper bout with the clothoid a-108.9.
As can be seen from the drawing, the upper bout is symmetrical. I obtain the points F and
G by simple measurements and determine that CF = 76.7 mm and GD = 85 mm. The
length of the pattern (403 mm) and the length FG (241.3 mm) are in the ratio of 1.67. The
line connecting the widest place of the lower bout with the eye of the scroll goes through
the lower and upper corners of the viola pattern (G¹A¹). Draw the line from the point F
through the upper corner and get FH, which makes an angle of 54.5º with CD.
The lower corner is found by a similar way, where the tangent of the clothoid a-238.5
gets close to the point P, and the line GH makes an angle of 51º with CD.
The configuration of the corners and C-bout are drawn with the clothoids a-69.1 and a-
51.5, and are in ratio of 1.34.
Figure 59. The geometric reconstruction of A.Stradivari's pattern for a viola contralto,
1672.
Figure 60. Superimposing the drawing of Amati's pattern "MB" for a violin (dashed lines)
and the drawing of A.Stradivari's mould for a viola contralto, 1672.
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the following ratios:
CJ(139 mm)/JK(96 mm) = 1.45; KD(168 mm)/JK = 7/4 KD/CJ = 6/5.
The mould for a viola contralto (1690) was made by A.Stradivari after the mould 'PG' for
a violin. So, in Figure 69 I have compared the proportions of these moulds.
The geometric reconstruction of A.Stradivari's pattern for the viola contralto of 1690
(Figure 61). The length of the pattern (CD) is 403 mm, the width of the upper bout (F¹F²)
is 177 mm, the width of the lower bout (G¹G²) is 233 mm, and the waist is 118 mm.
Moreover G¹G²/ F¹F² = 1.32. The point M, a root of the lower bulb, divides the length of
the pattern into two lengths in the ratio of 1.37, then MD = 233 mm and CM = 170 mm.
Having constructed the lower bulb I used the clothoid a-244. The root of the clothoid lies
at the point M, touches the C-bout and is inserted in the lower bout of the mould up to the
end block. Moreover, the angle KMP = 58º.
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-145 so that is a tangent to the horizontal line D, and
segments of both clothoids a-244 and a-145 are joined in such a way that the curvature is
continuous throughout.
As can be seen from the drawing, the lower bout was composed a little asymmetrically.
The upper bout is drawn with the clothoid a-184, the root of which does not lie on the
line CD. Since in this pattern the widest place of the upper bout to the waist is less than
that of a violin, the angle of the rotation of the clothoid is decreased, i.e. the angle FF²S¹
= 41.3º. I finish the design of the upper bout with the clothoid a-104.4.
As can be seen from the drawing, the upper bout is symmetrical.
I obtain the points F and G by simple measurement and determine that CF = 73 mm and
GD = 90mm. The length of the pattern (403 mm) and the length FG (240 mm) are in the
ratio of 1.68.
The line connecting the widest place of the lower bout with the eye of the scroll goes
through the lower and upper corners of the viola pattern (G¹A¹).
Draw the line from the point F through the upper corner and get FH, which makes an
angle of 54º with CD.
The lower corner is found by a similar method, where the tangent of the clothoid a-244
gets close to the point P, and the line GH makes an angle of 51º with CD.
The configuration of the corners and C-bout are drawn with the clothoids a-72.7and a-
48.4, which are in ratio of 3/2.
If the lines TJ and PK through the corners divide the length of the violin pattern into three
parts, I get three lengths in the following ratios:
CJ(134 mm)/JK(98 mm) = 7/5; KD(171 mm)/JK = 1.74; KD/CJ = 1.28.
Figure 61. The geometric reconstruction of A.Stradivari's pattern for a viola contralto of
1690.

