Professional Documents
Culture Documents
1
TA225 The Technology of Music
Investigating
Sound
Chapter 1 Sound Basics page 3
c
2 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 2
This publication forms part of an Open University course, TA225 The Technology of Music.
Details of this and other Open University courses can be obtained from the Course
Information and Advice Centre, PO Box 724, The Open University, Milton Keynes MK7 6ZS,
United Kingdom: tel. +44 (0)1908 653231, email general-enquiries@open.ac.uk
Alternatively, you may visit the Open University website at http://www.open.ac.uk
where you can learn more about the wide range of courses and packs offered at all
levels by The Open University.
To purchase a selection of Open University course materials visit the webshop at
www.ouw.co.uk, or contact Open University Worldwide, Michael Young Building, Walton
Hall, Milton Keynes MK7 6AA, United Kingdom for a brochure. tel. +44 (0)1908 858785;
fax +44 (0)1908 858787; email ouwenq@open.ac.uk
Chapter 1
Sound Basics
CONTENTS
Aims of Chapter 1 4
1 Introduction 5
1.1 Music and technology 5
1.2 What is sound? 5
1.3 Summary of Section 1 8
2 Sinusoidal pressure waves 9
2.1 Introduction 9
2.2 Pressure in the atmosphere 9
2.3 Pressure waves and cycles 11
2.4 Period 13
2.5 Wavelength 13
2.6 Pressure variations in one place 16
2.7 Summary of Section 2 18
3 Frequency 19
3.1 Frequency and period 19
3.2 Summary of Section 3 19
4 The speed of sound 20
4.1 The experimental result 20
4.2 Frequency, wavelength and the speed of sound 21
4.3 Summary of Section 4 22
5 Phase 23
5.1 Phase and phase difference 23
5.2 Cancellation and reinforcement 24
5.3 Summary of Section 5 26
6 Amplitude 27
6.1 Defining amplitude 27
6.2 Practical units of amplitude 29
6.3 Root-mean-square amplitude 30
6.4 Summary of Section 6 31
7 Pitch and loudness 32
7.1 The subjective experience 32
7.2 Summary of Section 7 34
8 The octave 34
8.1 The octave sound 34
8.2 Octave pitch and frequency increments 35
8.3 Summary of Section 8 37
4 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 4
AIMS OF CHAPTER 1
Q To introduce the concept of a travelling pressure wave as the
physical manifestation of sound.
Q To present a simple model relating pressure waves to molecular
activity.
Q To introduce the basic mathematical properties of sine waves.
Q To introduce the basic terms and measurements of sound used in
acoustics.
Q To relate the terms from basic acoustics to corresponding musical
terms.
Q To explore the subjective perception of pitch and loudness, in
particular their relationship to frequency and amplitude.
5 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 5
1 INTRODUCTION
1.1 Music and technology
Music technology in one guise or another is part of everybody’s life,
because music is a part of almost everybody’s life. For instance, if you
are an instrumental performer of music, professional or not, then your
instrument, be it the harp or the rock’n’roll drums, will be the result of
considerable technological expertise on the part of the instrument
maker. On the other hand, if you are not a performer but like to listen
to music, the chances are that most of your listening is done via a
home hi-fi system or a car radio rather than in a concert hall. In this
case your music is relayed to you via various technological devices;
and well before the music reaches a CD or radio it will have been
manipulated in various way using studio devices for recording, mixing
and mastering. From the traditional acoustic instruments to the
modern computer-based sampler or synthesiser, from the museum
gramophone to the latest MP3 player, whatever your choice of musical
style and whatever your relationship with music, technology is
virtually inescapable.
Technology and music have been closely associated since the first
musical instruments were constructed. This may come as a surprise to
people who are used to thinking of music technology in terms of
electric or electronic devices. The piano, for example, has taken
centuries of evolving technological expertise to become what it is
today. Throughout its history, music has freely exploited state-of-the-
art technological developments, and has had an intimate relationship
with sciences such as physics, mathematics and, more recently,
electronics and computing. Throughout this course you will see many
examples that highlight this close relationship of music with both
technology and the sciences.
This block of TA225, specifically, explores sound, which is the basis of
all music. Music, of course, implies sound. Music technology,
essentially, has the purpose of enabling sound production,
manipulation, storage and reproduction. But what is sound?
Throughout this block you will study the various aspects that, together,
comprise a basic model of sound. In this first chapter, in particular,
you will start by learning about the various ways of interpreting the
word ‘sound’, concentrating at first on a more formal approach to
sound from the perspective of physics.
Comment
Here are my descriptions. Your descriptions will almost certainly be
different from mine because people have different musical backgrounds
and experiences. This is to be expected and is not at all a problem, as
you should be able to relate your descriptions to the main points I will
be making shortly.
Track 1: A major and a minor scale played on the piano.
Track 2: A low-pitched note followed by a high-pitched note played on
the recorder.
Track 3: A drum solo.
Track 4: Humming of machinery.
Track 5: Sound of an ambulance siren.
Track 6: Sounds of sea waves.
Track 7: A tune sung by a male voice.
Track 8: Sounds of three female voices speaking, in different languages. Q
Comment
Naturally I do not know what your descriptions were like, but I would be
very surprised if most of them could not be put into one of these categories. Q
7 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 7
Let’s now take a closer look at my list of categories, starting with item
(a). In my descriptions of this sort, I referred to the source-causes of
the sounds, that is, objects or instruments (the sources of sounds) and
ways of using these to produce sounds (the causes of sounds). Here are
other examples of source-cause descriptions:
Violin (source) played pizzicato (cause); pizzicato is a plucking
technique for string instruments that are normally played with a bow.
Piano strings (source) struck with a felt-covered hammer (cause); this
is the basic mechanism of a modern piano.
Hand (source) clapping (cause); this is central to traditional Spanish
Flamenco music.
Source-cause descriptions are probably the most common way of
describing sounds. Your list probably resembles mine in this respect.
Such descriptions are particularly interesting because it is quite
remarkable how descriptions of sounds seem to rely so much on things
that are not the sounds themselves. Consider the way we would
generally describe an object we see. Normally we would describe an
object by mentioning a label (let’s say, chair) accompanied by some
qualities that, we feel, make the object distinct from others (for example, a
wooden, white-painted, chair in the corner of the room). With a verbal
explanation, we can characterise the object so that it can be distinguished
from other nearby or similar objects. Instead, where sounds are
concerned we tend to mention their origins. This sort of description of
sounds might be compared to describing objects we see by talking
about how they are produced. Perhaps this is because the recognition
of the source-causes of sounds is one of the most basic listening abilities,
an instinctive ability, as you will see in Chapter 5 of this block.
My second category of descriptions (ones that refer to musical elements)
is a bit more specific. If your background is not musical, you may not
have used or mentioned this sort of terminology at all, keeping your
description to the identification of the instruments in the first two
tracks, for example. On the other hand, if your background is musical,
you may have added yet more detail, for example, mentioning the key
in which the scales in the first track are played, the relationship
between them in harmonic terms, or, perhaps, the interval you identify
in the second track. Naturally, the use of musical terminology depends
on musical training, but some of the traditional terms have crept into
colloquial speech, perhaps due to some form or other of musical training
received in early schooling. ‘Note’, ‘key’ and ‘chord’, for example, have
specific meanings in music (which are not easy to pin down), but have
become part of everyday language, as in: ‘that strikes a chord’.
My final category, item (c) ‘Descriptions that use metaphors’, is a most
interesting one, as it relates to a sort of informal language used commonly
by musicians and experienced music listeners. (If you want to remind
yourself about what a metaphor is, see Box 1.) You may have heard
Box 1 Metaphor
According to the Oxford Concise Dictionary, Eighth Edition (1990), a metaphor
is ‘the application of a name or descriptive term or phrase to an object or action
to which it is imaginatively but not literally applicable (e.g. a glaring error).’
8 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 8
transmitted through other media, for instance water, so not all sound
consists of fluctuations in air pressure. However, for the purposes of
this discussion I will confine myself to sound in air.
These fluctuations in air pressure are caused by a local disturbance to the
air pressure, which might be sudden and transient – for example when
a paper bag is burst – or continuous and regular – for example when
someone sings a steady note. Whatever the nature of the disturbance,
the pressure variations spread outwards from the source through the
surrounding air, becoming gradually weaker. At a sufficient distance
from the source, the pressure variations die away completely.
For a listener in the vicinity of the sound source, the pressure variations
act on the listener’s hearing mechanism, causing the eardrum to move
in sympathy with the source of the pressure variations. The movements
of the eardrum are detected by a mechanism that will be described
later in this block, and are interpreted by the brain as sound.
equal pressure
The pressure variations we hear can also sometimes be felt. You may
inside and out be familiar with the experience of holding an inflated balloon and
feeling it vibrate in response to nearby sounds, or standing near a
loudspeaker and feeling vibrations from the bass notes. In the case of
the balloon, pressure variations cause the skin of the balloon to vibrate
in just the same way as they cause the eardrum to vibrate. Similarly,
all drummers and percussion players are familiar with the way
drumskins resonate in sympathy with sound from nearby instruments,
and often need to be damped to prevent unwanted noises being
generated.
reduced
pressure inside The air around us presses on everything it touches, and this is roughly
what is meant by the atmospheric pressure. Generally we are unaware
Figure 2
of the pressure because it acts equally in all directions, and so its
A plastic bottle
can be buckled effects are self-cancelling. However, if you removed some of the air
by unbalanced from an empty, thin-walled plastic bottle, the pressures inside and
pressure outside the bottle would cease to be self-cancelling, and the bottle
would buckle (Figure 2). The kinds of pressure imbalance that would
make a plastic bottle buckle are, however, much larger than the
pressure fluctuations associated with sound.
With regard to sound and the way it travels, we need to think about
pressure in relation to the arrangement of molecules in the air.
Atmospheric air is a mixture of gases, and at the sub-microscopic scale
consists of a mixture of gas molecules. These molecules are so tiny
high pressure that they are only detectable individually by sophisticated scientific
apparatus. In a moderately sized volume of air, such as inside a bottle,
there is a colossal number of such molecules, and between them there
is empty space. The molecules are not static, but continually move
around, bouncing off other molecules or off any solid or liquid objects
in their vicinity. In general, if there is no sound or other disturbance to
the pressure of a sample of air, the molecules are evenly (though
low pressure randomly) distributed throughout the sample.
Figure 3 The pressure of air (or any gas) is related to how closely packed its
High pressure
corresponds to molecules are (Figure 3). If the molecules are widely dispersed, the
closer molecular pressure is lower than when they are closer together – other things
spacing being equal (principally temperature).
11 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 11
When you pump up a bicycle tyre, by driving more air into the tyre
you squash the molecules together more closely than they are outside
the tyre. Hence the pressure inside the tyre is higher than that outside.
The message to remember from this section is that sound consists of
rapid fluctuations of atmospheric pressure, and that, at the molecular
level, high pressure corresponds to air molecules being bunched
together, and low pressure corresponds to air molecules being
relatively more widely separated.
2.4 Period
You saw in the computer animation in Activity 6 that the prongs of
the tuning fork vibrated cyclically. The animation explained that a
cycle of the prongs’ vibration is a complete sequence of motion up to
the point at which the motion starts to repeat itself. Another term for
this repetitive kind of motion is periodic motion. The time taken for
one cycle to occur is called the period of the vibration, and in
theoretical work it is usually represented by T, or by the Greek letter
tau, τ.
Properly speaking, in periodic or cyclical motion every cycle is
identical to every other. With a practical tuning fork, however, no two
cycles are identical. This is because each cycle is slightly weaker than
the one before, as the vibration of the prongs diminishes from the
moment the fork is struck. Nevertheless, it takes several seconds for a
tuning fork to become silent, during which time there will be
thousands of cycles of vibration. Thus over the course of a few cycles
there will be very little change from one cycle to the next and we can
regard the motion as periodic.
2.5 Wavelength
So far we have seen that sound is a pressure wave, and that the
spacing of the pressure variations is related to the period of vibration
of the source.
A graphical representation of the pressure wave from a tuning fork
closely approximates to a certain type of wave known as a sine wave,
as shown in Activity 9.
The last activity showed that if we freeze the pattern of high and low
pressure regions in the pressure wave, we have the pattern shown in
14 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 14
pressure distance
(b) wavelength wavelength
Figure 6 A graph of the pressure wave produced by a tuning fork is a sine wave
Comment
Using the sound-cause type of description, you might have described
some of the sine waves as flute-like, the flute being one of the few
common instruments that can produce a sine wave, or a close
approximation to one.
As far as metaphorical descriptions go, sine waves are often described
as ‘neutral’, ‘pure’ or ‘colourless’. You may disagree – particularly if
you are a flute player. Q
A full discussion of sine waves and their properties would entail quite
a lot of mathematics, which is beyond the scope of this course.
However, you may be interested to know that the oscillations of many
smoothly vibrating systems, when plotted as a graph, have the
characteristic sine-wave shape. Other examples would include the
oscillations of a mass on a spring and the swinging of a pendulum,
provided the oscillations are relatively small in each case.
15 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 15
Notice that in the pressure wave in Figure 6, the distance between any
two adjacent regions of high pressure (or low pressure) is the same.
This distance is called the wavelength of the sound, and is usually
represented by the Greek letter lambda, λ. In fact, the distance between
any two corresponding points of consecutive cycles is the wavelength.
For instance, in Figure 7, points A and B are one wavelength apart, as
are C and D.
pressure
C D
A P B
distance
wavelength λ wavelength λ
Figure 7 Wavelength
0 1 2 distance
(metres)
(a)
pressure
0 1 2 distance
(metres)
(b)
Note that in the time it takes a pressure wave to travel a distance equal
to one wavelength, the source performs one complete cycle of oscillation.
To see why this is so, consider Figure 9(a). The fork is at a particular part
of its cycle, and the point X on the pressure wave is adjacent to the fork.
