Sei sulla pagina 1di 17

Three entriesAcoustics, Harmonic series,

Temperamentfrom the Oxford Companion to Music


(Oxford Music Online is an e-resource available through the library webpage).

Acoustics
1. Introduction
Acoustics is the science of sound and hearing. Sound is a form of energy and involves
vibratory motion. When a piano is played in a concert hall, for example, the energy
exerted by the player causes the hammers to strike the strings and set them into vibration.
This vibration is taken up by the soundboard and radiated as a pressure wave through
successive layers of air particles. The members of the audience hear the sounds when the
air sets their eardrums in motion to produce signals that are communicated to the brain
via nerve-fibres. This simple description, of course, takes no account of the subtleties
introduced by the player, the instrument itself, and the acoustic properties of the hall.

2. Sources of musical sounds


In the present context we think first of the various families of musical instruments, but
there are many other sources of sound to which human ears can respond. The
characteristics of human hearing, evolved in prehistoric times, naturally set limits to the
ranges of loudness and pitch over which musical instruments and voices may usefully
extend.
It seems likely that the first musical instruments were simple percussive devices such as
blocks of wood or hollowed-out tree-trunks, the precursors of the present-day percussion
family. The introduction of stretched skins and metals would have added variety and
produced notes of definite pitch. The first wind instruments would have been reedless,
ancestors of the recorder and flute, their air columns being set into vibration by the player
blowing across the opening at one end of a bamboo or other pipe. Cutting out fingerholes
to alter the effective length of the air column would greatly have extended the scope by
producing a choice of notes of different pitch. Blowing edgewise on to blades of grass
could have led to the development of reed instruments in which the vibrating reed,

activated by the player's breath, became the source of the sound, and the air column acted
as a tuned resonator to determine the pitch of the note emitted.
In brass instruments, such as the trumpet and horn, air is forced between the player's lips,
placed more or less tightly against the mouthpiece. The lip vibrations act as the source
and, again, it is the effective length of the air column that sets the pitch. The human voice
works similarly by forcing air through a gap in the vocal cords, with the chest, mouth,
and throat cavities acting as resonators.

3. Music and noise


The traditional distinction between music and noise, defining one as pleasant and the
other as unwanted, has become blurred as composers have increasingly used sounds
from a wide variety of sources in their works. A better working definition of music might
be organized sound, which implies some human agency creating patterns of sound that
will entertain or intrigue the listener. These may include dissonant as well as tuneful or
consonant sounds, random or aleatory elements as well as strict measures, and editing
together of tape or digitally produced fragments of speech and other sounds (e.g. steamengines or traffic noise).
While music has pushed out its boundaries in this way, becoming louder at live pop
concerts and a public nuisance in its background or portable stereo manifestations,
noise has also increased. The sound of mechanized transport, on the roads and in the air,
has adversely affected the quality of many people's lives. However, in spite of these
overlaps between music and noise, Western music remains the province of the
traditional instruments of the orchestra and the singing voice. And the primary
characteristics of all such music are rhythm and a regular scale of pitch intervals.

4. Pitch of sound
It is well known that the pitch of a note on the musical scale is directly related to the rate
of vibration. If a circular saw is speeded up, the number of vibrations or sound impulses
per second (caused by the individual teeth striking the wood) goes up and so does the
pitch. The lowest vibration rate that produces a musical note, rather than a succession of
separate pulses of sound, is about 20 per second. At the high-pitched end the limit is
about 20,000 vibrations per second, though people vary in their ability to hear such
sounds.

A complete vibration is called a cycle; it consists of a full excursion of the vibrating


element from its rest position out to one side, back through the centre, out in the opposite
direction, and then to the centre again. This is illustrated in Fig. 1, where the motion of
the tip of a tuning-fork fitted with a small pen is traced on a moving roll of paper. The
amplitude of the vibration, corresponding to the loudness of the sound heard, is not
uniform in practice; the naturally occurring variation is known as the envelope of the
sound wave. The number of cycles per second is called the frequency, given in Hertz
(Hz), one Hertz being one cycle per second. For many years there was no agreed standard
for the true pitch of written scores, but in 1939 International Standard Pitch was set at
440 Hz for a (the note A above middle C; see pitch, 2). The precise pitch of one note on
the staff having been set, the rest follow a simple arithmetic sequence (but see scale;
temperament). It can easily be demonstrated, for example, that the basic musical interval
of an octave corresponds to a doubling or halving of the frequency. Thus the octaves of A
on a piano keyboard have frequencies and musical notation as shown in Fig. 2.