Figure 62. Superimposing the drawing of the pattern "PG" for a violin (dashed lines) and
the drawing of A.Stradivari's mould for a viola contralto of 1690.
Figure 63. Superimposing the drawing of the pattern for a viola contralto of 1690 (dashed
lines) and the drawing of A.Stradivari's mould for a viola contralto of 1672.
The geometric reconstruction of A.Stradivari's pattern for the tenor viola of 1690 (Figure
64). The length of the pattern (CD) is 468 mm, the width of the upper bout (F¹F²) is 207
mm, the width of the lower bout (G¹G²) is 257 mm, and the waist is 137 mm. Moreover
G¹G²/ F¹F² = 1.24.
Figure 64. The geometric reconstruction of A.Stradivari's pattern for the tenor viola of
1690.
Figure 65. Superimposing the drawing of the pattern "PG" for a violin (dashed lines) and
the drawing of A.Stradivari's mould for a tenor viola of 1690.
The geometric reconstruction of A.Stradivari's pattern for the 'Duport' cello (Figure 66).
The length of the pattern (CD) is 749 mm, the width of the upper bout (F¹F²) is 340 mm,
the width of the lower bout (G¹G²) is 435 mm, and the waist is 220.5 mm. Moreover
G¹G²/ F¹F² = 1.28.
The point M, a root of the lower bulb, divides the length of the pattern into two lengths in
the ratio of 1.38, then MD = 434.4 mm and CM = 314.6 mm.
Having constructed the lower bulb I used the clothoid a-458 Moreover, the angle KMP is
57.3º.
Similarly one constructs the opposite side of the bulb. The area around the end block is
completed by the clothoid a-272.7 so that it is a tangent of the horizontal line D, and
segments of both clothoids a-272.7 and a-458 are joined in such a way that the curvature
is continuous throughout.
As can be seen from the drawing, the lower bout was composed symmetrically.
The upper bout is drawn with the clothoid a-392.6. The point S divides the length of the
pattern in half so CS = SD = 374.5 mm. The angle of the rotation of the clothoid, i.e. the
angle JST = 54º. I finish the design of the upper bout with the clothoid a-220.
As can be seen from the drawing, the upper bout was composed a little asymmetrically.
I obtain the points F and G by simple measurement and determine that CF = 140 mm and
GD = 161mm. The length of the pattern (749 mm) and the length FG (448 mm) are in the
ratio of 1.67.
The line connecting the widest place of the lower bout with the eye of the scroll goes
through the lower and upper corners of a violin pattern (G¹A¹).
Draw the line from the point F through the upper corner and get FH, which makes an
angle of 53.5º with CD. The third line, which is a tangent of the clothoid a-392.6, goes
through the upper corner of the pattern.
The lower corner is found by a similar method, where the tangent of the clothoid a-458
goes through the point P, and the line GH makes an angle of 49.5º with CD.
The configuration of the corners and C-bout are drawn with the clothoids a-116 and a-85,
and are in ratio of 1.36.
If the lines TJ and PK through the corners divide the length of the cello pattern into three
parts, I get three lengths in the following ratios:
CJ(258 mm)/JK(173 mm) = 3/2; KD(318 mm)/JK = 1.84.
Figure 66. The geometric reconstruction of A.Stradivari's pattern for the 'Duport' cello.
THE ARCHES OF THE BELLY AND THE BACK
The arches of the belly and the back are surfaces. Before giving the rules of building the
arches of the belly and the back we must define what a surface is. The surface is a
continuous extent having only two dimensions (length and breadth, without thickness),
whether plane or curved, finite or infinite. The position of the point on it is defined by
two surface coordinates.
The curved surfaces are subdivided into regular, graphic, topographical, gravitational
and others.
Before defining what type of surfaces we shall rank the arches of the belly and the back,
address to the history of this question.
Many researchers of the creative activity of Italian luthiers developed the topographical
diagrams of the plates' arches to study them. The topographical diagram is marked with
contour lines, joining points of equal height of the curvature. Such a method of analysis
of the violin arches is very good in many events: it can be used for the study of the
deformations occurring in the belly and the back and it can be used during the making of
new plates as a particular case of checking on the symmetry of the curvature.
A.Stradivari himself has left absolutely other method of building of arches. His sixths
(one longitudinal and five transversal guides) give the rule of building for the arches of
the belly and the back and they are an analytical instrument for the study of the different
arches.
Proceeding from the nature of these sixths, the surface of a belly or a back can be defined
as a graphic surface, generated by moving a variable line (five transversal guides) along
a fixed direction (the sixth longitudinal guide) in accordance with the results of
calculations, satisfying assessed conditions. As the conditions in this instance there are a
typical curvature of the longitudinal guide, the horizontal plane and the outline of the
basis of the plate. Five transverse lines, as different positions of the variable generator,
together with a sixth longitudinal line as the guide, form the linear framework of the
surface of the belly or the back.
Though the nature of the arches of the different instruments of A.Stradivari is variable
(because of particularities in the wood), nevertheless it is geometrically analyzable and
can be expressed by a formula, i.e. the curvature of the arches can be described by the
clothoid. Then the generator, and guide of these surfaces will be regular curved lines and
the surface of the belly or the back will be a regular curvilinear surface with a variable
generator. The variability of a generator is assigned by the outline of the plate, the
curvature of the longitudinal guide and the plane of basis.
The initial position of the generator is defined by the geometric centre of the belly or the
back. Proceeding from the specifics of the belly and the back, these have different
curvatures. In Figure 67 I have presented the transverse section of the violin's belly (a)
and the back (b) of A.Stradivari. As can be seen from the drawing, the different positions
of the clothoid on the belly and on the back show the different nature of their curvature.
So the top of the belly has the greater radius of curvature, but closer to edges the radius is
decreased. At the back is quite the reverse - the smaller radius is located on the top, but
closer to edges the radius is increased.
The longitudinal curve (or guide, as defined mathematically) of the belly repeats the
nature of the curvature of the generator (Figure 68, a) i.e. in both cases the clothoid is
disposed from the centre to the edge.
That the guide of the back (Figure 68, b) has a more complex curvature than what is
clearly seen from the drawing (of which I have shown only half). If we dispose the
clothoid in the same direction, as with the belly, the clothoid is deviated from the guide
far enough from the edge, showing hereunder a rather big radius of curvature of the
back's guide beside the edges that correspond to the nature of the generator's curvature.
Thereby, the smallest radius of the back's curvature on the longitudinal line is located not
at its geometric centre (as the generator has), but at the centre of the upper and lower
bouts.