(a) distance
pressure
Y X
distance
(b) wavelength
one cycle of
oscillation at source
Figure 9(b) shows the situation one cycle later. The prongs of the fork
are back at the same part of the cycle as in (a), and the pressure at
point Y is at exactly the same part of the cycle of pressure variation as
X was. In the mean time, X has travelled away from the fork a distance
equal to one wavelength. Thus, in the time it has taken for the source
to go through one cycle of oscillation, the wave has travelled a distance
equal to one wavelength away from the fork. This is an important point
which I shall return to when we come to look at the speed of sound.
pressure
(a) 1 period 1 period
(b) time
Figure 10 (a) Pressure wave, with one point indicated by arrows. (b) Pressure
variations at the point indicated in (a) as time passes
pressure
pressure
0 1 2 3 time 0 0.1 0.2 time
(seconds) (seconds)
(a) (b)
3 FREQUENCY
3.1 Frequency and period
In Figure 11 you saw that waveform (b) had a much shorter period
than waveform (a). Hence waveform (b) completes more cycles of
oscillation in a second than does waveform (a). Waveform (b) is said to
have a higher frequency than waveform (a). The frequency of an
oscillation (usually represented by the symbol f ) is the number of
cycles there are in a second. This may or may not be a whole number.
For instance, a certain wave might have 25.5 cycles in a second;
another might have exactly 100. In either case, though, the number
quoted is acceptable as a frequency.
The unit of frequency used to be ‘cycles per second’, which had the
merit of being self-explanatory. Nowadays it is given the
internationally agreed unit hertz (symbol Hz), named after the German
physicist Heinrich Hertz (1857–94). A typical tuning fork might have a
frequency of oscillation of 440 Hz (or 440 hertz), meaning that the
prongs perform 440 oscillations every second. For high frequencies the
kilohertz is often used as a unit of frequency. One kilohertz (1 kHz) is a
thousand hertz. (Larger units than the kilohertz are not required for
sound, although they may be used in connection with equipment used
in sound technology.)
The frequency of an oscillation is directly related to the period of the
oscillation. Suppose a source of pressure waves vibrates at the rate of
100 Hz, that is, 100 cycles per second. It is fairly clear that each cycle
must last for one-hundredth of a second. Mathematically we express
this relationship as:
1 1
frequency = period =
period or frequency
1 1
In symbols, f = or T =
T f
1 1
f = or T =
T f
20 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 20
The time taken for the source to perform one cycle of oscillation is its
period T. So, in one period of oscillation, the wave travels a distance
λ. To determine the speed of the wave, we need to know how far it
travels in a second, rather than in one period. In a second there are
f cycles of oscillation, where f is the frequency, so in one second the
wave travels f times as far as it travels during just a single cycle of
oscillation. Thus,
speed = frequency × wavelength
Or, if we let the speed be represented by v:
v=f×λ
This equation above can be restated in two other ways:
speed v
frequency = or f =
wavelength λ
speed v
wavelength = or λ=
frequency f
5 PHASE
5.1 Phase and phase difference
In this section I am considering sine waves which have the same
frequency, but are out of step with each other. The following activity
shows one way in which this can arise.
Figure 13 shows how the pressure varies with time near the tuning
fork (Figure 13a) and at a distance (Figure 13b).
pressure
near fork
0 1 ms 2 ms time
(a)
pressure
at a distance
0 1 ms 2 ms time
(b)
Although the two graphs have the same frequency, they are not in step.
At any given moment, each is at a different part of its cycle. For
example, if you look at the 1 ms point on each graph you can see that
the curves are at different parts of a cycle.
We use the word phase to refer to the part of a cycle which a particular
vibrating system is in at any moment. In practice we are often less
concerned with the phase of a single wave than with the phase
difference between two (or more) waves having the same frequency.
(The stipulation ‘having the same frequency’ is necessary because we
cannot really speak of a fixed phase difference between sine waves
with different frequencies.) Another way of saying that there is a phase
difference between two sine waves is to say that they are out of phase.
When the waves have the same phase, they are said to be in phase.
In Figure 13, on each graph’s horizontal axis events to the right are
happening later than events to the left. Thus, because the first peak in (b)
is to the right of the first peak in (a), we say that the pressure variations
in (b) are lagging in phase behind those in (a). Alternatively, we can
say that the pressure variations in (a) are leading in phase those in (b).
In the last activity there was a phase difference of 0.2 ms, but phase
differences are not always expressed in units of time. Instead, the
following two methods are commonly used to express a phase
difference quantitatively:
(a) as a fraction of a cycle
(b) as an angle.
The first of these is fairly straightforward, as the following activity
demonstrates.
One of the reasons for being interested in phase arises from the
consequences of mixing, or adding, two sine waves that are phase-
shifted relative to each other. Figure 14 shows two sine waves that are
completely out of phase.
time
(a)
time
(b)
The consequence of adding or mixing two sine waves that are 180° out
of phase is complete cancellation of one wave by the other. This is the
basis of a noise-reduction technique sometimes used in noisy
environments: an out-of-phase version of the noise is played through
amplifiers and loudspeakers into the noisy environment, causing
cancellation and thus elimination of the noise. (Incidentally, the term
‘out of phase’ is used here to mean ‘180° out of phase’ rather than just
‘not in phase’.)
When two sine waves are in phase, there is mutual reinforcement. For
instance, in Figure 15 sine waves (a) and (b) are in phase. When they
are added or mixed the result is (c). Note that (c) is a sine wave with
the same period (and hence same frequency) as (a) and (b).
time
(a)
time
(b)
time
(c)
6 AMPLITUDE
6.1 Defining amplitude
Another important property of a sine wave we need to be able to
specify is its amplitude. In essence, the amplitude of a sine wave is its
size. Unfortunately there are various ways of defining what is meant by
the size of a sine wave, and you are likely to come across many of them
in material you look at outside the course. Before I explain what our
definition is, it will help matters if we look at what is meant by the
average value of a sine wave.
Figure 16 shows a sinusoidally alternating voltage. The curve is
symmetrical around the time axis, which is also the line of zero
voltage. The average value of this sine wave over many cycles is
therefore zero.
10 V
voltage
average value
0
time
All sine waves are symmetrical around a line running through the
middle, midway between the peaks and troughs of the wave. However,
it does not follow that all sine waves have an average value of zero.
Look at Figure 17. This is an exaggerated graph of a sinusoidal pressure
variation in the air.
Pmax
average value
atmospheric pressure Pa
pressure
0
time
Once again, the average value over many cycles runs midway between
the peaks and troughs, but now its value is not zero. The average value
is the prevailing atmospheric pressure.
A standard way of defining the amplitude of a sine wave is in terms of
its maximum departure from its average value. To see what this means,
look at Figure 18.
10 V
voltage
amplitude peak-to-peak
amplitude
0
time
a
The amplitude is the height of the peak relative to the average value. In
this case, because the average value is zero, the amplitude is just the
peak value of the sine wave, namely 10 volts. However, in Figure 19
the amplitude is not simply the peak value.
Pmax
amplitude a
peak-to-peak
atmospheric pressure Pa amplitude
pressure
0
time
Here, because the sine wave does not have an average value of zero,
the amplitude is the difference between the peak value and the
average value, that is, Pmax – Pa.
In general, the amplitude of a sine wave is the maximum deviation of
the sine wave from its average value. In other words, it is the
difference between its peak value and its average value. Another way
to express this is half the height of the wave from peak to peak.
Sometimes the peak-to-peak height itself is used as a measure of a sine
wave’s size, because it is easy to read from graphical displays. Such a
29 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 29
distance
0.4
0.3
0.2
0.1
voltage
0
0.1 0.2 0.3 0.4
time
–0.1 (seconds)
–0.2
–0.3
(a) –0.4
voltage
6
–8 –6 –4 0 2 4 6 8 10
–2 distance
–2 (metres)
–4
(b) –6
Comment
(a) I expect you found that changing the amplitude changed the
loudness, and the bigger the amplitude, the louder the sound.
(b) The effect of changing the frequency is rather harder to pin
down. If you are familiar with musical concepts you probably
said that changing the frequency changed the pitch. If you are
not familiar with ‘pitch’ as a musical term, then you might
have said the sound became higher and lower, and you could
regard this activity as supplying an illustration of what we
mean by pitch. Q
Comment
This activity is concerned with the subjective interpretation of
amplitude, so there is no hard-and-fast answer that everyone is sure to
agree with. In fact, you may have found the concept of halving the
loudness of a sound to be almost meaningless. Nevertheless, most
people who do this experiment find that halving the amplitude does
not halve the loudness. Usually a bigger reduction is required to give
the impression of a halving of loudness. Another way to express this is
to say that the amplitude must be more than doubled to give the
impression of a doubling of loudness. Q
8 THE OCTAVE
8.1 The octave sound
One feature of pitch that seems to be universal to all cultures is that
for musical purposes the pitch range is divided into discrete steps, for
instance the notes of a scale. This is not to say that musicians rigidly
adhere to those steps when they play, but the existence of these steps
is fundamental to the way music is conceived and organised. Different
cultures have different ways of defining the steps in their scale of
pitches, but nearly all cultures take the octave as their starting point. It
has a very characteristic sound, and it corresponds precisely to a
particular relationship of frequencies.
middle C X
Comment
Although the two pitches are clearly different (X has a higher pitch
than middle C), most people find that there is nevertheless
something very similar about them. One way to express this idea is
to say that they are two versions of the same musical sound. If you
have difficulty hearing a similarity between them, try the following.
Play all the white keys from middle C to X in order, starting at
middle C, and then play middle C again. When you reach X it
should feel like a return to base, although naturally it is not a
return to middle C. Q
The two pitches you played in the last activity are said to be an
octave apart. The pitch marked X is an octave above middle C. Its
note name is also C. The seventh white note to the right of X
(counting the white note next to X on the right as 1) is another C. It
is two octaves above middle C. The seventh white note to the left of
middle C is an octave below middle C and is also given the note
name C.
Sine waves whose pitches are an octave apart have frequencies in
the ratio 2:1. That is, the higher pitch has a frequency that is twice
that of the pitch below it. Alternatively, the lower pitch has a
frequency that is half that of the pitch an octave above.
1760
frequency (Hz)
880
440
220
110
55
27.5
A0 A1 A2 A3 A4 A5 A6
pitch
A4
440 Hz
B4 C5 B5 C6 B6 C7
C4
B2 C3 B3
B1 C 2
B0 C1 Middle C
D E F G A D E F G A D E F G A D E F G A D E F G A D E F G A
B0 C1 B1 C 2 B2 C3 B3 C4 B4 C5 B5 C6 B6 C7
Another system you might come across in books is the so-called Helmholtz
system. In this system C’’ stands for C0, C’ for C1, C for C2, c for C3, c’ for C4, c’’
for C5, etc.
Yet another system uses CCCC for C0, CCC for C1, CC for C2, C for C3, c for C4, c’
for C5, c’’ for C6, etc.
Although human hearing covers a range of, say, ten octaves at best,
seven of these octaves cover the bottom eighth of the range, from 20 Hz
up to 2500 Hz, which corresponds roughly to the pitch range from E b0
to E b7 (Figure 26). As far as music is concerned, this is where the action
is concentrated. Of the standard acoustic instruments, only the piano,
harp and piccolo go higher than E b7, and those not by very much.
approximately 2500 Hz
approximately 20 Hz
E 0 E 7
Middle C
At the lower end of the musical pitch range, the bottom note on a
double bass or bass guitar, E1, has a frequency of just over 40 Hz. Not
many instruments can go below this, except mainly the harp, piano,
double bassoon and organ. (The keyboard in Figure 26 is extended
below a normal piano keyboard.)
To say that musical instruments rarely produce pitches above Eb7 is
not the same as saying that they rarely produce frequencies above
2500 Hz. The sounds produced by virtually all instruments are not
pure sine waves. Instead, they are more or less complex mixtures of
sine waves covering a range of frequencies above that corresponding to
39 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 39
the pitch being played. I will say more about this in Chapter 2, but for
the moment the important point to bear in mind is that the mixture of
frequencies associated with a single musical pitch can extend well
above the frequency corresponding to the pitch that is heard. The
presence of these additional frequencies (sometimes called overtones,
partials or harmonics) is partly what gives individual ‘colour’ to
particular instruments.
Comment
I expect you answered ‘yes’ to (a) and ‘no’ to (b). (Depending on your
audio equipment, you may not have heard all the changes in (b). There
are as many steps in (b) as in (a).) Q
10 THE DECIBEL
10.1 Introduction
For a variety of reasons, not least the very wide dynamic range of
human hearing, the decibel (symbol dB) is often used as a unit for
the amplitude of sound waves. The decibel is also used in other
contexts, such as specifying the amplification of amplifiers or for
specifying the degree to which a signal is affected by noise. In the
context of sound, the use of the decibel as a unit captures
something of the subjective impression of the way loudness
changes with amplitude.
The decibel unit has two rather unusual properties in comparison with
other more conventional units you have probably met, such as the
metre or the second.
1 It indicates a ratio, rather than an absolute value. Thus the
decibel can be used as a way of comparing one amplitude with
another.
2 Equal decibel increments correspond to equal multiplications of
ratio.
Because the decibel expresses a ratio rather than an absolute value, it
cannot by itself specify the absolute amplitude of a sound. I shall
explain shortly how it can be adapted for the expression of absolute
values, but first I want to pursue the second feature I listed above,
namely that equal decibel increments correspond to equal
multiplications of ratio. I want to do this in the context of ratios of
amplitude.
41 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 41
36 64:1
40 100:1 400:1
60 1000:1
300:1
200:1
100:1
–20 –10 0 10 20 30 40 50 60
decibel scale
Comment
On each line, the decibel value is an equal increment on the decibel
value in the line above. The increment is 6 decibels each time. Thus in
the first part of the audio track in Activity 37 there was a 6 decibel
increase of amplitude with each recurrence of the sound. Q
The equality of the decibel increment you saw in the last activity
therefore matches our subjective sense of equal increments of loudness
when the amplitude is multiplied by a constant factor (2 in this case).