The natural or fundamental frequency of vibration of a stretched string is determined by


three factors: its length, tension, and mass (or weight) per unit length. The piano has
separate strings for each of its 88 notes, graduated in length and thickness to give
reasonably equal values of tension. The instruments of the violin family make do with
only four strings of equal length, which the player fine-tunes by adjusting the tension
before playing. The effective length of the strings is then varied by pressure (stopping)
from the fingers. Tension is also used to tune the skins of timpani, for example, whereas
the fundamental pitch of wind instruments is essentially fixed by the length of the
contained column of air. The player can alter this by means of fingerholes, keys, valves,
or a slide (on the trombone) and also, on brass instruments, by changing the lip pressure
against the mouthpiece.

5. Harmonics
Few sound sources perform such simple vibrations as to emit a single frequency. The
pure tone of a tuning-fork and some notes from a flute come close, and electrical
oscillators can generate a single frequency. The richer sounds heard from most musical
instruments result from the simultaneous setting up of several modes of vibration
whenever the instrument is played. A vibrating string, for example, can oscillate as a
whole to produce the fundamental note which establishes the pitch of the note we hear.
At the same time, the string will break up into partial modes of vibration, with each half,
third, or quarter behaving like a separate string (see Fig. 3). This will generate a series of
overtones having two, three, and four (etc.) times the frequency of the fundamental (see
Fig. 4). These are called harmonics (see harmonic series) and contribute much to the
richness of individual instruments (see below, 6). It will be seen that the octaves above
the fundamental (or first harmonic) correspond to the second, fourth, eighth (etc.)
harmonics, having two, four, and eight times the fundamental frequency. The seventh and
higher odd-numbered harmonics do not fall precisely on notes of the scale; thus they
introduce dissonance, so it is fortunate that the upper harmonics tend to weaken
progressively.

In wind instruments, an open pipe is equivalent to a stretched string, except that the two
open ends are points of maximum rather than minimum amplitude (antinodes; see Fig. 5).
The centre point, for the lowest (fundamental) vibration mode, is one of zero amplitude
(node) and the full series of harmonics is possible. In a closed pipe, however, one end
becomes a point of zero amplitude, and the fundamental frequency is an octave below
that for an open pipe of equal length. Only odd-numbered harmonics are formed, and a
different timbre is produced (see below, 6). Tapered or conical pipes set up acoustic

conditions of other kinds. A closed conical pipe produces the full harmonic series and
generally behaves like an open cylindrical (straight-sided) pipe of the same length.

String players are accustomed to playing harmonics, which the composer will indicate
by placing an over the notes in question. The technique used is to touch the string very
lightly at its centre, or at a third of its length. This inhibits the formation of the
fundamental, and the note heard is respectively an octave, or an octave plus a 5th, higher,
with a thin, silvery quality.
Note production in brass instruments is based firmly on the harmonic series. The player
can select any one of the first dozen or so harmonics by altering lip pressure and
technique. Naturally the pitch intervals are wide apart at the lower end of the range for
the given tube length. The ability to play only notes of the harmonic series is the
characteristic feature of such simple instruments as the bugle or posthorn. Further notes
became available when added lengths of tube, known as crooks or shanks, could be fitted.
This gave a new (lower) fundamental and its associated series of harmonics. The later
introduction of valves extended the versatility of brass instruments to cover the full
chromatic scale, the player merely pressing on piston keys to select fixed crooks of
different lengths. The trombone already had this facility, as well as gliding, portamento,
and vibrato effects, since one tube slides into another to give continuous control of
overall tube length.