Figure 67. The transverse section of the violin's belly (a) and the back (b) of A.Stradivari.
Figure 68. The longitudinal section of the violin's belly (a) and the back (b) of
A.Stradivari.
In Figure 69 I have drawn the sixths of curvature of the violin belly, which are placed:
A)- at centre of the C-bout; B) - corresponding with the upper corners; C) - at the
maximum width of the upper bout; D) - corresponding with the lower corners; E) - at the
maximum width of the lower bout; F) - corresponding with the upper half of the
longitudinal section; G) - corresponding with the lower half of the longitudinal section.
The distance between the vertical lines, dividing the central guide (A) into even areas, is
6 mm; the height of the arch in the centre is 15.5 mm; the width of the belly here is 112
mm.
The distance between the upper corners (B) is 150 mm, and the height of the arch in this
place is 14.8mm.
The widest place of the upper bout (C) is 167 mm whilst the arch here is 12.0 mm.
The distance between lower corners (D) is 179 mm, and the height of the arch in this
place is 15.3mm. The widest place of the lower bout (E) is 207 mm whilst the arch here is
13.0mm. The longitudal curve (FG) is symmetrical and on my drawing it is divided in
half. The distances between the vertical lines in the guides B, C, D, E, F, and G
correspond to the points of their intersection with the lines of the topographical diagram.
One can see the location of the sixths on the belly in Figure 71, where for greater clarity I
have combined the topographical diagram with the graphic framework. The numbers on
the drawing indicate the heights of the arch of the belly.
The topographical diagram of the belly is shown in Figure 70. The numbers on the
diagram indicate the distances between lines.
In Figure 72 I have drawn the sixths of curvature of the violin back, which are placed: A)
- at the centre of the C-bout; B) - corresponding with the upper corners; C) - at the
maximum width of the upper bout; D) - corresponding with the lower corners; E) - at the
maximum width of the lower bout; F) - corresponding with the upper half of the
longitudinal section; G) - corresponding with the lower half of the longitudinal section.
The distance between the vertical lines, dividing the central guide (A) into even areas, is
6 mm; the height of the arch in the centre is 14.8 mm; the width of the back here is 112
mm. The distance between the upper corners (B) is 150 mm, and the height of the arch
here is 13.4mm.
The widest distance of the upper bout (C) is 167 mm and the height of the arch here is
10.1 mm. The distance between the lower corners (D) is 179 mm, and the height of the
arch here is 14.2mm. The widest distance of the lower bout (E) is 207 mm and the height
of the arch here is 11.0 mm. The longitudal curve (FG) is symmetrical and on my
drawing it is divided in half, and I have added the button to the upper half of the back.
The distances between the vertical lines in the guides B, C, D, E, F, and G correspond to
the points of their intersection with the lines of the topographical diagram. One can see
the location of the sixths on the belly in Figure 81, where for greater clarity I have
combined the topographical diagram with the graphic framework. The numbers on the
drawing indicate the heights of the arch of the belly.
The topographical diagram of the belly is shown in Figure 73. The numbers on the
diagram indicate the distances between lines. Hereinafter in Figures from 75 to 88 I show
the drawings of arches of violas and cellos.

Figure 69. The sixths of the violin belly.

Figure 70. The topographical diagram of the violin belly.
Figure 71. Combining the topographical diagram of the violin belly with the sixths.
Figure 72. The sixths of the violin back.
Figure 73. The topographical diagram of the violin back.
Figure 74. Combining the topographical diagram of the violin back with the sixths.
Figure 75. The sixths of the viola belly.
Figure 76. The topographical diagram of the viola belly.
Figure 77. Combining the topographical diagram of the viola belly with the sixths.
Figure 78. The sixths of the viola back.
Figure 79. The topographical diagram of the viola back.
Figure 80. Combining the topographical diagram of the viola belly with the sixths.
Figure 81. The fifths of the cello belly.

Figure 82. The longitudinal section of the cello belly.

Figure 83. The topographical diagram of the cello belly.
Figure 84. Combining the topographical diagram of the cello belly with the sixths.
Figure 85. The fifths of the cello back.

Figure 86. The longitudinal section of the cello back.

Figure 87. The topographical diagram of the cello back.
Figure 88. Combining the topographical diagram of the cello back with the sixths.
***
The afore-cited analysis of the scrolls, ffs and moulds of the stringed and bowed
instruments allows me to draw a conclusion about the goal-directed usage of some of the
proportional relations and geometric angles by great masters. There is only the need to
settle the problem of clothoid usage by them, in certain patterns, for the drawing of the
contour of the instrument.
Neither A.Stradivari, nor fellow masters used the clothoid in the manner I have. Quite
simply, they could not draw with the precision we can nowadays.
But they could shape it from a string or some other suitable material.
This conclusion is confirmed by the fact that a variety of curved lines, made by Stradivari
are still extant them very close to or exactly modeled on this remarkable mathematical
curve, the clothoid. The small inexactnesses are explained by the fact that Stradivari did
not strive for perfect accuracy when making the drawing; additionally, the string cannot
be as accurate a method as that which can be produced with help of a computer.
In the next chapter I will try, as far as possible, to reproduce the method of Stradivari for
the drawing of all contoured curved lines of the violin, including the scroll, ffs and the
pattern.
Chapter three
The Reconstruction of the Stardivarius Method of Violin Design
In this chapter I will try to show a method of violin design using compasses and a ruler to
determine sizes and proportions only; and a way of drawing contour lines with the help of
strings (a method that could have been used by olden day masters).
I already noted that my designs were produced with the help of two modules: the number
π and the clothoid.
At the turn of the 16th century, when such prominent masters as Stradivari and
Guarnerius were working, mathematics and geometry had reached a stage of rather high
development, and there was decimal calculus around the world, but the inchwas the
fundamental unit of measurement.
It is impossible to talk about the size of one inch at that time with sufficient accuracy
because it has only been since 1959 that the inch has been defined officially as 2.54
centimeters. This unit derives from the old English unce, or ynche, which in turn came
from the Latin unit uncia, which was "one-twelfth" of a Latin foot, or pes. The old
English ynche was defined by King David I of Scotland about 1150 as the breadth of a
man's thumb at the base of the nail. During the reign of King Edward II, in the early 14th
century, the inch was defined as "three grains of barley, dry and round, placed end to end
lengthwise. Sometimes the inch has even been defined as the combined length of 12
poppy seeds.
If at various times the grains or the thumb were used for quickly finding this size, it is
important to search for the true origin of the inch in geometric proportions. I mean the
number π and the golden ratio: 1 in. = πμ/2 cm = 2.5416 cm.
Thereby, the number π is connected with the British inch and the modern metre by means
of the golden division. One can say that the inch, which was used by the ancient masters,
was approximately, as I have indicated in my formula.
Consequently, the ancient constructors of violins did not use the number π as a module,
but one inch instead. Thereby, the main sizes of violin can be defined as a geometric
progression of the British inch in the proportional attitude of the golden division. In
Figure 29 this progression can be defined by the following lengths: AB = 82.25mm
(2in,μ BC = 133.08mm (2inμ²); AC = 215.33mm (2inμ³); etc.
I remind the reader that as the modulor I used the golden division and numerically
simple ratios. Concerning the numerically simple ratios, their usage is very simple and
does not require any special explanations. In Figure 2 I had demonstrated the calibration
of lengths in the ratio of the golden division.
Now I begin to design the violin with the help of spirals, using simplest ways, which
could use the luthiers of the past.
THE SCROLL
Let us begin with the Bernoulli spiral. We can see that because of the small size of the
violin scroll, it is unreasonable to use of the method of drawing known to Vitruvius
[Marcus Vitruvius Pollio, an author of the De architectura libri decem (Ten Books on
Architecture)]. The degree of reduction of the arcs of the circumferences (decay) in his
method does not help us to solve our problems.
For the drawing of the scroll I offer to use the scale compasses (Figure 89), one variety
of which was used by the masters for the fretting out of guitars, lutes, mandolins and so
on.
If one pair of arms has, for instance, a length of 4 inches (101.6 mm), then the other pair
must be shorter by a percentage chosen by the luthier himself.