SUMMARY OF CHAPTER 1
Sound can be considered objectively and travels a distance equal to one wavelength.
subjectively. (Section 1.2) (Section 2.5)
Amplitude is the size of a sine wave. It is of the same sound, and they are given the
the maximum value of the wave’s departure same musical letter-name (A, A flat, B etc.).
from its average value. (A sine wave (Section 8.1)
oscillates symmetrically about its average
value.) (Section 6.1) Subscripts may be used to distinguish
different pitches which share the same
Peak-to-peak amplitude is the vertical letter-name, for instance A1, A2, A3, etc.
distance from peak to trough. It is twice the (Section 8.2)
amplitude. (Section 6.1)
Equal separation of pitch does not
Amplitudes of sounds are typically correspond to equal separation of
measured in units of pressure or volts, or frequency, but to equal ratios of frequency.
in decibels. (Sections 6.2 and 10.3) (Section 8.2)
A steady source equal in value to the r.m.s. The pitches used in music correspond
amplitude of a sinusoidally oscillating roughly to frequencies in the seven-octave
source supplies energy at the same rate as range from 20 Hz to 2500 Hz. (Section 9.1)
the oscillatory source (other things being
equal). (Section 6.3) Generally, musical tones are not pure sine
waves but are mixtures of sine waves.
Pitch is the subjective property of sound (Section 9.1)
that is heard to change while frequency
changes. (Section 7.1) The dynamic range of human hearing is the
range of amplitudes the ear can cope with.
Loudness is the subjective property that is It covers a range of more than 1 000 000:1.
heard to change while amplitude changes. (Section 9.2)
(Section 7.1)
The decibel (symbol dB) is a way of
Equal changes of frequency are not expressing a ratio. It can be used to express
experienced as equal changes of pitch. absolute values by referring measurements
Equal changes of amplitude are not to a fixed standard. (Sections 10.1 and 10.3)
experienced as equal changes of loudness.
(Sections 7.1, 8.2, 10.1 and 10.3) Adding decibels is equivalent to
multiplying their corresponding ratios.
For musical purposes, pitches are organised (Sections 10.1 and 10.2)
into discrete, named steps. (Sections 7.1
and 8.1) Decibels can be used to express a ratio of
amplitudes, for instance in specifying the
In the pitch standard known as concert voltage gain of an amplifier. (Section 10.2)
pitch, the pitch A4 is defined as the pitch
of a sine wave of frequency 440 Hz. The sound pressure level (SPL) is a unit for
( Section 7.1) expressing the amplitude of a sound wave
relative to the threshold of hearing. Its unit
An octave pitch increase corresponds to a is the decibel. (Section 10.3)
doubling of frequency. (Section 8.1)
Equal increments of SPL are heard as
Pitches which are one or more octaves apart approximately equal increments of
are heard musically as different instances loudness. (Section 10.3)
48 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 48
Activity 7
The following are the correct pairings:
1 and (c)
2 and (b)
3 and (a).
The following is the correct text:
Sound waves are pressure waves, because they consist of cyclical
changes of pressure.
Sound waves emanating from a single source in the open (away from
buildings etc.) are travelling waves because the pressure variations radiate
outwards from their source, conveying energy away from the source.
Sound waves are longitudinal waves, because the molecular
oscillations are along the line of travel of the wave.
Activity 8
It took two complete cycles, and a little bit more, to create this pattern.
To see why, notice that the right-hand fork is at the centre of the low-
pressure part of the cycle, when the prongs are closest. To the right
there are two other low-pressure zones, so to create this pattern two
complete cycles were needed (Figure 30), plus a bit extra to account
for the region to the right of the diagram.
one cycle one cycle
‘extra’
Activity 11
(a) Each cycle of the pressure wave occupies 1 metre, so this is the
wavelength.
(b) We cannot directly read a single wavelength here, but four cycles
occupy 1 metre, so the wavelength is 0.25 metre.
49 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 49
Activity 12
It travels 300 metres. To see why, recall that in the time it takes for one
cycle of the fork the wave travels one wavelength, which is 1.5 metres.
So in the time required for 200 cycles the wave travels 200 times as
far, which is 200 × 1.5 metres = 300 metres.
Activity 14
With calibrated graphs like these, it makes sense to measure a period
from the point where the curve crosses the horizontal axis to the
corresponding point on the next cycle.
(a) It is clear from this graph that one cycle takes one second. Hence
the period is one second.
(b) In this graph it is not so easy to read the time for a single cycle.
However, five cycles clearly take 0.1 second, so a single cycle
would take one-fifth of this. The period is therefore 0.02 second.
Activity 15
The sine wave with a period of 1 second has a frequency of 1 Hz. The
sine wave with a period of 0.02 second has a frequency of (1/0.02) Hz,
or 50 Hz.
Activity 16
To members of Group A, Group B appears to be lagging by 0.2 seconds.
It takes 0.1 second for the sound to travel 34 metres from A to B at a
speed of 340 metres per second, and a further 0.1 second for the sound
to travel from B back to A. This makes a total delay of 0.2 second for
the round trip.
Activity 19
The wavelength is less than 1 metre. Using λ = v/f, we can see that the
wavelength in metres is 340 ÷ 384. Without using a calculator this can
be seen to be less than one. In fact its value is about 0.89 metres.
Activity 21
The delay is 0.2 ms. Figure 31 is a repeat of Figure 13.
pressure
near fork
0 1 ms 2 ms time
(a)
pressure
at a distance
0 1 ms 2 ms time
(b)
One way to answer the question would be to look at where the first
peaks occur in each graph and to try to read the time difference. This
50 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 50
is not easy because neither peak occurs on a grid line, and finding the
exact summit of a smoothly curving sine wave is not easy. In cases like
this it makes more sense to compare the points where the graphs cross
the horizontal axis at the end of corresponding cycles. In Figure 31(a)
the end of the first cycle occurs at 1 ms. In Figure 31(b) the end of the
corresponding cycle occurs at the next vertical grid line after 1 ms. As
there are five grid lines between 1 ms and 2 ms, the space between
each line represents 0.2 ms. Hence a delay of 0.2 ms.
Activity 22
The phase difference is 0.2 milliseconds, and the period is 1
millisecond, so the phase lag is 0.2 of a cycle, or one-fifth of a cycle.
Activity 24
The phase difference is half a cycle, or 180 degrees. We could equally
say the waves were one-and-a-half cycles apart, or two-and-a-half, and
so on, which would give phase differences of 540 degrees and 900
degrees respectively. However, it is customary to speak of a phase
difference like this as 180 degrees.
Activity 26
(a) The amplitude is 0.35 volts. The tops of the peaks fall half way
between 0.3 and 0.4 volts, hence the value of 0.35 volts.
(b) The amplitude is 6 volts.
Activity 27
(a) The amplitude is half the peak-to-peak amplitude, so the answer
is 1 volt.
(b) The r.m.s. amplitude is 0.71 of the answer in (a), so the answer is
0.71 volts.
(c) The r.m.s. amplitude is the size of a steady source that would
deliver energy at the same rate, other things being equal. So the
steady source needs to be 0.71 volts.
Activity 30
(a) False. The sum of A and B will have double the amplitude of A or
B because A and B are in phase. A doubling of amplitude does not
produce a doubling of loudness.
(b) This is true.
(c) True. The ear is most sensitive around 4 kHz.
Activity 32
The pitch three octaves above this A has a frequency of 3520 Hz.
Unless you are familiar with this type of calculation, it is sensible to
do it an octave at a time, as follows.
One octave above has a frequency of 2 × 440 Hz = 880 Hz.
Two octaves above has a frequency of 2 × 880 Hz = 1760 Hz.
Three octaves above has a frequency of 2 × 1760 Hz = 3520 Hz.
The reason for doing the calculation a step at a time is to avoid a
couple of traps that can easily be fallen into. First, note that a three
octave rise does not correspond to a tripling of frequency. Secondly,
51 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 51
Activity 33
(a) Figure 24 showed that the frequency increments become smaller
as we go down in pitch, so the bottom end of the keyboard (the
left-hand end) would be where the keys were closest together.
(b) It is safest to do this calculation in stages. The width of the second
octave would be 2 × 20 cm = 40 cm. The width of the third octave
would be 80 cm, and the width of the fourth octave would be
160 cm. The total width would therefore be:
20 cm + 40 cm + 80 cm + 160 cm = 300 cm, or 3 metres
Activity 34
Ten octaves. A simple approach is to divide 20 000 Hz repeatedly by
two until we reach the lower limit of the human frequency range.
The number of times the division can be carried out is the number of
octaves. Thus, starting at the upper end of the range:
20 000 Hz ÷ 2 = 10 000 Hz
10 000 Hz ÷ 2 = 5000 Hz
5000 Hz ÷ 2 = 2500 Hz
2500 Hz ÷ 2 = 1250 Hz
1250 Hz ÷ 2 = 625 Hz
625 Hz ÷ 2 = 312.5 Hz
312.5 Hz ÷ 2 = 156.25 Hz
156.25 Hz ÷ 2 = 78.125 Hz
78.125 Hz ÷ 2 = 39.0625 Hz
39.0625 Hz ÷ 2 = 19.53 Hz
Hence the span is ten octaves.
Activity 35
One octave, corresponding to a halving of the upper-frequency limit.
Activity 36
It is above the frequency of the pitches produced by virtually all
instruments. However, it is within the range of the harmonics of some
instruments.
Activity 39
(a) The overall amplification is 12 dB + 18 dB = 30 dB. From Table 1
this can be seen to be equivalent to a ratio of 32:1. Multiplying the
amplifications of the individual stages confirms this. The first
gives an amplification ratio of 4:1 and the second an amplification
ratio of 8:1.
(b) The overall amplification ratio is 80:1. Table 1 does not give a
decibel equivalent for this, but it must be the sum of the decibel
equivalents for each stage. From Table 1 these are 18 dB and 20
dB, so the overall amplification is 18 dB + 20 dB, or 38 dB.
52 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 52
Activity 40
From Table 1 or from the graph, an amplitude ratio of 1000:1 is
equivalent to 60 dB. This is the sound pressure level of this sound.
Activity 42
Table 1 does not give us a decibel equivalent directly for an amplitude
ratio of a million to one. However, Table 1 shows that a ratio of 1000:1
has a decibel equivalent of 60 dB. Hence 1000 000, which is 1000 ×
1000, has a decibel equivalent of 60 dB + 60 dB, which is 120 dB. This
sound pressure corresponds to a jet taking off when heard from a
distance of 100 m.
53 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 1 SOUND BASICS 53
LEARNING OUTCOMES
After studying this chapter, you should be able to:
Acknowledgement
Grateful acknowledgement is made to Robert Harding Picture Library/
Alamy Images for permission to reproduce Figure 12 [St Marks
Cathedral].
55 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 55
Chapter 2
Sound Shape and Colour
CONTENTS
Aims of Chapter 2 56
1 Introduction 57
2 Waveshape and timbre 58
3 Fourier’s theorem 59
3.1 Introduction 59
3.2 The harmonic series 60
3.3 The effect of phase 62
3.4 Synthesis and analysis 64
3.5 Summary of Section 3 64
4 Frequency spectrum 65
4.1 Line spectrum 65
4.2 Bandwidth 66
4.3 Summary of Section 4 67
5 The problem of synthesis 67
5.1 The problem of timbre solved? 67
5.2 The evolving sound 68
5.3 Summary of Section 5 72
6 Repetition rate and fundamental frequency 72
6.1 Summary of Section 6 75
7 The missing fundamental 75
7.1 Summary of Section 7 76
8 Formants 77
8.1 Summary of Section 8 78
9 Pitches in the harmonic series 79
9.1 Introduction 79
9.2 Identifying the pitch classes 80
9.3 Consonance and dissonance; the triad 82
9.4 The complete harmonic series 84
9.5 Summary of Section 9 85
10 Intervals in the harmonic series 86
10.1 The perfect fifth 86
10.2 The perfect fourth 87
10.3 The major third 87
10.4 Other intervals 88
10.5 Summary of Section 10 88
56 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 56
AIMS OF CHAPTER 2
Q To introduce the concept of a timbre.
Q To investigate some of the factors that affect timbre, principally
frequency spectrum, attack and decay and formant.
Q To introduce the harmonic series.
Q To introduce Fourier’s theorem and to show the role of harmonics
in creating non-sinusoidal periodic waves.
Q To introduce the concept of a time-varying frequency spectrum and
show its relevance to timbre.
Q To introduce the concepts of attack, steady state, decay and
formant.
Q To show how the harmonic series can be used to define pitches and
intervals.
Q To show some of the anomalies that arise from using the harmonic
series to define pitches and intervals and to show how equal
temperament overcomes them.
Q To introduce the phenomenon of beating.
57 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 57
1 INTRODUCTION
Much of Chapter 1 was concerned with sound waves and their
properties. These properties were broadly of two sorts: objective and
subjective. Three objective properties were found to be particularly
important: frequency, amplitude and phase. Two of these, frequency
and amplitude, were found to be closely related to audible, subjective
properties, namely pitch and loudness.
I mentioned in passing that the sounds produced by conventional
musical instruments, if plotted as graphs, are seldom pure sine waves;
rather, the sounds are mixtures of sine waves. In this chapter I shall be
looking at this idea more closely. The timbre of instruments is related
to the non-sinusoidal nature of the sound waves they produce. Timbre
is the characteristic sound, or ‘colour’, of an instrument or voice which
enables us to differentiate it from other instruments or voices even
when they are playing the same note. For instance, a trumpet sounds
very different from a clarinet. Timbre turns out to be a very complex
phenomenon, partly related to the particular mixture of sine waves
produced by instruments, and partly related to a number of other
factors, not least of which is the playing style of the performer.
A large part of the chapter will be concerned with the harmonic series.