6. Timbre
As a rule, then, each note from a musical instrument consists of a fundamental tone,
which usually sets the pitch of the note, together with a number of harmonics (overtones
at frequencies which are simple multiples of the fundamental). The tone-colour or special
timbre of each instrument is largely a function of the numbers and strengths of the

harmonics present. Trained listeners can not only distinguish between the different
families of instruments but even recognize individual violins, flutes, clarinets, etc.
In common with most percussion instruments, the piano is incapable of producing
continuous notes. The notes begin to decay from the instant of striking, with the higher
harmonics tending to die out first so that the timbre smooths out as the loudness
diminishes. The ability to recognize different instruments is also helped by the attack or
transient with which each note begins. The transient can contain very high frequencies
up to and beyond the limit of human hearing. This explains one difficulty experienced by
the makers of electronic musical instruments: they may successfully generate and mix a
family of overtones to simulate the sound of a flute or oboe, for example, yet the sound
may disappoint because the attack is missing (see electronic musical instruments).
Transients are also difficult to record and reproduce faithfully since they require extended
bandwidth and fast-reacting circuitry and loudspeakers.

7. Resonance
For a sound source to radiate sound waves effectively it must be capable of setting a
substantial volume of air into vibration. Tuning-forks and violin strings, for example, are
inefficient since their motion cuts through the air without imparting much energy to it.
Much louder sounds can be produced, however, if the stem of a tuning-fork or the ends of
a stretched string are touched on to a table top, or better still a hollow wooden box. Then
the initial vibrations are transmitted into the table or box and the resultant sympathetic
vibrations are able to produce a greater amount of air movement.
As musical instruments evolved, ways were found of increasing this reinforcement of
sound volume using the fundamental principle of resonance: that any structure having
both mass and elasticity will exhibit one or more natural frequencies of vibration which
are relatively easy to stimulate. To take the simple example of a garden swing, children
soon find that applying even a tiny push in time with each cycle of the swing's natural
motion can build up much higher amplitudes. The swing is behaving like a tuned
resonator which responds strongly to a driving force at its own natural frequency but is
less responsive at other frequencies. Musical equivalents include the resonating tubes
mounted under each bar of a marimba or vibraphone.
However, when constructing the soundboard (or body) of a piano or violin, it is necessary
to produce a broadly tuned resonator capable of reinforcing the string tones over a wide
range of fundamental and harmonic frequencies. In practice, instruments of the violin
family do not resonate with equal efficiency at all frequencies. They remain to some
extent frequency-selective, in spite of the special shaping, tensioning, and varnishing
techniques evolved to achieve the most pleasing tonal quality, responsiveness, and

sonority. Each instrument will therefore colour the sound to some extent, and small
differences will always exist.
Many instruments tend to reinforce harmonics in a particular band of frequencies, no
matter which fundamental is being sounded. This pitch region is known as a formant,
and it may be necessary for the player consciously or subconsciously to keep it under
control. The human voice is a prime example of physical differences contributing
individual colorationsand indeed each vowel sound is characterized by two fixed
formant regions.
The important effects of resonance on musical sounds are also found in any room where
music is being performed. Every room has natural resonant frequencies, referred to as
eigentones and related to the dimensions of length, breadth, and height. The room is not
unlike a complex organ pipe in which standing waves are set up between parallel wall,
floor, and ceiling surfaces. Selective resonance at these eigentone frequencies will
inevitably colour the sound, especially in small rectangular rooms where the resonant
frequencies are high enough to fall within the musical range. Some control can be
introduced using soft furnishings to damp out the resonances or irregular wall shapes to
increase diffusion of the sound energy. See architectural acoustics.

8. Combination tones and beats


When more than one note is sounding it is often possible to distinguish each one by ear,
but side effects occur depending on how close the notes are in pitch. Imagine that one
instrument plays the note a constantly (440 Hz) while another plays a note that can be
varied in pitch. If the second instrument also plays 440 Hz, there is perfect unison and the
note simply sounds louder. But if the second note is sharpened slightly, say to 445 Hz, a
note of some intermediate pitch is heard that pulsates in loudness as the peaks and
troughs of the two waves drift in and out of step. These pulsations are called beats, and
in this example (known also as a difference tone) there will be five beats per second.
Piano tuners and others listen for, and work to eliminate, beats as a means of accurate
adjustment of correct pitch.
Raising the pitch of the second instrument produces a degree of unpleasant dissonance
which varies with the frequency difference and falls effectively to zero at the familiar
consonant intervals of a major 3rd (4:3), perfect 5th (3:2), and octave (2:1). In the special
case where the octave is sounded, and for all adjacent numbered harmonics, the
difference frequency is in fact the fundamental. This accounts for the effect known as the
subjective fundamental, heard when a small radio set, for example, appears to
reproduce bass notes which its small dimensions would normally be unable to radiate

effectively. The beating between adjacent harmonics causes the brain to hear the nonexistent fundamental (see acoustic bass, 1).