Figure 89. The scale compasses.

I have chosen the 15% reduction of radius and proceed to draw the violin volute (Figure
90).
A spiral is constructed as a series of circular arcs. I begin by drawing a quadrant of the
circular arc ab centre O and radius of 16 mm with the spread of the big pair of
compasses. Then I turn the compasses over and with the spread of the small pair of
compasses draw a quadrant with a circular arc bc centre x, which rests on the line ob. The
new radius of 13.6 mm has only 85% of the original radius (16 mm). Spread the big pair
of compasses an equal distance to this radius and draw a quadrant with a circular arc cb
centre y with the spread of the small pair of compasses.
Next arc de with the centre at point z also is a quarter of circumference.
The further process of drawing the volute completely repeats the previous one and is
easily understood from the drawing. Naturally we can use the compasses with another
percentage reduction of the radius (as I did when analyzing the violin scrolls).
The quality of the volute is dependant on the accuracy of the present drawing. If I made
my drawing with the help of a computer with a higher degree of accuracy, than masters
of the past, who using only compasses and a ruler, certainly allowed for some inaccuracy,
which, of course influenced the final result.
When I analyze the violin scrolls, certain inexactness in the building of the Bernoulli
spiral by some Italian masters is striking. This inexactness was not so much the effect of
carelessness as non-observance of the main rule for designing the Bernoulli spiral: each
new arc must be reduced by a constant percentage.
Figure 90. The phased drawing of the volute.
Figure 91. The volute of Pietro Guarneri I.
Using the scale compasses, one can allow a certain inaccuracy, but in very small limits.
But if we draw the volute with the usual compasses and produce a reduction 'by eye', then
the spiral although it will be rather graceful (Figure 91), cannot be identified as a genuine
Bernoulli spiral. The parameters of this spiral are as following: arc ab - radius of 27.5
mm; bc - 22.8 mm; cd - 16 mm; de -14.6 mm; ef - 11.3 mm; fg - 9.5 mm; gh - 7.3 mm; hi
- 6.3 mm; ij - 5.4 mm; jk - 4.1 mm; kl - 3.9 mm; lm - 3.1 mm; mn - 2.5 mm. The
percentage reduction of the radius from the greater radius is as follows: 82.8%, 70%,
91%, 77%, 84.3%, 77.4%, 85.3%, 87%, 75.9%, 94%, 80.1%, 80.1%. The average factor
is 82%.
Return to Figure 90. On the last eleventh drawing of my construction the centre of the
volute, which will serve us for the further construction of the violin scroll, is already seen
distinctly. I have only to draw three parallel lines: through the centre of the volute, above
the volute through point a and the lower line, which goes through the point a at a distance
of 2 inches (50.8 mm) from the line c. The centerline is limited by the vertical line at the
distance of 2inμ from the centre of the volute.
The next stage of constructing the scroll is concerned with the drawing of the curved
lines, which I have done with the help of the clothoid in the First Chapter.
To this effect I use the springy string, which can take the shape of a clothoid when we
roll it. Using the rolled string shaped as a loop (Figure 92) I continue to draw the outer
face of the volute. Certainly, we can vary not only the size, but also the configuration of
the curvature, created by a curled string, forcing the string with the thumb of the left hand
towards the nut.

Figure 92. Modeling the external curvature of the volute.