This is a commonly occurring set of frequency relationships found in
the mixture of sine waves produced by many instruments. The harmonic
series is important not only in the story of timbre but also as the basis
of some of the pitches used in musical scales in many cultures.
The word ‘harmonic’ will occur frequently in this chapter. If you are a
player of a stringed instrument, you will be familiar with the word
‘harmonic’ to denote a way of playing which achieves a particular
timbre through the player lightly touching the string at certain places,
rather than by pressing it against the instrument’s neck. This meaning
of the word harmonic is different from the one that will feature in this
chapter, although the two meanings are not unrelated.
58 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 58
0 1 2 3 4 5 6 7
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
time (milliseconds)
Figure 3 Violin playing G3
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38
Notice that, apart from the flute (Figure 1), all these waveforms are
distinctly non-sinusoidal.
3 FOURIER’S THEOREM
3.1 Introduction
The French mathematician Joseph Fourier (1768–1830, roughly
contemporary with Beethoven) made important discoveries regarding
periodic waves, which have profound implications for the analysis of
sound, and indeed for innumerable other branches of science,
mathematics and engineering. He discovered that a non-sinusoidal
periodic wave can be created by adding together sine waves of
appropriate frequencies, amplitudes and phases. This observation is
usually referred to as Fourier’s theorem. Another way to think of
Fourier’s theorem is that a non-sinusoidal periodic waveform is
equivalent to a number of sine waves of appropriate frequency,
amplitude and phase added together. The amplitudes of the sine waves
that are added can be different from each other (in fact, they usually
60 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 60
are), but they are constant; that is, they do not change as time passes.
The frequencies and phases of the sine waves are similarly constant as
time passes.
A few restrictions must be placed on the interpretation of the phrase
‘non-sinusoidal periodic wave’ for Fourier’s theorem to work, because
not every conceivable non-sinusoidal periodic wave can be regarded as
a combination of sine waves. In principle, Fourier’s theorem applies to
non-sinusoidal periodic waves which satisfy both of the following
criteria:
1 infinite duration,
2 not too many gaps, sudden jumps, or other discontinuities.
The first condition looks impossible to meet, but turns out not to be a
major obstacle in practice. Provided the duration of a periodic wave is
long in relation to the duration of one cycle, then effectively the wave
is infinitely long for the purposes of Fourier’s theorem.
The second condition in the above list is designed to exclude rather
strange waves of a kind that do not really occur in music. Provided a
periodic wave shape is sufficiently well formed to be capable of being
drawn or printed on a piece of paper, then it satisfies the second
condition. In practice, then, the two restrictions above do not pose a
problem for the use of Fourier’s theorem in connection with real,
periodic musical waves such as those in Figures 1 to 4.
The following activity gives you the chance to see and hear for yourself
what happens when harmonically related sine waves are added.
200 Hz
0 5 10 15 20 time (ms)
(a)
400 Hz
(b)
(c)
200 Hz
0 5 10 15 20 time (ms)
(a)
400 Hz
(b)
(c)
Comment
I am sure that you found the sounds to be identical. However, this does
not mean that the effects of phase change are always inaudible. This is
still a somewhat controversial area, but you would probably be able to
hear the effects of a transition from the wave at the bottom of Figure 5
to the wave at the bottom of Figure 6 during the transition. Q
4 FREQUENCY SPECTRUM
4.1 Line spectrum
In the last section I discussed the idea that a non-sinusoidal periodic
wave has a harmonic composition, that is to say, it is composed of
harmonics. The complete set of harmonics that make up a periodic
wave is known as the frequency spectrum (or sometimes just
spectrum) of the wave.
It is useful to have a simple, graphical way to represent a frequency
spectrum. To specify a spectrum fully we need to be able to represent
three kinds of information:
1 The frequencies present.
2 Their amplitudes.
3 Their phase relationship.
Because the effects of phase relationship tend to be inaudible (except
in special circumstances), we can afford to drop the third type of
information when looking for a simple way to represent the frequency
spectrum of sound. Information relating to the first two types of
information can be captured in a particular type of frequency spectrum
graph in which harmonics are represented as vertical lines. Figure 7 is
an example.
amplitude
Figure 7 shows that there are only two harmonics in this wave form, at
frequencies of 200 Hz and 400 Hz. The vertical lines at these frequencies
have the same height, indicating that these harmonics have equal
amplitude.
A diagram like Figure 7 can have various names. Sometimes it known
as a frequency spectrum (because it shows the frequencies present).
Sometimes it is referred to as an amplitude spectrum, because it shows
the amplitudes of the harmonics but not their relative phases. (Bear in
mind that a complete representation of a frequency spectrum would
have to show both the amplitudes of the harmonics and their relative
phases, but we are choosing to ignore the phase information.)
As it happens, Figure 7 is also a line spectrum, because the harmonics
are shown as discrete lines. Any spectrum consisting of sine waves of
discrete frequencies (whether harmonically related or not) must
consist of isolated lines, as in Figure 7. However, sometimes frequency
spectra may consist of a continuous spread of frequencies rather than
discrete frequencies. Such spectra are usually associated with
unpitched sounds, for example a cymbal crash or the sound of the
wind. The important point to bear in mind is that not every frequency
spectrum graph consists of isolated lines. Phase information is
sometimes represented as a separate diagram.
66 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 66
4.2 Bandwidth
The concept of a frequency spectrum naturally gives rise to the concept
of frequency bandwidth, or just bandwidth. This is the range of
frequencies over which a frequency spectrum extends. In the case of
line spectra like Figure 7 the bandwidth is simply the highest
frequency present minus the lowest. In Figure 7 this is
400 Hz – 200 Hz = 200 Hz
As you can see, bandwidth is measured in the unit of frequency, hertz.
For frequency spectra that do not consist of isolated lines, bandwidth
must be defined rather differently, but it still conveys the same
essential idea, namely the range of frequencies over which a spectrum
is extended.
Comment
I expect you found that the pre-set values gave a sound that was not
very convincing as a simulation. All the same, you probably also found
that changing the settings did not give any significant improvement. Q
0 2 4 6 8 10 12 14 16 18 20 22 24 26
time (milliseconds)
Figure 9 Flute in low-register (F4)
Figure 10 shows the envelope of the whole of the oboe note of which
Figure 2 shows a part. (The arrow in Figure 10 shows from which part
of the overall waveform Figure 2 was taken.)
attack 35 ms
Figure 2
Oboe playing C#
time
Figure 10 5
Three distinct phases can be distinguished in Figure 10: the attack, the
steady state and the decay. Very many instrumental and vocal sounds
(but not all) have a three-part structure.
During the attack phase (sometimes also called the onset) the note is
establishing itself and growing rapidly in amplitude. Although the
attack phase is relatively short, it is nevertheless an extremely
significant part of an instrument’s timbre.
After the attack, there is typically a steady-state part. However, as
Figure 10 shows, the term ‘steady state’ can be something of a
misnomer, as the wave may vary quite a lot during this portion. The
cause of the variation may be voluntary or involuntary. Very few
instrumentalists or singers can maintain a note with absolute
steadiness, and even with mechanical instruments, such as the organ,
where you might expect the steady state to be absolutely steady, there
is often a degree of unsteadiness. In addition, many instrumentalists
and singers habitually add vibrato to a note. Vibrato is a periodic
variation in the amplitude or frequency of a waveform, typically with a
period of about 0.2 or 0.3 seconds (which is much longer than the
period of the waveform itself). The amount of vibrato added to a note is
a matter of taste and style, and at different historical periods different
amounts of vibrato have been used. However, even supposedly vibrato-
less performers often turn out to be using a small amount of vibrato.
The final phase is the decay (sometimes called offset). This is much
less significant than the earlier two phases in determining the timbre.
Depending on the instrument, the decay phase may or may not be
under the player’s control.
70 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 70
time
time
0 5 10 15 20 25 30 35 40 45
time (milliseconds)
Figure 13 From the attack phase of an oboe
attack 5 ms
decay 2 s
Figure 4
time
Figure 14 Envelope of a piano sound
200 Hz
0 5 10 15 20 time/ms
(a)
400 Hz
(b)
(c)
Part (b) of the last activity would have shown you that the result of
combining this series of frequencies
200 Hz 300 Hz 400 Hz 500 Hz 600 Hz 700 Hz 800 Hz
74 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 74
Comment
(a) With only even harmonics present, you see two cycles of the
combined waveform, as in Figure 16.
(b) The two cycles of the combined wave have the same period as two
cycles of the second harmonic, Figure 17.
Hence the repetition rate of the combined wave is the same as the
frequency of the second harmonic. Q
then the combined wave has a repetition rate of 200 Hz, not
100 Hz. It is not difficult to see why this should be the case.
Although the above frequencies are all harmonically related to a
fundamental of 100 Hz, being respectively the second, fourth, sixth
and eighth harmonics, they are also harmonically related to a
fundamental of 200 Hz.
8 FORMANTS
In Section 5 I mentioned that the wave shape produced by an
instrument was apt to differ depending on where in its range the
instrument was being played. In terms of frequency spectra, we would
say that the frequency spectrum changes through the instrument’s
range. In one sense the frequency spectrum must change, because
when an instrument plays a high note its frequency spectrum is
further to the right on a frequency spectrum graph than when it plays
a low note (Figure 20).
amplitude
0 100 200 300 400 500 600 700 800 900 1000
frequency (Hz)
Figure 21 A flat frequency spectrum
The spectrum is described as flat because all the harmonics have the
same amplitude, which is a rather unrealistic state of affairs. Suppose
this spectrum is subjected to a process which amplifies the amplitudes
of all harmonics between 750 Hz and 950 Hz by a factor of two (6 dB).
This region of frequency is the formant region for this particular case.
As a result of this process, the spectrum now has the appearance of
Figure 22.
78 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 78
amplitude
0 100 200 300 400 500 600 700 800 900 1000
frequency (Hz)
Figure 22 The spectrum of Figure 21 showing the
effect of a formant region from 750 Hz to 950 Hz
ACTIVITY 20 (SELF-ASSESSMENT)
(a) When the fundamental was 100 Hz, which harmonics were
emphasised by the formant?
(b) When the fundamental was 200 Hz, which harmonics were
emphasised by the formant? Q
We have already seen that the second harmonic belongs to the same
pitch class as the first harmonic, because they are one octave apart.
The following activity is designed to identify other members of this
pitch class.
You may have realised during the last activity that every even-
numbered harmonic must be one or more octaves above an odd-
numbered harmonic. In other words, as you ascend the harmonic
series, every time you meet an even-numbered harmonic you
encounter a member of a pitch class that has already been introduced
by a lower-numbered odd harmonic. For instance, the sixth harmonic
belongs to the same pitch class as the third, so the sixth harmonic does
not introduce any new pitch class to the series.
*
*
G2 D4 B4 F5 A5 C 6 E6 F 6
Middle C * *
harmonic number
1 3 5 7 9 11 13 15
Apart from the pitch G2, none of these pitch correspondences is exact.
That is to say, the pitch of the fifth harmonic (say) is not exactly the
same as the pitch B4 as you would find it on a modern piano. The
difference arises because, as I hinted earlier, the harmonic series is not
in practice used to define the pitches to which most modern
instruments are tuned. We shall be looking at this in more detail in
Section 11. For harmonics 3, 5, 9, 13 and 15, the discrepancy in pitch
between that supplied by the harmonic series and tuning of a modern
piano is relatively small, and many listeners would either not notice it
or would have to listen very carefully to hear a difference.
It is a different story with harmonics 7 and 11, marked with an asterisk
in Figure 23. Here there are very noticeable differences between the
pitches supplied by the harmonic series and those used in modern
tuning, and the pitch names that are allocated (F5 and C # 6 ) to these
harmonics are approximations. (Alternatively, you could say that the
pitches used in modern tuning are only approximations to the pitches
supplied by the harmonic series.)
The series of pitch classes in the last activity is not quite enough to
give us a scale of G, which is G, A, B, C, D, E, F # and G, because it
lacks C. Actually it would not be difficult to establish the pitch class C
via a harmonic series on a different fundamental, because any series
that had G as its third harmonic would have C as its fundamental. It
looks, then, as though the harmonic series could, with a certain
amount of octave transposition, be used to give us the pitches of a
standard major scale. In principle, therefore, the idea that our
subdivision of the octave derives from the harmonic series is not
implausible. However, as a practical method of deriving all the pitches
for, say, tuning a keyboard instrument, this approach leads to
problems. For example, if we use the series in Figure 23 to define the
82 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 82
Comment
Whether you regard these pitches in combination as dissonant is a
subjective matter. However, most people regard them as consonant. Q
The pitch class of the third harmonic is regarded as being the most
consonant in relation to the fundamental after the octave. The pitch
class of the third harmonic is therefore regarded as being specially
important in music. The pitch class represented by the fifth harmonic
is consonant in relation to both the fundamental and the third
harmonic. It is also particularly important in music.
The pitches you used in Activity 24 were the first three in Figure 23,
reproduced here in Figure 24.
G2 D4 B4
Middle C
harmonic number
1 3 5
If you play these pitches on a piano or any other instrument, the result
will sound very different from what you heard in Activity 24. This is
because the pitches you play on an instrument will not be pure sine
waves, whereas the harmonics you were adding in the activity
consisted of pure sine waves.
The pitch classes corresponding to the first three odd harmonics are
used to create the most basic type of chord used in music, the major
triad. A triad of G major is formed by playing the pitches G, B and D in
their closest possible arrangement, which is that shown in Figure 25.
I have not put subscripts on the pitches
because this arrangement of pitches is a triad
wherever it is played on the keyboard.