9. Transmission of sound
The invisible nature of sound waves has encouraged those engaged in research to
improvise models or analogies both to aid their own understanding and to explain the
nature of sound to the world at large. In 1660, for example, Robert Boyle proved by
suspending a watch inside a glass jar that sound needs a physical medium for its
transmission. The alarm bell of the watch could initially be heard quite clearly, but when
the air was gradually pumped out the ringing became fainter and eventually inaudible.
Returning air to the jar restored the ringing sound. Similar experiments with sound
sources immersed in water prove that sound travels as easily through liquids as through
gases. Solids too make an efficient medium for the transmission of sound, as witness the
stories of Amerindians pressing their ears to a railway track to listen for trains several
miles distant.
Clearly the medium itself does not travel from the source to the observer; it merely hands
on the vibratory motion in the same way as a chain of firefighters might pass buckets of
water from a tank to the place of a fire. In a sound wave, each particle of the medium
passes on the energy as imitative vibration about its normal position. The speed of
transmission is greater when the vibrating particles are closer together, which suggests
that the speed of sound is greater in liquids than in gases, and greater still in solids.
Typical values are as shown in Table 1.
Speed of sound in various media at 15C medium metres per second feet per second
air
340
1120
water
1420
4600
wood (oak)
4400
14,500
16,700Table 1
aluminium
5100

The speed of sound in air increases by about 0.61 metres per second for each 1C rise in
temperature. It should be noted that the nominal speed is only about 1200 km per hour.
Comparison of this with the speed of light and radio waves (297,600 km per second)
explains why the sound of distant wood-chopping, for example, is heard some little time

after the axe is seen to strike the wood. It also makes clear the need to group musicians
reasonably close together if time lags caused by distance are to be kept to a minimum.
A popular analogy for the transmission of sound is dropping a stone into a still pond,
causing ripples to spread outwards in the form of ever-widening circles. The stone drags
down some water particles, these drag adjoining particles, and the motion is progressively
imitated by successive layers of particles. The original particles then swing upwards
again, and so on. This single shock wave is analogous to a handclap, but it would be
possible to produce the equivalent of a sustained musical note by vibrating a plunger up
and down to generate continuously radiating waves. It should be noted first that the
advancing wave is a perfect circle, showing that the speed of transmission is equal in all
directions; second, increasing the vibration rate does not alter the speed of the travelling
waves but merely causes the wave crests to move closer together; third, increasing the
amplitude of plunger motion (more energy) again leaves the speed of the waves
unchanged but it does increase their magnitude and the distance they will travel before
the energy is dissipated.
As has been stated, sound is vibration, and one way of demonstrating the motion of the
tip of a tuning-fork was illustrated in Fig. 1. A more versatile method for showingand
even recordingthe vibratory nature of sound waves was the Phonautograph, devised by
Leon Scott in 1857. This consisted of a large horn, to collect the incident sound energy,
with a thin membrane at the narrow end. A bristle was attached to the membrane and
rested against the lampblacked surface of a revolving cylinder. If a flute was played near
the horn, the bristle vibrated in sympathy and inscribed a wavy line on the cylinder.
Counting the number of wave cycles traced per second gave the frequency of the note,
and the distance occupied by one cycle was defined as the wavelength. Given that the
velocity of a wave is the distance travelled per second, we arrive at the universal formula
which holds for wave transmissions and recording media of all kinds: velocity =
frequency wavelength.
The nominal velocity of sound in air having already been noted as 340 metres per second,
it follows that the wavelength of a note at 340 Hz is 1 metre, that at 34 Hz it is 10 metres,
and so on. For the generally accepted range of audible frequencies, 2020,000 Hz,
wavelengths in air range from 17 metres to a mere 17 mm.

10. Radiation and reflection of sound


The concept of wavelength has a direct bearing on how sound waves travel out from
various sound sources, are reflected by obstacles in their path, and build up or decay in
rooms or concert halls.