I model the rear sides of the peg box, which has an S-form, in one step, holding the string
in the way shown in Figure 93. As can be seen, the curled string exactly copies the
sidebar of the scroll. It would not be out of place to observe that the nature of the
curvature depends in full measure on the way of holding the string, on the distance
between hands, on the pressure of the right thumb, etc. So the violin scrolls, drawn in this
way, can be different (with a fluent change of the radius), but cannot be attained with the
use of compasses alone.
Figure 93. Modeling the rear sides of the peg box.
The upper part of the peg box is also modeled with the help of a curled string, which can
be held by a master in the way shown in Figure 101. Even A.Stradivari's scrolls have a
different curvature in this part of the peg box. A smaller radius is used in the centre, and
closer to the volute or to the upper nut. So the nature of the curvature is completely
dependent upon the method of holding the string, when the thumb of the right or left hand
plays the leading role. In my photograph, for instance, the right thumb forces the string,
creating the smallest radius of the curvature closer to the nut.
Figure 94. Modeling the upper part of the peg box.
THE VIOLIN PATTERN
The violin moulds, made by A.Stradivari, have several round holes, which help to
assemble the ribs. They correspond with all six blocks. But there are other holes in the
moulds, which were changed into a semicircular shape when cutting the spaces for the
upper and lower blocks. In some moulds such semicircular cuts are placed at the centre of
the space (for instance, the model B, 6/12/1692), some moulds have two semicircular cuts
(the model PG, 1689); one has three cuts (the model P, 1705). I deduce the reason for
these cuts, was that they were former holes; i.e. auxiliary holes when modeling the
contour of the pattern.
When modeling the violin scroll with the help of a springy string, I held the latter with
both hands. As a variant, I have offered to fasten one end of the string by any known
way. Now I will show one of the possible variants of such fastening.
In the hole, drilled in the board of the willow wood for a future pattern, on the axis of
symmetry the violin peg is placed in such a way that the small hole for the string is at the
surface of board exactly. Spring a string from the peg's hole and roll it to determine the
curved shape of the bout. Then trace the curvature of the string with a pencil. As can be
seen in the photograph, the string repeats the contour of the clothoid completely (Figure
95).
Figure 95. Modeling of the upper bout of the violin pattern from the hole on the axis of
symmetry.
It will be useful to notice that the correct angle of rotation of the peg, holding the string,
is very important in this matter. Rotating the peg to one side or another, I hereby change
the curvature of the string and the form of the future pattern too. In my case a string
passes through the peg at an angle of 54º to the axis.
We see that the position of these holes was attained by experiment and not affixed
formula. On the Stradivari's moulds we see the residuum of these holes of different sizes:
from deep cuts (the model B-3/6/1692) to small cuts (the model SL), and also their total
absence. This is explained by the different distance between the hole and the edge of
pattern, whereas the depth of cuts for the blocks is always 15 mm.
On the next photograph (Figure 96) I investigate the possibility of modeling the upper
bout, placing the peg into the hole, which is not located on the axis of symmetry.
Figure 96. Modeling of the upper bout of the violin pattern from the hole not on the axis
of symmetry.
As can be seen, this works with greater success, because in this case the curvature of the
string includes the area of the upper bout, something the clothoid and the previous
experiment with the string did not do. In both cases the string repeats the contour of the
pattern exactly, only right now it passes through the peg at an angle of 60º to the axis.
The left side of the pattern is built as well as the right one. Some asymmetry between the
left and right sides is discernible because it is practically impossible to repeat this
procedure with such a flexible thing, as string. This asymmetry that does not spoil the
general appearance of the instrument at all, in fact it even contributes certain
individuality to the final contour of the violin.
In Figure 97 I demonstrate the modeling of the lower bout of the violin pattern. This
procedure does not differ from the previous work with the upper bout. The angle at which
the string passes through the peg is 62º.
To finish the modeling of bouts at the area of the upper and lower blocks is not laborious
with the use of the same string (Figures 98, 99).
Figure 97. Modeling of the lower bout of violin pattern.
Figure 98. Finishing of the modeling of upper bout.

Figure 99. Finishing of the modeling of lower bout.

The next stage is the determination of the position of the upper and lower corners of the
pattern. Connect the edges of the widest place of the lower bout and the volute of the
scroll with a straight line. Then the lines, passing through the centres of the bouts at
certain angles (as I defined this in the First Chapter), intersect the first line. In the
intersections of these lines drill the holes for the pegs in such a way that the edge of the
peg touches the cross point, and the line through the centre of the bout crosses the hole
strictly at the centre (Figure 100).

Figure 100. The position of the hole for the peg when modeling the C-bout.
The pegs are fixed in such a position that their holes are compliant with the lines, which
pass through the centres of the bouts. The string is placed in the pegs and pulled through
them up to the desired size of the waist (Figure 101). Though in this case the string did
not repeat the contour of Stradivari's pattern, such C-bouts (the same curvature of the
upper and lower corners, and a rounded centre of the waist) are found in the violins of
different masters, as with those of Andrea Amati. I will name this method: modeling with
a free string.
Figure 101. Modeling of the C-bout of the violin pattern with a free string.
I model the Stradivari's C-bout by another way, limiting the excessive concavity of the
string inside of the pattern with help of a stop (Figure 102). Since this stop exerts
pressure on the string, not at the centre, but a little above, the curvature of the upper
corner will be bigger than the curvature of the lower corner. I will name this method:
modeling with a restricted string.
Figure 102. Modeling of the C-bout of the violin pattern with a restricted string.
I model the outer face of the corners with the help of the same pegs, directing the string
along other curved lines, which must touch the bulbs of the corresponded bout (Figures
103).
Figure 103. Modeling the outer face of the corner.
THE FF-HOLES
In Figures 104, 105, 106 and 107 I show four stages of modeling of the ffs. Here I have
placed the pegs into the eyes, but from the back of the board only in order not to overlay
the drawing of the f-hole by the head of the peg. The string, which I use for the modeling
of the f-hole, is thinner than in previous cases.
As to the strings which helped me to draw the violin, I used a piano string for the
modeling of large objects: the bulbs of the bouts and the scroll; a guitar string for the C-
bout and a violin string for the ffs. In short, the smaller the object of modeling, the
thinner the string used.
Figure 104. Modeling of the upper part of the f-hole.