G B D
Any selection of pitches drawn from the pitch classes of the triad,
such that each class is represented by at least one pitch, is a major
chord. For instance, this set of pitches is a chord of G major: G2, G3, G4,
D5, G6, B6; but it is only one of innumerable chords of G major that can
be assembled from the pitch classes G, B and D. A major triad is thus
the most basic form of a major chord. As you would have heard in
Activity 25, a major chord and its corresponding major triad share a
characteristic sound. Thus in the published scores of a lot of popular
music, chords are represented in shorthand by letters, such as G (to
mean a chord of G major), C (a chord of C major), D (a chord of D
major), etc. It is left to the player to decide the particular arrangement
of pitches from the corresponding triads to use.
In the next activity we start to explore the higher harmonics.
Comment
Generally adding the seventh harmonic to the other three would be
regarded as producing dissonance, although for many people the
dissonance is mild. Q
84 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 84
Activity 26 shows that not all the pitch classes in the harmonic series
are consonant with the fundamental, or with each other (to check this
point you might try listening to just the fifth and seventh harmonics,
with none of the others present).
Comment
You probably found the ninth and eleventh harmonics dissonant in
relation to just the first, third and fifth harmonics. Adding the seventh
harmonic as well would generally be regarded as increasing the
overall level of dissonance. Q
G2 D4 B4 F5 A5 C 6 E6 F 6
Middle C * *
G3 G4 D5 G5 B5 D6 F6 G6
*
odd harmonics
1 3 5 7 9 11 13 15
even harmonics
2 4 6 8 10 12 14 16
Table 1 shows that the frequency ratio for this interval is 3:2, or 1.5:1,
meaning that the frequency corresponding to the higher pitch is 1.5
times that of the lower pitch.
Notice in Figure 27 that the fifth is spanned by five white notes. These
five white notes are the first five notes of the scale of G, and this is the
origin of the name ‘fifth’. A perfect fifth, however, does not have to
begin and end on white keys, and not all groups of five adjacent white
keys span a perfect fifth. Unless you are conversant with music theory,
the surest way to check whether an interval is a perfect fifth is to
count the semitone steps.
Perfect fifths occur elsewhere in the harmonic series than between the
second and third harmonics. Any two members of the harmonic series
whose frequencies are in the ratio 3:2 (or 1.5:1) are a perfect fifth apart.
In fact, any harmonics whose harmonic numbers are in the ratio 3:2
are a perfect fifth apart. For example, harmonics 6 and 4 have
frequencies in the ratio 6:4, which is the same as 3:2. Alternatively,
you could say that the frequency of harmonic 6 (frequency 6f1) is 1.5
times that of harmonic 4 (frequency 4f1). Hence sixth and fourth
harmonics are a perfect fifth apart.
87 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 87
(a) G B
(b) B D /E
(c) D /E G
Figure 29 Three major thirds
within one octave
Notice that in Figure 29, (b) and (c) have been chosen such that each
begins where the one before ends. In other words, the lower note in (b)
was the upper note in (a), and the lower note in (c) was the upper note
in (b). Thus three interlocking major thirds span exactly one octave, a
fact we shall make use of in Section 11.
The major third occurs elsewhere in the harmonic series than between
the fifth and fourth harmonics. Any frequencies in the ratio 5:4 (or
1.25:1) are a major third apart.
The intervals of a major third and a perfect fifth are used in the major
triad. Taking the home note of the triad as G, for instance, the note
above it is B, which is a major third above G, and the top note is D,
which is a perfect fifth above G.
G
Figure 30 Keyboard for Activity 34
If we take the pitches visited in part (a) of Activity 34 and regard them
as representatives of their respective pitch classes, we can arrange
them in ascending order of name, to produce this sequence of pitch
classes:
G – G# – A – A# – B – C – C# – D – D# – E – F – F# – G
This series of pitch classes covers all the pitches of a chromatic scale,
a chromatic scale being one where all the white and black notes are
sounded between the lower and upper notes of the scale. Thus by
starting from an arbitrary pitch and rising through a chain of perfect
fifths we can derive all twelve pitch classes of the chromatic scale.
This assumes that we regard enharmonic pairs, such as G# and Ab, as
equivalent.
You might be wondering how these pitch classes relate to those
derivable from the harmonic series in Figure 26. For the pitch classes
G, D and A (the first three from the ascending chain of fifths) the two
methods give the same result. All the other pitch classes, however,
differ slightly. Thus there is actually more than one way to derive a
scale of pitches from the harmonic series.
90 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 90
ascending pitch
C
G /A G /A
E E
C
The first sound is an exact major third (using the ratio from Table
1). It consists of the pitches C and E played simultaneously, as
represented by (a) in Figure 31. The second sound is an exact major
third with a lower note of E, as in (b). However, in the third sound,
as represented by (c), the upper C is not an exact major third above
the lower note of the pair. Instead, it is an exact octave above the C
in (a). Although the major thirds in (a) and (b) sound harmonious,
that in (c) sounds harsh. Q
12 EQUAL TEMPERAMENT
The previous section has demonstrated the audibility of the anomalies
arising from the use of the ratios in Table 1 to define intervals. Can
anything be done to rescue the situation? In general, musicians have
compromised in various ways, and indeed it is possible to make
sufficient adjustments on most string and wind instruments during
playing to correct for certain amounts of ‘out-of-tuneness’. With a
keyboard instrument such as the piano, organ or harpsichord,
however, this is not possible, since the pitch of the notes has been set
by the person who tuned the instrument. Normally, there is just one
key that has to serve, for example, as both G# and Ab. How, therefore, is
the pitch of the note which is played by this key to be determined?
There does not seem to have been much of a problem in Mediaeval
and Renaissance times, because only a restricted range of keys (that is,
home notes) was in use, and so you tuned the notes as was most
appropriate for your commonly used keys. In the case of our G#/Ab
example, the note would have been biased towards G#, because that
note was employed in the commonly used key of A minor, while there
was very little use of the keys in which the combination of Ab and C
was required.
In later periods various tuning compromises have been adopted which
have entailed putting some intervals out of ‘perfect’ tuning. For
example, some major thirds might be true (that is, in a ratio of 5:4),
whereas other major thirds would not. The ideal was to keep as many
intervals as true as possible. This process of adjusting pitches and
intervals away from their true values is called tempering. Many kinds
of tempering bias the tuning towards a particular key or home note. In
such systems a set of pitches that sounds acceptable in music with G
as its home note might sound less acceptable with A as its home note.
These difficulties became more acute as composers began to use a
wider range of modulation – that is, changing key or home note within
a piece. As this happened, the carefully set bias for the original key
became a cacophony if your music moved to more distantly related
keys. (A distantly related key was one in which the notes of the scale
had very few in common with the scale of the original key.) All kinds
of tunings were worked out, in theory and in practice, to accommodate
the musical needs as they developed. The particular combination of
pitches in a tuning system is known as a temperament.
Towards the nineteenth century, musicians’ needs were developing in
a direction that required complete freedom of modulation – that is, the
ability to move from any major or minor key to any other major or
minor key in the course of a piece of music. This was accommodated
by a system of equal temperament, in which the ‘out-of-tuneness’ was
spread equally around the pitch-relationships. In equal temperament,
only the octaves remain perfectly in tune: every other interval is
deliberately ‘out of tune’ to some degree. Perfect fifths are all flattened
by a regular amount: they are narrowed, in fact, from a ratio of 3:2 to
2.9966:2. One consequence of this is that twelve perfect fifths do now
become equivalent to seven octaves. Major thirds, on the other hand,
are expanded; instead of a ratio of 5:4 they have a ratio of 5.04:4 (or
93 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 93
13 BEATS
13.1 Close frequencies
In the last section, mention was made of a ‘throbbing’ sound that might
be heard when intervals are not quite perfectly in tune (that is, not
quite as given by the harmonic series). The term used for this effect is
beating. It is most readily demonstrated using sine waves with pitches
that are close in frequency, but it is not exclusively associated with
sine waves nor with close-frequencies.
time
beats
smooth tone rough smooth
0 15 Hz frequency difference
f2 – f1
critical bandwidth
Figure 33 The emergence of two distinct pitches as frequency separation increases
The explanation for these phenomena lies in the operation of the ear
rather than in the physics of sound, and in particular the significance
of the critical bandwidth in Figure 33 will become clear later, in
Chapter 5, when we look at the working of the ear. (The critical band
has an upper an lower limit; the lower limit is to the left of the zero in
Figure 33.)
The phenomenon of beats and roughness created by closely spaced
notes played simultaneously has many musical consequences. For
instance, sine waves at the frequencies of G2 (98 Hz) and G # 2 (103.83 Hz)
are close enough for beats to be very audible when they are played
simultaneously. Whether the beats are audible in a musical context
depends very much on the instrument(s) playing the notes. If they are
played simultaneously on the piano the throbbing is not so easy to
hear; however, on a guitar, for example, beats can be heard and felt by
97 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 97
SUMMARY OF CHAPTER 2
Timbre is the characteristic sound of an Graphical representations of frequency
instrument or voice. (Section 1) spectra are examples of frequency-domain
representations. Waveform graphs plotted
The sound waves produced by instruments against time are time-domain
are usually not sinusoidal. The non- representations. (Section 4.2)
sinusoidal character of the wave is related
to the instrument’s timbre. (Section 1) A complete representation of a frequency
spectrum would contain information about
A harmonically related series of frequencies frequencies, amplitudes and phases.
has this pattern: (Section 4.2)
f1 2f1 3f1 4f1 5f1 ....
Frequency f 1 is called the fundamental A frequency bandwidth is the range of
frequency. (Section 3.2) frequencies in a spectrum. For a line
spectrum, the bandwidth is the highest
frequency minus the lowest frequency.
Sine waves with harmonically related
(Section 4.3)
frequencies are called harmonics. Harmonics
are numbered: the first harmonic has
frequency f 1 , the second harmonic has Synthesising an instrumental timbre
frequency 2f1, and so on. (Section 3.2) convincingly is complex, and generally
requires more than simply adding
Fourier’s theorem states that, with some harmonics according to a simple scheme.
restrictions, a non-sinusoidal periodic wave (Section 5.1)
can be created by combining harmonics with
unvarying amplitudes. The non-sinusoidal The wave forms produced by musical
periodic wave so created has the same period sounds typically have an attack phase
as the first harmonic. (Sections 3.1 and 3.2) (onset), a steady-state phase and an decay
phase (offset). The attack phase is very
Changing the relative amounts of the characteristic of an instrument’s or voice’s
harmonics changes the shape and timbre of timbre. (Section 5.2)
the resultant periodic wave. (Section 3.1)
The frequency spectrum of a musical sound
Changing the phase relationship of is typically a time-varying spectrum.
harmonics changes the shape of the resultant (Section 5.2)
wave, but providing the phase relationship
is not changing, the effect is inaudible. Periodic non-sinusoidal waves are said to
(Section 3.3) have a repetition rate rather than a
frequency. (Section 6)
Combining harmonics to create a non-
sinusoidal periodic wave is called Fourier Repetition rate is the reciprocal of the
synthesis. Analysing a non-sinusoidal period, and is measured in hertz. (Section 6)
periodic wave into its constituent harmonics
is called Fourier analysis. (Section 3.4) An incomplete harmonic series may be
harmonically related to more than one
A frequency spectrum is a range of frequency. (Section 6)
frequencies. (Section 4.1)
A series of harmonically related
A frequency spectrum can be represented as frequencies from which the fundamental
a frequency spectrum graph (also known as frequency has been removed may be heard
an amplitude spectrum). (Section 4.1) to have a pitch corresponding to the
missing fundamental frequency. This is the
The frequency spectrum of a periodic non- phenomenon of the missing fundamental.
sinusoidal wave is a line spectrum. (Section (Section 7)
4.1)
Harmonically related sine waves combine
In a line spectrum, each line represents a to create a fused tone which has a pitch
harmonic, and the height of the line indicates corresponding to the frequency of the
the amplitude of the harmonic. (Section 4.1) fundamental. (Section 7)
99 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 99
A formant is a fixed range of frequencies The major third is the interval between the
over which harmonics are emphasised. fifth and fourth harmonics. It has a
(Section 8) frequency ratio of 5:4. It is also the interval
between the home note of a major scale and
Instruments and voices have characteristic the third note of the scale. (Section 10.3)
formants. (Section 8)
A chain of interlocking perfect fifths can
A pitch class is a set of pitches that share the supply twelve pitch classes of the
same letter name. (Section 9.2) chromatic scale. (Section 11.1)
The odd numbered pitches of a harmonic A chain of ascending perfect fifths with the
series can be used to define several pitch ratio 3:2 eventually arrives at a pitch which
classes relative to the pitch class of the is slightly higher than seven octaves above
fundamental. (Section 9.2) the starting pitch. A chain of three
ascending major thirds using the ratio 5:4
The seventh and eleventh harmonics are covers slightly less than one octave.
markedly out of tune in relation to the rest (Sections 11.1 and 11.2)
of the series. (Section 9.2)
Adjusting the frequency ratios for intervals
The pitch class of the third harmonic away from those of the harmonic series is
corresponds to the fifth note of a major called tempering. (Section 12)
scale. (Sections 9.3 and 10.1)
Most schemes of tempering (i.e. most
temperaments) favour certain keys over
The pitch class of the fifth harmonic
others. (Section 12)
corresponds to the third note of a major
scale. (Sections 9.3 and 10.3)
Equal temperament removes the anomalies
inherent in using the simple ratios of the
A major triad consists of the home note (the harmonic series for intervals of less than
tonic), the third note and the fifth note from an octave. It treats all keys equally,
a major scale. (Sections 9.3, 10.1 and 10.3) allowing free modulation from one key to
any other. (Section 12)
A major triad is the most basic form of the
major chord. A major chord consists of at In equal temperament, the only intervals
least one representative of each of the pitch whose frequency ratios agree exactly with
classes of the major triad. (Section 9.3) those of the harmonic series are octaves.