When a sound source is small compared with the wavelength of the note it is radiating,
the waves travel outwards with equal strength in all directions in a three-dimensional
version of the pond waves discussed above, the wavefront being an expanding sphere.
Since the original amount of energy is being spread over a larger area, the intensity of the
sound (i.e. the energy passing through a given area of the wavefront; see below, section
11) diminishes progressively. Given that the area of the surface of a sphere is
proportional to the square of the radius, the intensity will fall as the square of the distance
travelled, the well-known inverse square law.
By contrast, a source that is larger than the wavelength generates a plane (or flat)
wavefront whose intensity falls off only slightly with distance. Since sound wavelengths
range from several metres down to a few millimetres, it can be seen that most musical
instruments and loudspeakers are intermediate in size. They therefore tend to radiate bass
notes (long wavelengths) rather inefficiently in all directions and higher notes (short
wavelengths) strongly in specific directions, generally along a major axis.
Like other waves, sound waves are bounced or reflected when they meet a wall or
obstacle. Again the wavelength is important, and an obstacle must be relatively large for
reflection to take place. Longer wavelength sounds effectively bend round the obstacle
and continue outwards (the process known as diffraction). When a wave meets a large
surface at right angles, it is reflected back along its original path. This sets up an
interference pattern called a standing wave with peaks and troughs of sound energy at
multiples of a quarter-wavelength from the reflecting surface. This explains the
appearance of dead spots in auditoria and unwanted resonance in small rooms (see
architectural acoustics). Reflection is most efficient from a hard surface like marble, of
course, whereas soft or porous surfaces absorb some of the incident sound energy.
Similarly, an irregular or indented surface tends to scatter or diffuse the sound.
When a reflected sound reaches an observer after the direct wave by an interval of more
than about 120; of a second, it appears as an echo and two distinct sounds are heard.
This implies a distance of about 17 metres, so that echoes can become audible in large
halls, churches, and city streets, for example. Another common experience is the change
in pitch that occurs when there is relative movement between source and listener. This is
known as the Doppler effect; good examples are the sudden lowering of pitch of a
passing train siren, or the similar effect heard when one is on a train passing a stationary
bell, perhaps at a level crossing. The actual frequency heard in such cases depends on the
sum of the velocities of sound and the train. Thus more cycles per second reach an
observer moving towards a source (higher pitch) and fewer when moving away (lower
pitch).

11. Intensity of sound


The sensation of loudness is evidently related to the magnitude of the air vibrations
close to the ears, and will therefore depend on the power of the source (measured in
watts) and the distance between source and observer. Only about 1 per cent of the energy
expended by a musician emerges as sound, the rest being lost as heat in overcoming
friction, etc. Table 2 shows the wide range of total powers radiated by typical sources.

Typical power levels musical source power (watts)


75-piece orchestra (fff)
70
bass drum
25
pipe organ
13
cymbals
10
trombone
0.6
piano
0.4
75-piece orchestra (mf)
0.09
flute
0.06
normal speech
0.000024
0.0000038Table 2
violin (ppp)

In practice, only a small fraction of the sound radiated by a source reaches the ears of any
one listener; the intensity (the power passing through a given area; see above, section 10)
is therefore very small indeed. The human ear can respond to an astonishing range of
intensities over a ratio of about 10 million million to one. The relationship between
intensity and subjective loudness is not strict, but a tenfold increase in intensity roughly
equals a doubling in loudness. There are 12 such steps between the quietest and loudest
sound thresholds (see ear and hearing). Each step is called a Bel, and a smaller, more
convenient unitthe decibel (dB)is commonly used. Table 3 gives some examples of
intensity levels. Note that the decibel is a unit of relative rather than absolute level and
the figures listed are with respect to the standard threshold of hearing intensity, one
million-millionth of a watt per square centimetre. This is the weakest sound audible to
normal ears and corresponds to an eardrum amplitude of less than the diameter of an
atom.