Figure 105. Modeling of the lower part of the f-hole.

Figure 106. Modeling of the inner part of the f-hole's arm.
Figure 107. Modeling of the external part of the f-hole's arm.
THE ARCHES OF THE BELLY AND THE BACK
When speaking of the arches of the belly and the back of A.Stradivari's violins, we noted
that to date two methods of construction of their buildings and geometric analysis is
known. First, using the sixths of the curvature and of course the topographical
diagrams. Both of these ways are very good for copying, but completely unsuitable for
primary modeling. Even my mathematical descriptions of the arches of the belly and the
back, although suitable for the modeling of a new violin, do not answer the most
important question: "How was the primary modeling of the arches of a violin produced
by the great luthiers of 16th -18th centuries?"
After a series of experiments with the string, which has helped me to create the
hypothesis about the designing of the outline of the instrument's body with the help of
some flexible material, and after finding the mathematical relationship between a curled
string and the clothoid, for a long time I searched for anything similar regarding the
arches. In Reality, neither Amati nor Stradivari, nor other luthiers have left us any traces
of this creative process. Common sense tells us that the sixths of the curvature, made by
Stradivari, are nothing more than auxiliary instruments, used by the master for the
transfer of the curvature of some arch on the real violin plate. I hold the view, that there
was a certain object, identical to the arch of the plate, conceived by the master, which
must have been copied in wood by him alone. No doubt he created this object himself, to
meet the requirements of each new idea. It is obvious that this object, unlike the string,
had a constant hard form, which allowed him to use its measurement easily.
I do not think that Amati and Stradivari created the form of the violin arch intuitively,
using merely the talent of a sculptor and an aesthetic vision of the beautiful. In the first
place they were concerned with sound, and so, engineering problems for them rose above
pure aesthetics. Thus, it is possible to conjecture that firstly they created some model of a
future arch, but afterwards they copied it with the help of the sixths. My suggestion is
that this model wasn't made of wood. To reach this conclusion I can list several more or
less cogent arguments:
- they considered that it is impossible to do the model of wood;
- if it were easy to make of wood, that we do not need the model, because it is possible to
make the arch of the plate directly;
- the material for the making of the model must be flexible and springy as was the string
for the design of the violin outline;
- the physical force, which changes the form of material, must be natural and directed to
its whole surface to obtain a high-quality construction from an engineering perspective;
- after the completion of the model the material must lose its springy qualities and
become enough hard as a matter of convenience to make a copy of it.
In my experiments I tried a great deal but the results of those attempts will remain outside
this book. For now I will show only the final result of my studies. In the section The
Arches of the Belly and the Back I listed several types of curvilinear surfaces with the
generator of the variable type: regular, graphic, topographical, gravitational and
others? I referred to all, except the gravitational one. Now it is time to turn our attention
to it.
Make a frame by cutting out the hole in the form of the violin plate without corners from
some board (Figure 108) and glue a knitted fabric around the frame. By wetting the
horizontally disposed frame with water, you will find it will sag under the weight of its
own gravity exactly in the shape of the violin arch. This demonstrates that the violin plate
can be seen to be a result of the gravitational surface.
Figure 108. The frame for modeling of the arches of the plates.
This soft construction can be stiffened with plaster of Paris (gypsum). Of course, if the
wet plaster is spread too much, then the fabric will sag too deeply, more than the required
11 mm. Additionally, unequal spreading of the plaster on the surface of the fabric will not
create the smooth surface of an arch, unless it is spread equally with very fluid
consistency. So, I dilute the plaster to the consistency of milk, and spread it on the fabric
with the wide and soft brush (Figure 109).
Figure 109. Modeling of the arches.
The difficulty in properly fulfilling this work is akin to the varnishing of the ready violin.
In practice, it is necessary to cover the fabric with about 5-6 layers of the gypsum milk to
get more or less a hard construction. Any incidental bulges in the layers can change the
nature sag of the fabric and disturb the arch of the model. Spreading the fluid gypsum on
the second and even the third layer must be very careful, since it is possible to break the
still thin gypsum shell and to spoil all the work. If an attempt is made to spread a layer of
thicker gypsum (for instance with the consistency of sour cream) on a yet not hardened
shell with one or two layers, then we risk spoiling the configuration of the arch, because
it is very difficult to spread the thick gypsum on the whole surface evenly.
When the fabric becomes rather stiff and already can not be bent, the gypsum with the
consistency of sour cream is spread in several layers in the same way (Figure 110). I
finish the work by flooding the model with the gypsum up to the edges (Figure 111).
Figure 110. Spreading the layer of thicker gypsum.
Figure 111. Flooding the model with the gypsum up to the edges.
We know that the configuration of the arch of the belly differs from the arch of the back.
If we spread the gypsum absolutely evenly in the first layers, then we will get the arch of
the belly (Figure 112). For obtaining the arch of the back it is necessary to spread the first
layer unevenly, i.e. thicker in the centre and thinner on the edges. The greater sag in the
centre, because of the excess of gypsum, will give us the necessary arch of the back
(Figure 113).
Figure 112. Plaster model of the violin belly.
Figure 113. Plaster model of the violin back.
In general, the first layer of gypsum forms the future arch, and the resilience of the wet
fabric allows forming the arch smoothly, without "big nubs" and other roughness. But if
the second layer will be spread unevenly, then the excess gypsum will bend the weak fine
first layer more and the arch will be uneven.
It is known that A.Stradivari had been defining the height of each arch before the making
of a violin and the sixths of the curvature prove it. How does one form the arch of the
model with a given height for the intended violin? We know that his sixths do not yet
show the curvature at the edges of the instrument, because this was done after the
purfling is inlaid. So before the purfling is inlaid the height of the arch in the centre of the
plate will be about 11 mm disregarding the thicknesses of the edges. And therefore it is
necessary to allow for the sagging of the fabric to such a depth during the fixing of the
first layer of gypsum. If the depth of sagging fabric is not enough, then it is necessary to
spread a second layer of gypsum milk, when the first layer is not yet dry. It is necessary
to repeat the procedure until the sagging fabric will meet the intended size. Technically
these layers when completed will be considered as the first layer. The second and the next
layers must not be spread until the previous layer is dry.
If we work with the model of the back, then we increase the sagging by spreading
additional gypsum only along the vertical central line, more in the centre and less on the
edges.
But if after spreading the first layer of gypsum we get too much sagging of the fabric,
then it is necessary to correct one of two things: either reduce the consistency of the
gypsum milk (and replace the fabric), or change the stretching fabric on the frame. In the
last case two variants are possible:
- replace the fabric (as in the previous case) or
- if the glue is not yet dry, it is possible to tighten the fabric.
In the last case, possible slanting of the fabric will influence the general quality of the