Enharmonic pairs have the same pitch. The
An interval is a step in musical pitch. The semitone step is defined by a constant
harmonic series can be used to define multiplicative factor of the twelfth root of
frequency ratios for intervals. ( Section 2 ( 12√2). (Section 12)
10.1)
Two sine waves with close frequencies f1
The perfect fifth and the major third are and f2, when played simultaneously, are
especially important musical intervals. heard as a single pitch that beats at the rate
(Sections 10.1 and 10.3) equal to the difference between the two
frequencies. (Section 13.1)
The perfect fifth is the interval between the
third and second harmonics. It has a Beating is also heard with frequencies that
frequency ratio of 3:2. It is also the interval are close in ratio to those in the harmonic
between the home note of a major (or series. Beating is sometimes heard in the
minor) scale and the fifth note of the scale. intervals of equal temperament. (Section
(Section 10.1) 13.2)
100 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
time (milliseconds)
Figure 34 One cycle of the violin wave form
From the horizontal axis we see that one cycle takes just over 5 ms. To
get a more accurate answer would require careful measurement of the
amount by which the separation is greater than 5 ms. In fact, careful
measurement shows that the period is certainly in the region of 5.1 ms, so
within the tolerance to be expected when reading such graphs, we can
say that the period is the same as in part (a). In fact, it is often (but not
always) true that the period of a wave produced by an instrument is
the same as the period of a sine wave tuned to the same pitch.
Activity 3
Yes, all these frequencies are whole-number multiples of 150 Hz.
Activity 4
The first six terms in this series are these multiples of 98 Hz.
(1 × 98 Hz), (2 × 98 Hz), (3 × 98 Hz), (4 × 98 Hz), (5 × 98 Hz), (6 × 98 Hz)
These work out as follows:
98 Hz, 196 Hz, 294 Hz, 392 Hz, 490 Hz, 588 Hz
Activity 5
No, these frequencies are not harmonically related to a fundamental of
100 Hz. For instance, there is no whole number by which 100 Hz can be
multiplied to give any of the frequencies in this series apart from the first.
Activity 6
This fundamental frequency is 100 Hz, so the next four above 800 Hz
are 900 Hz, 1000 Hz, 1100 Hz and 1200 Hz.
Activity 9
A tuning fork’s waveform is a pure sine wave (to a very close
approximation), so there is only one frequency in its spectrum, at
440 Hz. Hence the spectrum is as shown in Figure 35.
101 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 101
amplitude
10
Activity 10
The highest frequency present is 450 Hz. The lowest frequency is
75 Hz. Hence the bandwidth is (450 – 75) Hz = 375 Hz
Activity 16
Given a fundamental frequency of 200 Hz, the first four harmonics
have the following frequencies:
200 Hz × 1 = 200 Hz (first harmonic)
200 Hz × 2 = 400 Hz (second harmonic)
200 Hz × 3 = 600 Hz (third harmonic)
200 Hz × 4 = 800 Hz (fourth harmonic)
Thus harmonics 1 to 4 are present in the frequencies given.
Activity 19
When the fundamental is 200 Hz, the harmonics are at frequencies of
200 Hz, 400 Hz, 600 Hz, 800 Hz and 1000 Hz. The effect of the formant
is to emphasize the 800 Hz harmonic. Figure 36 shows the result.
amplitude
0 100 200 300 400 500 600 700 800 900 1000
frequency (Hz)
Figure 36 Effect of the formant on a harmonic series
with fundamental 200 Hz
Activity 20
(a) With a fundamental of 100 Hz, the eighth and ninth harmonics
were emphasised.
(b) With a fundamental of 200 Hz, the fourth harmonic was emphasised.
Activity 21
The fourth harmonic, with frequency 4f1, is an octave above the second
harmonic, and therefore two octaves above the fundamental. The eighth
harmonic, 8f1, is three octaves above the fundamental, and the sixteenth
harmonic, 16f1, is four octaves above the fundamental, and so on.
102 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 102
Activity 22
There are eight pitch classes. As we go up through the series adding
harmonics, we can ignore the even-numbered harmonics because they
do not contribute new pitch classes. Only the odd-numbered
harmonics contribute new pitch classes, and there are eight of those.
Activity 23
To establish the pitch classes from Figure 23, we simply take the note
names, remove the subscripts, and arrange the letter-names in
ascending order. This is the result:
G, A, B, C# *, D, E, F* and F# .
Activity 28
The twenty-second harmonic would be out of tune. It is an octave
above the eleventh, which is marked with an asterisk.
Activity 30
There are seven
semitone steps.
Figure 37 shows
them.
1 3 6
Figure 37
2 4 5 7
Seven semitone steps
comprise a perfect fifth G D
Activity 33
(a) A perfect fifth has a frequency ratio of 3:2, which is the same as
1.5:1. The upper note of a perfect fifth therefore has a frequency of
1.5 times the frequency of the lower. The lower frequency here is
6f1, so the upper frequency must be 9f1. Hence the ninth harmonic
is a perfect fifth above the sixth.
(b) In this case the upper frequency is 1.25 times the lower (because
the frequency ratio is 5:4, or 1.25:1). The lower frequency is 12f1,
so the upper frequency is 15f1. Hence the fifteenth harmonic is a
major third above the twelfth.
(c) No. The tenth harmonic has a frequency of 10f1. The frequency a
major third above this is 1.25 × 10f1, which is 12.5f1. This does not
appear in the harmonic series, being between the twelfth and
thirteenth harmonics.
Activity 34
(a) These are the pitches passed through on the way:
G – D – A – E – B – F # – C # – G # – D # – A # /B b – F – C – G
You may have given some of these pitches their alternative ‘flat’
names. Figure 38 shows these perfect-fifth steps. If your answer was
G D A E B F C G D A /B F C G
Figure 38
103 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 103
different, you might like to check that there are seven semitone steps in
each of these perfect-fifth intervals.
(b) As Figure 38 shows, it takes twelve perfect fifths in succession to
return you to the note G.
Activity 35
There are seven octaves from the bottom G to the top G.
Activity 36
Seven octave steps bring you to the figure of 128 (which is 27). Twelve
steps of a perfect fifth in a 3:2 ratio gives you approximately
129.74634. If you went up though twelve ‘perfectly’ tuned fifths, you
would therefore end up with a note that was sharper in pitch than if
you went up through seven octaves.
Activity 37
Three steps of the ratio of 5:4 (or 1.25:1) starting from 1 Hz gives a
frequency of
1 Hz × 1.25 × 1.25 × 1.25 = 1.953 Hz
A single octave step gives a frequency of 2 Hz.
Activity 40
Any whole-tone step from Table 2 can be used. Taking the ratio of
D4:C4 we get 293.66:261.63, which is about 1.122425:1. The ‘just’ ratio
of 9:8 is equivalent to 1.125:1. Thus the equally tempered whole-tone
step is slightly smaller.
Activity 43
(a) The frequency difference is 3 Hz, because this is the rate of
beating. Hence the higher frequency must be 183 Hz.
(b) The second harmonic of wave A has a frequency of 180 Hz. It is
the lower frequency of a pair that beats at 3 Hz, so the higher
frequency must be 183 Hz. The 183 Hz frequency is that of the
third harmonic of wave B, so the fundamental frequency of wave
B must be 183 Hz ÷ 3 = 61 Hz.
104 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 2 SOUND SHAPE AND COLOUR 104
LEARNING OUTCOMES
After studying this chapter, you should be able to:
1 Explain correctly the meaning of the emboldened terms in the main
text and use them correctly in context.
2 Derive the period and repetition rate of a non-sinusoidal periodic
wave from its graph. (Activity 2)
3 Explain in simple terms Fourier’s theorem, and explain the
distinction between Fourier synthesis and Fourier analysis.
4 Recognise whether a series of frequencies is harmonically related.
(Activities 3, 5)
5 Calculate the frequency of any harmonic given a fundamental, and
calculate the fundamental given a harmonic. (Activities 4, 6, 16)
6 Draw a line spectrum from appropriate frequency data, or interpret
frequency data from a line frequency spectrum. (Activities 9, 10)
7 Calculate bandwidth from a line frequency spectrum. (Activity 10)
8 Modify a line spectrum to show the effect of a formant, or deduce
the effect of a formant from appropriate frequency spectrum data.
(Activities 19, 20)
9 Discuss the main factors that affect timbre and explain why the
timbre of instruments cannot easily be recreated artificially.
10 Relate pitches to pitch classes. (Activity 23)
11 Perform simple frequency calculations involving the intervals of
the octave, perfect fifth and major third. (Activities 33, 36)
12 Identify members of the harmonic series which are related by
intervals of the octave, the perfect fifth and the major third.
(Activity 33)
13 Perform simple frequency calculations given a table of equally
tempered frequencies. (Activity 40)
14 Discuss some of the difficulties that arise when the harmonic series
is used to define musical intervals other than the octave.
15 Discuss the system of equal temperament and why it has evolved.
16 Perform simple calculations relating to rates of beating. (Activity 43)
105 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 105
Chapter 3
Sound and Time
CONTENTS
Aims of Chapter 3 106
1 Introduction 107
2 The time dimension in music 108
2.1 Relative durations 108
2.2 Allocation of time to notes 110
2.3 Beat and pulse 110
2.4 Beats and subdivisions 112
2.5 Subdividing unequally 115
2.6 Are beats always regular? 116
3 Time signatures 117
3.1 Accents and bars 117
3.2 Interpreting the time signature 118
3.3 Spacing of notes on the page 122
3.4 Variability of beats and accents 122
3.5 Compound time signatures 123
3.6 Rhythm 125
4 Expressivity in performance 126
4.1 The work of Carl Seashore 126
4.2 Categorical perception 128
5 Synchronising an ensemble 131
Summary of Chapter 3 135
Answers to self-assessment activities 136
Learning outcomes 139
106 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 106
AIMS OF CHAPTER 3
To explain the way the temporal aspect of music is represented in
notation, and how the temporal dimension of music is related to an
underlying pattern of beats and accents.
To introduce the concept of metre and the time signature as a way of
representing it.
To introduce the concept of categorical perception and relate it to the
notational categories of notated music.
To discuss expressivity in terms of categorical perception.
To discuss synchronisation of musicians in terms of underlying pulse,
and show some of the counting strategies used by musicians.
107 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 107
1 INTRODUCTION
So far we have spent quite a lot of time investigating the pitch and
frequency aspect of musical sounds. Music, however, unfolds in time;
musical notes have duration as well as pitch, and music has a tempo.
The temporal aspect of music is important in a course on music
technology for a number of reasons. A very straightforward reason is
that the complexities of human performance need to be appreciated if
one is trying to create or recreate music convincingly by electronic
means. Many of these complexities relate to the temporal aspect of
music: when notes begin and end, how they relate to an underlying
beat, and so on. Another reason is the widespread use of terminology
in music technology that relates to the temporal aspect of music, such
as ‘beat’, ‘rhythm’ and ‘tempo’. These have very particular meanings,
which I shall explain. Another issue relates to synchronisation and
co-ordination, which inevitably arise when musicians perform
together.
To discuss the temporal aspect of music I shall draw on terminology
and ideas from conventional music notation. I should, however, stress
that you do not need to be able to read music to follow the discussion
in this chapter, nor is it my intention to teach the reading of music, in
the sense of teaching how to perform it or how to transcribe it.
However, I shall spend quite a lot of time explaining the way the
temporal aspect of music is represented in notation, and by the end
you should understand some of the basic principles of the notation of
music’s temporal aspect. I shall also be using both British and
American terminology for note values. I am including American
terminology (‘quarter notes’, ‘eighth notes’, etc.) not because I think it
is generally preferable to British terminology, but simply because it
makes some of the arithmetic easier. If you are already familiar with
British terminology, you should have no problem. If you are not
familiar with either system, but find the American system easier to use
and would prefer to use it exclusively, that will be acceptable for the
purposes of this course.
I should also stress at the outset that what I will be discussing is the
modern conception of time in relation to music as it has existed in the
broadly European and American musical traditions for two or three
centuries. That may seem like a long time, but in terms of musical
history it is fairly recent. If you go back to the twelfth, thirteenth and
fourteenth centuries you find different conceptions of the temporal
and metrical aspects of music. Also, in music traditions outside Europe
and America, for instance in Java and India, you again find different
ways of conceiving the temporal aspect of music. Furthermore, I shall
only be concerned with metrical music, that is, music with a beat or
pulse (I shall be explaining these terms in the chapter). Many of the
concepts I discuss are not relevant to styles of music such as musique
concrete (music based on the manipulation of recorded natural sounds)
or electro-acoustic music.
108 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 108
time
Perhaps the first important point to grasp about the temporal aspect of
notation is that the durations of notes or silences (usually called rests)
are relative. That is to say, nothing in Figure 1 indicates directly how
long each note lasts in units of time, such as seconds or fractions of a
second. Instead, the note shapes embody a system of time
relationships. As you can see, most of the notes in Figure 1 are of this
type: q (crotchet, or quarter note in American terminology). In
addition, at the end of the line, there is one of this type h (minim, or
half note), which is preceded by one of this type e (quaver, or eighth
note), and then by one of this type ß (dotted crotchet, or dotted quarter
note). We can regard any type of note as a basic unit of time, but the
crotchet (quarter note) is commonly used. In terms of this note, the
relative durations of the other notes are as follows:
h (minim, or half-note) is twice as long as q (crotchet, or quarter-note)
e (quaver, or eighth-note) is half as long as q (crotchet, or quarter-note)
ß (dotted crotchet, or dotted quarter-note) is 50% longer than q
(crotchet, or quarter note)
The pattern of note values used here is just a selection from a larger set
of possible values. Box 1 ‘Standard note values’ gives the note values
that are in common use in notated music.
Once a basic duration for any particular note value has been decided, the
durations of all the other notes in a piece (with some provisos) are fixed.