Typical sound levels source or situation sound level (dB)


threshold of pain
130
symphony orchestra (fff)
110
underground-train interior
94
average street traffic
74
conversational speech
60
suburban sitting-room
45
broadcasting studio
20
0Table 3
threshold of hearing

The ear is not equally sensitive at all frequencies, but reaches a peak in acuity in the
10002000 Hz region. The standard frequency for acoustic measurements is 1000 Hz
(kiloHertz, abbreviated to kHz) and the loudness unit or phon is used to describe the
loudness level of any sound by reference to the intensity level in dB of an equally loud
tone at 1000 Hz. Coincidentally, one phon is approximately the smallest detectable
difference in loudness.
John Borwick

Bibliography
A. Wood, The Physics of Music (London, 7/1975)
C. A. Taylor, Sounds of Music (London, 1976)
M. Campbell and C. Greated, The Musician's Guide to Acoustics (London, 1987)

Harmonic series.
A series of frequencies that underlies music in many ways (see acoustics, 5; ear and
hearing, 4) and features in the playing technique of stringed and wind instruments.

The series 1, , , , etc. is musically represented in the divisions of the length of a


string or, in wind instruments, of an air column. The corresponding frequency-values
(and therefore the musical pitch) follow a reciprocal series 1, 2, 3, 4, etc., since twice the
Comparison of intervals in the equal-tempered
scale and the harmonic series (cent values
Harmonics
shown in parentheses) equal temperament
C (0)
1
2

C (100)
D (200)
E (300)
E (400)

16
17
(105)
9 (204) 18
19
(298)

5
10
(386)

21
(471)

F (500)
11
(551)

22
23
(628)

F (600)
G (700)

20

3
6
(702)

12

24

G (800)
13
(840)
27
(883)

A (900)
B (1000)
B (1100)
c (1200)

7
14
(969)
15
(1088)

wavelength corresponds to half the frequency, and so on. To demonstrate the series
musically, the note C (C below the bass staff) is conventionally taken as the root or
fundamental. Ex. 1 illustrates the series up to the 24th harmonic, since this is found in
some 18th-century music for brass instruments.

Except for certain stringed instruments, on which natural harmonics are produced by
touching the string at particular points along its length, the standard scheme numbers the
octave harmonic 2 (as in Ex. 1); this has the great advantage of bringing the numbers in
line with the interval ratios. For example, the Cs 1, 2, 4, 8, etc. are each an octave apart,
i.e. in a ratio of 1:2. Every interval in the series is expressible as the ratio of the two notes
involved: for example, the interval GE (3:5) is the natural or just major 6th, as is DB
(9:15 = 3:5).
The higher one goes through the series, the smaller the successive intervals become; thus,
in Ex. 1, the intervals between harmonics decrease progressively from an octave (1:2) to
about three-quarters of a semitone (23:24). This means that the series does not line up
with a musical scale, in which the intervals are repeated in each octave. In practical
music-making the differences are evened out by either mean-tone or equal-temperament
tuning (see temperament).
The differences between intervals in the harmonic series and in the equal-temperament
scale can be shown in terms of cents, a cent being 1100 of an equal-temperament
semitone. This is illustrated in Table 1, again based on C below the bass staff, and
beginning each octave from zero cents. These are calculated values and may differ in
practice. For example, harmonics (partials) of a stretched string may become increasingly
sharp at higher frequencies because of string stiffness.
Anthony Baines/John Borwick

Temperament.
A method of tuning in which some concords are made slightly impure so that few or none
will be unpleasantly out of tune. This became essential with the introduction of keyboard
instruments. Voices and many other instruments can modify their notes according to
context, varying the pitch slightly to keep in tune, but with keyboards all pitches are
fixed. A major scale which is perfectly in tune starts with a major whole tone, followed
by a minor whole tone, and then a semitone, measuring 204, 182, and 112 cents
respectively, together making a perfect 4th of 498 cents (one cent is 1/100 of an equaltempered semitone). Such a scale could be set on a keyboard instrument but it would be
impossible to start a new scale on the second of those notes, because the next step would
be a minor instead of a major tone.
The first medieval tempered scale was the Pythagorean, where every tone is a major tone
and all 5ths except one are pure, exactly in tune. One 5th must be smaller than the others
by 24 cents (an eighth of a tone, termed a Pythagorean comma), because the sum of 12
pure 5ths, each 702 cents, is 24 cents greater than that of seven 1200-cent octaves. A
scale built in pure 5ths will never return to a pure octave without compromising one of
the 5ths.
A further difficulty arises from the fact that the sum of three major 3rds, each 386 cents,
is smaller than an octave by 41 cents, almost a quarter-tone. The result of compensating
for this is that the better in tune one makes the 3rds, the worse the 5ths become, and vice
versa. The Pythagorean temperament has perfect 5ths but some appalling 3rds, so sharp
that the 3rd was regarded as a dissonance in the Middle Ages simply because it was
indeed dissonant. By the mid-15th century, and perhaps earlier, musicians including
Arnaut de Zwolle were carefully planning their use of Pythagorean temperament
starting on B and tuning 5ths downwards from there, for instance, so that the bad 5th was
the little-used G E and there were four almost pure 3rds (DF , AC , EG , BD ) in
keys in which they wanted to write.
When harmony had evolved to the stage when almost any 3rd was required, a new
temperament had to be devised, with all 3rds pure and 4ths and 5ths as nearly pure as
possible. This was achieved by halving the 386-cent 3rd, taking the mean, or average,
size of whole tone: 193 cents. The resulting temperament, called mean-tone, was
constructed by tuning CE pure and then tuning each 5th within that 3rd (CG, GD, D