work, and so, the stretching of the fabric during its fixing to the frame is better.
One can make a copy of the model using both ways: making the sixths of curvature and
drawing the topographical diagram.
It seems to me that the arches of the belly and the back, built in such a way, with their
smooth and plastic lines of curvature meet the requirements of the distribution of strain in
them perfectly. Certainly, in the process of working with specific wood a master can
change the height of the arches mildly, but the general nature of the curvature must be the
same as we have determined in the model. And A.Stradivari did it like this, in those
violins which have not been built by according to his sixths of curvature absolutely
exactly. These arching guides only helped Stradivari to operate correctly, reserving the
right of partial changes for him to adjust according to the acoustic qualities of the wood.
The frame is made with due regard for the desired width of the trough along the edges of
the plates, since this curvature at the edges of the instrument will be done only after the
purfling is inlaid. In my case I used the outline of the first internal curved line, shown at
Figure 70 for the belly and at Figure 73 for the back.
One can make the sixths of the curvature of the arches in this way: to make cross sections
of the plaster model along lines, intended for copying (Figure 114), to place those
segments of the arch on a wooden plate, to trace a sidebar of the curvature with a pencil
and to cut out the patterns according to this drawing.
As this way is the simplest way of copying the plaster model, then A.Stradivari probably
used it. If for some violin forms he made not only fifths, but also the sixth longitudal
pattern, then the last was made before making cross sections of the plaster model.
Common sense tells us that we can expect that A.Stradivari did not save these plaster
segments.
In Figures 115 and 116 I have demonstrated the coincidence of the drawings of the fifths
of the belly and the back with their plaster models.
Figure 114. The cross sections of the plaster model.
Figure 115. Superposing the drawings of the fifth and the photos of the cross sections of
plaster model of the violin belly.
Figure 116. Superposing the drawings of the fifth and the photos of the cross sections of
plaster model of the violin back.
Conclusion
I have finished the geometric analysis of stringed and bowed instruments of the great
Italian masters and hope that the presented considerations will help other masters get to
know the creative process in the design of these instruments better. This, in turn, will
inspire them in the quest of making their own models, which will possess the
particularities of the style of the great Italians and still have the feature of individuality,
which a modern master possesses.
Practically on each page of my book I fought with a wish to begin to prove the advantage
of my method in comparison with the investigations of other authors.
The fact that different methods of study of the same instruments yield different results is
absolutely natural. Anyway, all theories, known to the present-day luthiers, reflect the
real characteristics of the violin geometry with a greater or smaller degree of
approximation and are equally right from the standpoint of the abstract feature of its
proportions, but they are incompatible with a reconstruction of the method of working of
the old-time Italian masters.
Comparing different methods, in the first place a researcher turns his attention to such
qualities of a theory as its systematic character, fullness, accuracy of results and
versatility (i.e. wide range of using). But often it is not enough to confirm that this system
of proportions of the instrument was the result of methods, which the master himself
used. Rather, it is a secondary, side result.
How do we find the methods which were used by the luthier, when he was designing
violins? The fact is that the working of the master leaves not only final results, but also
direct traces of the ways by which it was made. The most observable of them is seen in
the different sort of 'mistakes' they made along the way.
A mistake itself can be accidental, but its nature depends upon the intention, when it was
made. Let us take, for example, the asymmetry of the upper or lower bout of some violin
moulds of Stradivari. What can we see? The configuration of the left and the right curved
lines is the same absolutely, but they have no common axis of symmetry. It can happen
either when using the pattern of the whole curved line between the block and the corner,
which have been placed both times a little bit differently, or when using the springy
string, which although it resembles the configuration of the clothoid in all cases, can not
repeat itself as a mirror image exactly.
Since Stradivari did not use special patterns for drawing the outline of the violin's body,
then nothing else is left, but to suppose, he used the springy string, which does not retain
its form after having been used. The use of the compasses would give another kind of
distortion, which is unlike that described above.
We can see the nature of the mistakes in designing the curled line with the compasses in
the example of designing the violin scroll, when a mistake in the determination of the
radius of the following arc of the Bernoulli spiral changes the nature of the curved line
itself.
There was a time when I tried to analyze the violin scroll by the clothoid, but it was
unsuccessful. It is possible, certainly, to find the scroll, which could be described by the
clothoid, but it would be a rather occasional event, not a regular one.
Failure overtook me when using the Archimedes' spiral, which after C.F.Sacconi
everybody considered as central for the development the scroll. Below I adduce two
drawings of the Archimedes' spiral: 1st - drawn by the hand of C.F.Sacconi, and so, not
quite exact; and 2nd - reconstructed with the help of a computer with the sufficient
degree of accuracy.
The particularity of the Archimedes' spiral is that distance between the whorls of the
spiral is alike everywhere, but such a case exists extremely seldom and only in some
violin scrolls.
Figure 117. The Archimedes' spiral.
If this analysis may seem unconvincing to whomever and he/she will want to correct or
even completely change the modus operandi, then I will say: "Good luck!" If such new
studies will be logical and motivated enough, I am ready to consider these forthcoming
opinions.
Bibliography
ABELL, A.M. - The Partello collection of violins, 1909.
ALEXANDRE, A. - Les Stradivari. Paris, 1945.
APIAN-BENNEWITZ, P.O. - Die Geige, der Deigenbau und die Bogenverfertigung.
Weimar Voigt, 1892.
BAGATELLA, A. - Regole per la construzione dei violini, viole violoncelli e violoni.
Padova, 1914.
BACHMANN, A - An Encyclopedia of the Violin. New York, 1966.
BOYDEN, D.D. - The History of Violin Playing from its origins to 1761 and its
Relationship to the Violin and Violin Music, London, 1965.
COVENTRY, W.B. - Notes on the construction of the Violin. London, 1902.
DAVIDSON, P. - The violin. Glasgow, London, 1871.
GOODKIND, H.K. - A violin Iconography of A.Stradivari. New York, 1972.
HAND, L. - How to make a fiddle. Chicago, 1903.
HILL, W. - Antonio Stradivari: his life and work (1644-1737). London, 1902.
JALOVEC, K. - Italianische Geigenbauer. Prag, 1957.
JALOVEC, K. - Böhmische Geigenbauer. Prag, 1959.
JALOVEC, K. - Die Schönsten Italienischen Geigen. Prag, 1963.
JALOVEC, K. - Enzyklopädie des Geigenbaues. Prag, 1965, 2 voll.
JALOVEC, K. - Deutsche und Österreichische Geigenbauer. Prag, 1967.
KRESÁK, M. - Husliarske Umenie na Slovensku. Tatran, 1984.
MOECKEL, O. - Die Kunst des Geigenbauers. Berlin, 1954.
NICHOLOSON, J. - Design and plans for the construction and arrangements of the new
model violin. London, 1880.
OAKES, W.W. - Construction of the violin. (In: Music, A monthly Magazine, Chicago,
1897).
PANCALDI, C. - Progresso italiano nella costruzione del violino operato da Antonio
Gilbertini da Parma. Palermo, tip. Maddalena.
PELUZZI, E. - Tecnica costruttiva degli antichi liutai italiani. Firenze, 1978.
PUCCIANTI, A. - Antonio Stradivari. Cremona, 1959.
SACCONI, S.F. - The 'secrets' of Stradivari. Cremona, 1979.