A composer can establish the duration of a particular note value by
means of a metronome mark (see Box 2 ‘The metronome’). A
metronome mark says how many of a particular type of note there are
in a minute. For instance, a metronome mark for Figure 1 might be
written as q = 120, meaning that a hundred and twenty crotchets
(quarter notes) would last a minute, and hence the duration of a
crotchet comes out as half a second. From this, the durations of all the
109 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 109
q or Q
3 3
quarter note crochet 1/4
Various extensions of the
e or E eighth note quaver 1/8 schemes described here are used
x or X sixteenth semiquaver 1/16 to allow for other subdivision of
note notes, and also to allow for
r or R thirty-second demisemiquaver 1/32
asymmetrical subdivision, where,
for instance, a note might be
note
subdivided into unequal parts.
other notes follow. The metronome mark is a precise way of setting the
tempo or speed of a piece of music. Less precise ways of specifying
tempo involving Italian terms, such as presto (very fast), allegro (fast,
or lively), andante (moderate walking speed), adagio (leisurely), and
lento (slow) have a long tradition in classical music and are still widely
used, either in conjunction with a metronome mark or in place of one.
time
Comment
I expect you found that you were quite accurate when counting to five,
but probably less so when counting to ten. However, you may have
found that with a little practice you could markedly improve your
results. Many people also find that making a physical movement, such
as tapping their foot or a finger, enables them to improve their
accuracy. ■
Comment
You probably had no difficulty finding the beat in the first extract. With
the second example I expect you found it very difficult to tell where the
beat was, and you may have thought there was no beat at all. However,
the players and conductor of the piece would have been very well aware
of where the beat was. Because relatively few musical events coincide
with the beat, the listener cannot easily detect the background pulse. ■
I should introduce a word of caution here about the word beat, which
tends to be used in two senses in music. In one sense it means what I
have outlined above: regularly recurring, instantaneous markers of
time (for instance, hand claps, drum beats, clicks). In its other sense,
beat means the time interval between two consecutive markers of time.
It is important to grasp that the word has these two meanings, which
may occur within one sentence, for instance when a musician says
something like, ‘Start on the third beat and hold the note for four
beats.’ I will use both senses in this chapter, though it should be clear
from the context which is meant. Figure 4 summarises these two
meanings of ‘beat’.
beat 1 beat 2 beat 3
pulse
markers
1 beat
time
Figure 4 ‘Beat’ as an instantaneous time marker and as
a measure of duration
crotchet
(quarter note)
duration
time
Comment
You probably did not have much difficulty with this activity. In fact,
the ease with which you can subdivide a beat into two equal parts
probably accounts for why the most basic subdivision in conventional
music is a halving. ■
In (a) we set up a regular pattern of beats, and each beat coincides with
the commencement of a note. I am using the crotchet (quarter note) to
represent the time gap between two consecutive beats, although, as
always, any note value could be used. (The horizontal line after each
‘left’ and ‘right’ indicates that the sound of the word lasts through to
the next beat. The use of a line like this is common in notated vocal
music.)
In (b) each ‘left’ and ‘right’ lines up with those in (a), but between each
‘left’ and ‘right’ there is an ‘and’. In (b), the words are now represented
by quavers (eighth-notes). (We could also include a short horizontal
line after each ‘left’, ‘right’ and ‘and’, although this is not so necessary
when the notes are close together, as here.)
One thing that should be apparent from Figure 6, although you may
not have realised it during the activity, is that in line (b) the words
‘left’ and ‘right’ must be said more quickly to make room for
interpolated ‘ands’.
Other subdivisions of the beat, such as into three, four, or more equal
parts can be a bit tricky, but, once more, playing them or recognising
them is a matter of practice. Once the sound of a particular subdivision
has become familiar it then becomes almost second-nature for a
performer to play it. Sometimes music students use mnemonics to help
them get used to the sound of subdivisions, common ones being based
on words or phrases with the right
number of syllables to give the
required subdivision. For
instance, saying the words
‘higgledy, piggledy, higgledy,
piggledy, ...’ with one word at each
step while walking (or at each
hand clap) gives you a regular
three-part or triplet division of a Figure 7 Subdividing a beat equally
into three
beat, as in Figure 7.
In music where the beat is
frequently subdivided into
triplets, the notation can be made
tidier by using a dotted note to
represent the beat. Almost
invariably a dotted crotchet
(dotted quarter note) is used, as in
Figure 8. This gets rid of the need
Figure 8 Use of a dotted crotchet
to put a figure ‘3’ above or below to represent the beat simplifies
the triplet groups. the notation of triplet subdivision
To help get a regular four-part
division of the beat, the word
‘caterpillar’ can be used, as shown
in Figure 9.
each crotchet into two quavers (eighth notes). Each pair is beamed
together. I have emphasized the equality of the subdivisions by the
putting a 1:1 ratio under each beamed pair, although a ratio such as
this does not form part of the musical notation.
Now, to make this subdivision unequal, we can either extend the first
note of each pair at the expense of the second, or extend the second
note at the expense of the first. These two schemes are adopted
respectively on the left and the right of Figure 10. Thus, on the left-
hand side of Figure 10, at (c), the first note is now twice as long as the
second (indicated by the 2:1 ratio), whereas on the right-hand side the
second note is twice as long as the first (indicated by the 1:2 ratio).
The pattern is continued in (d). On the left, the two notes are now in
the ratio 3:1 whereas on the right they are in the ratio 1:3.
As with even subdivisions, standard patterns such as these (and a few
others ) are very common and have a characteristic sound which
musicians learn to recognise. The 2:1 ratio at (c), for instance, when
repeated in consecutive pairs of notes, has a characteristic ‘rumty
tumty rumty tumty’ sound that occurs in much British (and especially
English) folk music. The notation for this subdivision in Figure 10(c)
is rather awkward, as you probably noticed. There is a simpler way to
notate this subdivision which I shall mention later in connection with
compound time signatures.
The important point to grasp here is that there are standard patterns of
subdivision, which are very common and have characteristic sounds.
These standard patterns are based on simple numerical ratios. For the
purposes of this course you are not expected to be able to recognise
the sound of these subdivisions or to be able to perform them.
Comment
Segovia was famous for his flexible attitude to the beat, and to the
subdivision of the beat, but his flexibility was not unusual when this
recording was made (1927). Many musicians of his era played with a
fairly flexible pulse, and this was considered to be musically
expressive. Nowadays performers tend to be less flexible. However, in
folk music and other non-classical genres you can still often hear quite
variable pulses. ■
117 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 117
3 TIME SIGNATURES
3.1 Accents and bars
When presented with a sequence of identical beats, equally spaced,
people generally hear them as unequally accented. A familiar example
is the ticking of a clock. The ticks of a clock are all equal, but are
usually differentiated by the listener as tick, tock, tick, tock, .... The
listener interprets the sounds as forming a recurring two-beat cycle:
‘tick, tock’.
With a little imagination, the same ticking of a clock can be heard as a
three-beat cycle: ‘tick, tock, tock, tick, tock, tock, ....’ Four-beat cycles
and five-beat cycles can also be imagined, and other numbers of beats.
What defines the start of each new cycle is a feeling of accentuation on
the first beat of the cycle, making that beat a ‘strong’ beat. A cycle,
then, consists of a pattern of strongly and weakly accented beats. For a
two-beat cycle it goes like this:
S W S W S W
where ‘S’ stand for ‘strong’ and ‘W’ for ‘weak’. For a three-beat cycle
the pattern goes like this:
S W W S W W
For a four-beat cycle it goes like this:
S W M W S W M W
(Here ‘M’ stands for ‘medium’, an accentuation intermediate between
strong and weak.)
It is usual to attach count-numbers to the beats of a cycle. For a two-
beat pattern, the count goes like this:
1 2 1 2 1 2
A two-beat cycle is often found in marches, the two beats
corresponding to ‘left’ and ‘right’.
A three-beat cycle is counted as follows:
1 2 3 1 2 3, etc
Such a pattern is found in waltzes and minuets.
A four-beat pattern, predictably enough, is
1 2 3 4 1 2 3 4
What all these patterns have in common is that beat 1 is the strongest
of each cycle
The question of what an accent actually consists of is complex. The
example of the ticking clock shows that listeners can hear an accent in
sounds that are identical, and that they can choose to hear
accentuation in different ways. However, in actual musical
performance everyone should ideally be hearing the same pattern of
accentuation. A sense of a strong beat can be created in several ways,
one of which is to increase the loudness, but loudness is not the only
way. Anything that draws attention to a particular beat could be said to
118 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 118
Figure 12 shows the first few bars of Figure 11. At the start of the staff,
there is a clef, a key signature and a time signature.
119 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 119
key signature
I do not want to say too much about the first two, except to say that the
clef identifies the second line from the bottom of the staff as
representing the pitch G (specifically G4, using the pitch notation
introduced in Chapter 2). Hence the pitches of all the other lines and
spaces can be derived from it. Clefs other than the G clef are in use.
They assign other pitches to the lines or spaces of the staff.
The key signature defines the key, that is to say, the scale from which
the notes of the music are taken. Here it consists of two sharp signs,
but other numbers of sharps and flats are possible. (A key signature
never mixes sharp and flat signs.) The sharp signs here are attached to
the line corresponding to the pitch F5, and to the space corresponding
to a pitch of C5. This key signature means that all members of the pitch
class F (not just the F5) are sharpened by a semitone, and all members
of the pitch class C (not just C5) are sharpened by a semitone, unless
there is some instruction to the contrary later in the music. These two
pieces of information tell the performer that the music is in the key of
D major or B minor, because these are the only two keys where only the
notes F and C are consistently sharpened. (The scale of D, for instance
is D, E F#, G, A, B, C#, D.) As it happens, this piece is in D major rather
than B minor, but this fact can only be established either by listening
to the music (major and minor keys sound distinctly different) or by
looking at whether the note D or B acts as a ‘home’ note for the piece.
The symbol I want to concentrate on here is the time signature,
because it tells us two vital pieces of information:
1 How many beats there are per bar.
2 What note value has been selected to represent the beat (or, if you like,
what note value has a duration equal to the space between two beats).
120 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 120
The time signature for this piece is 4/4, said as ‘four-four’. Another
way to express the same idea is to say that the metre of the piece is
four-four. Notice in Figure 11 that the time signature is not repeated at
the start of each staff, although the clef and key signature are.
Although a time signature looks like a fraction, it is not a fraction in
the usual arithmetical sense. Nevertheless, for our purposes you can
think of the time signature being the result of a fraction multiplication,
like this:
4 1
= 4×
4 4
This way of rewriting the time signature brings out the two pieces of
information listed above. First of all, we have the number of beats per
bar. This is the multiplying term, 4. There are four beats per bar in this
piece. The second piece of information is the note value chosen to
represent the beat. It is, in American terminology, the quarter note, or
crotchet.
A time signature or 4/4 is extremely common in all musical styles,
especially in popular music, show tunes, and jazz. In fact a metre of
four-four is often referred to as common time. Other very common time
signatures are 3/4 (used in waltzes) and 2/4 (much used in marches),
and 2/2, 3/2 and 4/2 (often used in hymns). Time signatures of 5/4 and
7/4 are relatively much less common, though quite often used in East
European folk music. Note that the American name ‘quarter note’, is
not derived from the fact that there are four of them in a bar of 4/4. The
name derives simply from the fact that it is a quarter of the longest
commonly encountered note value.
Bars 4, 8, 10, 11 and 16 of Figure 11 show that not all bars in this piece
consist of four crotchets (quarter notes). However, the durations of the
notes in each of bars 4, 8, 10, 11 and 16 add up to the same duration as
occupied by four crotchets (quarter-notes). You can check this with a
little arithmetic.
If we take bars 10 and 11, each consists of this sequence of note values:
q e e q q
Using American terminology the time 1 1 1 1 1 4
+ + + + =
value of this sequence is easily calculated: 4 8 8 4 4 4
So these bars are each equivalent to a bar of four quarter notes, or
crotchets.
1 1 1 1 1 4
The addition sum for this bar is + + + + =
4 8 8 4 4 4
121 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 121
Beethoven’s tune in Figure 11 begins on the first beat of the bar, but
plenty of tunes begin at a different part of the bar. Figure 13 shows a
familiar tune that begins on the last beat of the bar.
In Figure 13 I have added a couple of rests before the tune starts to clarify
where in the bar the tune starts. However, rests such as these are
generally not added to an incomplete bar at the start of a piece of music.
One reason why it can be hard to tell a piece’s time signature from the
sound alone is that there is generally more than one way in which the
music can be notated. For instance, Figure 14(a) and (b) would sound
identical. The notational difference is due to the use of a different note
value to represent the beat. In (a), a crotchet (quarter note) is used; in
(b), a minim (half note) is used.
(a)
(b)
(c)
(b)
Figure 15
Spacing of notes gives only
an approximate indication
of their relative durations
sign, >. For instance, a bar of music in four-four might have this
symbol over a normally weak beat, such as the fourth, to indicate that
the note falling on this beat should be accented (Figure 16).
Figure 16 Accent
on a weak beat
Beethoven achieves this very effect in bar 12 of Figure 11, but without
using this symbol. Instead he ties this normally weakly accented note
to the one following, which is normally strongly accented, as shown in
Figure 17.
tie
Figure 17 Tying notes together
The curved line joining the last note of bar 12 to the first of bar 13
makes them into a single note with a combined length of the two notes
added together. (Naturally only notes having the same pitch can be
tied.) Because of the tie, in bar 13 there is no note beginning on the
first beat, and thus no note to receive a strong accent. In effect, the
accent that would normally fall here is transferred forward to the weak
beat at the end of bar 12, as in Figure 18.
Figure 18 A syncopation
Besides being used when the prevailing subdivision of the beat is into
three rather than two, compound time signatures are frequently used
when the beat is often subdivided into unequal parts in the ratio of 2:1
(or 1:2), as in Figure 10(c). Figure 19 shows how the notation works
when a dotted crotchet (dotted quarter note) represents the beat.
(a) beat
The dotted crotchet (dotted quarter note) which represents the beat in
Figure 19(a) is equivalent to three quavers (eighth notes), as in
Figure 19(b). If we tie the first two notes together (Figure 19c), we have
125 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 125
effectively subdivided the beat into two parts which are in a ratio of 2:1.