A, AE) a quarter of a comma flat. The only disaster that resulted was the size of the
discrepancy between G and A : the two notes, which are the same pitch in equal
temperament, are 41 cents apart in mean-tone, and using one instead of the other
produces a chord so out of tune that it howls like a wolfhence the expression wolf
5th. There were also four wolf 3rds, wildly sharp, but these were kept in keys which
composers took care to avoid (e.g. C , F , B, and G in a tuning cycle starting on C: it
was always possible to move the wolves by starting the tuning on a different pitch).
Quarter-comma mean-tone was first discussed by Zarlino in 1571. Sixth-comma was an
improvement as music became more chromatic in style, because though the 3rds were
very slightly worse, the 5ths and 4ths were equally slightly better and the wolves were
smaller and howled less. It is often used today for performances of early music. Wolves
of some sort are inevitable in any temperament which uses the same correction all the
way (a regular temperament) with a specific fraction of a comma. The only exception is
equal temperament, which has the disadvantage that every interval is out of tune except
the octave and that the 3rds (400 cents instead of the pure 386) are almost as bad as the
Pythagorean. For that reason musicians try to avoid it except when playing with a piano.
It is produced by tempering the 5ths, flattening each by 2 cents, to spread the 24-cent
comma equally through the octave.
From the 17th century, irregular temperaments such as those devised by F. A. Vallotti,
Andreas Werckmeister, Kirnberger, and others were used instead of either equal
temperament or mean-tone. These were based on improving the keys most likely to be
used, but because one key cannot be made better in tune without another being made
worse, the keys less often used were degraded. It has been demonstrated that Bach's
well-tempered keyboard for the 48 used such a temperament, with some movements
showing the purity of the better keys and others, with rapid passage-work, disguising the
faults of the poor ones.
The one sure way of avoiding wolf notes but still keeping 3rds and 5ths almost pure was
by increasing the number of notes in the octave. On a keyboard with separate keys (and
strings or pipes) for D and E , and for G and A (the notes worst affected in any
normal temperament started on C), it was possible to play in almost any key with
confidence. Because of the number of pipes involved, few organs were built in such a
way, but many harpsichords had split keys, the front half producing one pitch and the
back half the other. A few Italian instrumentsthe arcicembalo, for instancewere built
with a large number of keys, and many theorists have calculated tunings involving
anything up to 51 pitches to the octave. A problem is that the more keys there are on the
keyboard the more difficult it is for the player to remember which should be used in
which chord, so that few players have ever tried more than the two basic split keys.

Jeremy Montagu

Bibliography
D. Devie, Le Temprament musical: Philosophie, histoire, thorie et pratique (Bziers,
1990)
M. Lindley, An historical survey of meantone temperaments to 1620, Early Keyboard
Journal, 8 (1990), 129
R. Covey-Crump, Pythagoras at the forge: Tuning in early music, Companion to
Medieval and Renaissance Music, ed. T. Knighton and D. Fallows (London, 1992)
M. Lindley, Temperaments: A Brief Survey, Bate Collection Handbook (Oxford, 1993)
W. A. Sethares, Tuning, Timbre, Spectrum, Scale (London, 1998)

Potrebbero piacerti anche