 Комментарии: 1, последний от 19/08/2009.

 © Copyright Sergei Muratov
(sergeimuratov(a)bigpond.com)
 Размещен: 16/06/2008, изменен: 27/01/2013. violin_design m/muratow _s_w
173k. Статистика. Оценка:
 Монография: История, Естествознание, 6.17*71 Ваша оценка:
Изобретательство
?? ????? ??

Связаться с программистом сайта.

Новые книги авторов СИ, вышедшие из печати:

Н.Жильцова "Узоры тьмы" К.Медведевич "Ястреб халифа" Т.Форш "Долгая дорога к трону"
П.Комарницкий "Всего один шаг" Ю.Васильева "Урожай собрать не просто" В.Корн "Небесный
странник" С.Кусков "Игрушки для императоров:лестница в небо" А.Гутник "Солдаты шестой
заповеди" Е.Звездная "Экстремальное интервью,или Девушка для героя" И.Мороз "Сталь и песок"
В.Дубровский "Ворон.Кочевые дороги" С.Василика "Обыграть темного эльфа" С.Банцер "Точка
невозврата" А.Черчень "Разное счастье нам выпадает" В.Горъ "Щит" Ю.Шолох "Звериный подарок"
В.Чиркова "Бегущие по мирам" Е.Никольская, К.Зимняя "Девушка для дракона" В.Крабов "Барон"
К.Демина "Изольда Великолепная" К.Назимов "Рыскач.Артефакты истинных магов" Е.Щепетнов
"Слава.Гладиатор поневоле" М.Михеев "Стрелок" В.Контровский "Холодная нефть с горячим
запахом крови" А.Бобл "Аномальный континент" Ю.Иванович "Амазонки Янтарного мира"

Как попасть в этoт список

Сайт - "Художники"
Доска об'явлений "Книги"