The two tied quavers (eighth notes) can be replaced by a crotchet
(quarter note), as in Figure 19(d).
3.6 Rhythm
In connection with Beethoven’s ‘Ode to Joy’ tune in Figure 11 we
noticed that not all bars consisted of four crotchets (quarter notes),
although all the bars had a 4/4 metre. For instance, bars 4, 8, 10, 11 and
16 consist of other note values than four crotchets (quarter notes). We
express this idea by saying that these bars have a different rhythm from
that of a bar of four crotchets (quarter notes). However, the rhythms of
bars 4, 8 and 16 are identical, and the rhythm of bar 10 is the same as
that of bar 11. A rhythm is thus the pattern of note values used in any
section of music, usually together with their metre.
The concept of rhythm does not apply to just one bar’s worth of note
values. There is, for instance, a distinctive rhythm in the two-bar group
formed by bars 12 and 13 arising from the syncopation. Similarly, we can
speak of the two-note pair in Figure 19(d) as having a distinctive rhythm.
4 EXPRESSIVITY IN PERFORMANCE
4.1 The work of Carl Seashore
The modern temporal notation of music, as you have seen, divides
time into beats and subdivisions of the beat. A note either falls on a
beat or not on a beat. If it does not fall on a beat then it falls half-way
between the beats, or on some other relatively simple numerical
subdivision of the beat such as two thirds or one quarter.
In using a relatively limited set of categories of note values and
placements, the temporal representation in notation is a counterpart to
pitch representation. In the representation of pitch a note has to be one
pitch or another, F# or G, for instance, and the set of available pitches
is finite and quite limited. But what really happens in the performance
of music? Are these categories of placement and pitch an accurate
representation of what musicians actually play? I have already hinted
that in an expressive performance there might be some flexibility of
pulse, but are the other parameters (temporal placement and pitch)
also more flexible than the notation indicates?
The answer is yes, but the extent to which they are flexible was not
fully appreciated until the pioneering work of an American musical
psychologist called Carl Seashore in the 1930s. Using equipment that
now seems quite crude, he analysed recorded performances of singers
and instrumentalists from which he was able to make precise
measurements of pitch, loudness and time. He found that the
apparently rigid framework of notation is approximate to a degree that
surprised even many highly skilled musicians.
Figure 20 is based on one of Seashore’s analyses.1 The top line is the
first twelve notes of the folk tune ‘All though the night’. Do not worry
if you cannot read music; you should be able to follow the analysis.
Above the staff, the notes have been numbered, and the words, or lyrics
added above the numbers. (The usual place for the lyrics of a song is
below the staff, but it is more convenient here to put them above.)
This section of the tune uses only five pitches, F, G, A, Bb and C, and
the lines and spaces corresponding to these pitches have been labelled
at the left of the staff. Unusually, the horizontal placement of the notes
on the staff in Figure 20 accurately represents the temporal placement
of the notes indicated by the notation. You can check the horizontal
placement of the notes by referring to the short vertical lines below the
staff, which represent evenly spaced beats. As you can see, there are
four per bar, and note 2, for instance, lies half-way between the second
and third beats of the first bar.
C 1 2 3 4 5 6 7 8 9 10 11 12
B
A
G
F
beats
C
B
A
G
F
Figure 20 ‘All through the Night’ in conventional notation (top) and a performance
analysed by Carl Seashore (bottom)
Comment
The first note, nominally B b , varies almost between B and A, almost a
whole tone. Note 3, at its widest deviation, varies almost between G #
and F # , again almost a whole tone. ■
Regarding pitch, notice that the average value of the pitch during vibrato
is not always the nominal pitch. For instance, during note 7 the average
pitch is lower than the nominal pitch of the note (A), and during note 8
the average pitch is rising. Notice also that notes 2 and 3, and 7 and 8, run
into each other. From the diagram it is hardly possible to say where note 2
ends and 3 begins (for instance). In addition, several notes are approached
from below their nominal pitch. Notes 9 and 10, in particular, start below
their nominal pitch and rise. (At note 11, Seashore’s singer performs a
version of the tune where this note is given the same pitch as note 12.)
As far as the temporal placement of the notes goes, if the singer were
giving a literal interpretation of the note values, then the notes in this
bottom part of the diagram would be aligned with the notation in the
upper part of the diagram. Notice that although notes 1 to 4 fall close to
where they would be expected to fall, notes 5–10 are all begun early.
One thing that is immediately clear from Seashore’s results is that what
happens in a real performance is not as rigid as conventional notation
indicates. The beginnings and endings of notes are not always clearly
defined, nor are their pitches, and the placement of notes can also be
quite flexible.
128 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 128
5 SYNCHRONISING AN ENSEMBLE
When more than one musician is involved in a performance of a piece,
the question of synchronisation naturally arises. How do two performers
ensure that their notes coincide? The potential for non-synchronisation
escalates as more musicians are involved until, with a large choir or
orchestra, there may be in the region of 70 to 100 performers.
The conductor (if there is one) has a score, or, more correctly here, a
full score, which shows what every performer should be doing at any
instant. Figure 22 shows one page of the score from part of Beethoven’s
Ninth Symphony. (The first two bars of Figure 11 are in the final two
bars of Figure 22.) I have indicated the instruments or voices to which
each staff applies. Numbers in brackets indicate the number of
instruments playing each staff, except for the choral staves and string
staves, where the precise number of performers is not indicated by the
composer. (The first and second violins of a modern symphony
Flutes (2)
Oboes (2)
between
Clarinets (2) beats 2
and 3
Bassoons (2)
Double
Bassoon
French
Horns (4)
Trumpets (2)
Timpani
vertically
First Violins aligned
Second
Violins
Violas
Soprano
Alto
Tenor
Baritone
Sopranos
Altos
Tenors
Basses
Cellos and
double basses
Figure 22 Page of the full score of Beethoven’s Ninth Symphony (from M. Unger, Eulenburg)
132 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 132
However, if the music is being taken fairly slowly, then the beat as
indicated by the notation may be too slow to count comfortably. For
instance, if the extract shown in Figure 24 is played so that each group
of four notes takes two seconds or longer, the performer may increase
the counting rate, and count in the places indicated by ‘X’ in Figure 25,
or even, in extreme cases, count each note individually. This is known
as dividing the beat.
On the other hand, if the music is played quickly, the performer may
not count each beat. For instance in fast music in 4/4, the performer
may count in the places shown in Figure 26.
The disparity between what a musician may count, and what the time
signature indicates the beat to be, indicates that what is meant by ‘beat’
is not entirely straightforward. Notational subdivisions of the beat can
come to be regarded as beats in slow music; and notational beats can
come to be regarded as subdivisions in fast music. Behind this
phenomenon lies the experimentally observed fact that people are
most likely to feel comfortable tapping their feet at around one beat
every 0.6 seconds. So when conductors are conducting, or musicians
are counting, or dancers are dancing, they will tend to respond to
repetitive events in the music that are happening at about this rate, and
have a physical sensation that this is the beat. Whether these beats are
actually the same as the beat indicated in notation (if the musicians are
playing from notation) depends very much on the speed at which the
music is performed.
SUMMARY OF CHAPTER 3
In music notation, note durations are Except for compound time signatures, the
relative. Durations are notated using a upper number of a time signature represents
hierarchical system of note values which the number of beats per bar, and the lower
represent successive doublings or halvings number, interpreted as the denominator of
of duration. In American terminology, note a fraction, indicates the note value that
names indicate relative durations (half note, represents the beat. (Section 3.2)
quarter note, eighth note, etc.) This system
can be adapted to accommodate other A change to the pattern of accentuation
subdivisions than equal, binary within a bar can be indicated by the use of
subdivisions. (Section 2.1) an accent symbol (>). A syncopation occurs
when a strong accent is shifted to a
A metronome mark can be used to specify normally weakly accented beat. (Section
an absolute duration for a particular note 3.4)
value, and hence for any other note value.
A metronome mark is a precise way of Compound time signatures are used where
indicating tempo. (Section 2.1) the principal subdivision of the beat is into
three parts rather than two, or where the
Music conventionally has a beat or pulse. prevailing subdivision is into two parts
Beats are regularly spaced, instantaneous with durations in the ratio of 2:1. In music
temporal markers, to which musical events of this kind, a dotted note represents the
are related. The word ‘beat’ also refers to beat. Most commonly, a dotted crotchet
the time interval between two consecutive (dotted quarter note) is used. (Section 3.5)
temporal markers. In much music, audible
repetitive events coincide with the beat, Modern musical notation has evolved to
making the beat an explicit part of the represent certain ways of arranging the
music. In other music, the beat may be temporal and pitch aspects of musical
inaudible. (Section 2.3) material. Performances are typically much
more flexible than the notation suggests.
The durations of notes are conceived in Notation appears to represent the categories
terms of the underlying beat rate. A note of musical perception of its originating
lasts for a certain number of beats, or culture. (Sections 4.1 and 4.2)
fractions (subdivisions) of a beat.
Commonly used subdivisions of the beat The expressive quality of music appears to
are based on simple numerical ratios and reside in its flexibility in relation to the
have characteristic sounds. (Section 2.4) categories of organisation used in notation.
(Sections 4.1 and 4.2)
A stream of unaccented beats will begin to
seem regularly grouped if listened to for Acceptable computer-based transcription
more than a few seconds. Groups of evenly or performance requires intelligent
spaced beats appear to be set off from one computer systems that are able to interpret
another by equally spaced accented beats. or generate expressive effects correctly.
Regular beats fall into repeating cycles of (Section 4.2)
strong and weak accentuation. Each cycle
begins with a strong accent. A cycle of Synchronisation of ensembles is achieved
accentuation is a bar. (Section 3.1) through matched counting by the
performers. If the performers are playing
The time signature is a way of indicating from notation, their counts may or may not
how many beats there are in a bar, and what coincide with the notational beats,
note value represents the beat (in the sense depending on the tempo of the
of the time interval between two beats). performance. (Section 5)
136 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 136
Activity 2
(a) There are 90 crotchets (quarter notes) per minute, and hence 180
quavers (eighth-notes) per minute. So in five minutes there will be
900 notes.
(b) There are 120 quavers (eighth notes) per minute, so each one lasts
half a second.
(c) A minim (half note) lasts four times as long as a quaver (eighth
note), so each lasts for 4 × 0.5 second, or two seconds.
Activity 7
A beat is one of a series of regularly spaced, instantaneous temporal
markers, for example the clicks of a metronome or the silent markers
inside a performer’s head. The word is also used to represent the time
between two such temporal markers. This length of time can serve as a
temporal unit. By using a particular note value to represent this
temporal unit, the durations of other note values can be specified in
terms of beats, either as whole numbers or fractions of a beat.
Activity 9
This is a three-beat pattern:
⏐1 2 3 ⏐ 1 2 3 ⏐ 1 2 3 ⏐
⏐S W W ⏐ S W W ⏐ S W W ⏐
Activity 11
Each of these begins with a dotted crotchet (dotted quarter note), ß
which is 50% longer than a crotchet (quarter note). In other words, this
note is worth
1 1
+
4 8
The whole bar is therefore
⎛ 1 1⎞ 1 1 4
⎜ + ⎟+ + =
⎝4 8⎠ 8 2 4
Activity 12
All bars will have an equal total of note values and rests, so, as long as
the duration of each note value remains the same, the duration of each
bar will remain the same. However, the duration of the note values might
not remain the same, for instance if the music speeds up or slows
down. If there are no tempo changes, the bars are all the same length.
137 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 137
Activity 15
Twenty-four bars of music in 6/8 has forty-eight beats (two beats per
bar). If forty eight beats are to last for 30 seconds, then in a minute
there would be 96 beats. So the music must be played at 96 beats per
minute. Thus the metronome mark must be ß = 96.
Activity 17
As usual, the use of American terminology helps considerably.
(a) A time signature of 2/4 means that there are two beats per bar, and
the beat is represented by a crotchet (quarter note). Thus the
duration is one beat.
(b) A quaver (eighth note) is half a crotchet (quarter note), so an
undotted quaver (undotted eighth note) lasts half a beat. Adding a
dot extends the duration for 50%, so the duration of one dotted
quaver (dotted eighth note) becomes three-quarters of a beat. For
two such notes the duration is one-and-a-half beats.
Activity 18
A given series of notes would have a different rhythmic character if
played in one metre rather than another. Four instance, four crotchets
(quarter notes) played in 4/4 are rhythmically different from four
crotchets (quarter notes) played in 3/4 because of the differing
arrangements of strong and weak accents.
Activity 19
Patterns of regularly accented beats are naturally grouped into
repeating cycles of strong and weak accentuation. A bar is a cycle of
accentuation, beginning with a strong accent.
Activity 20
The time signature is a way of indicating how many beats there are in
a bar, and what note value has been chosen to represent the beat (in
the sense of the time interval between two beats). In non-compound
time signatures, the upper number of a time signature represents the
number of beats per bar, and the lower number indicates the note
value that represents the beat.
Activity 21
(a) In compound time signatures, a dotted note represents the beat
(usually a dotted crotchet or dotted quarter note). In non-
compound time signatures an undotted note represents the beat.
(b) Compound time signatures are used where the principal
subdivision of the beat is into three parts rather than two, or
where the prevailing subdivision is into two parts with durations
in the ratio of 2:1.
Activity 24
Both the Seashore-style diagram and conventional notation are
graphical representations of music. The Seashore diagram has a chart
showing how pitch varies (and another showing how loudness varies).
As with conventional notation, the Seashore pitch chart represents
pitch in the vertical direction and time along the horizontal axis. The
138 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 138
Activity 27
(a) Categorical perception is a psychological concept. It is the process
by which newly experienced phenomena are assimilated to pre-
existing categories, which they may not exactly fit. The mismatch
between the phenomenon and the category is, however,
overlooked until such time as it leads to unignorable anomalies.
(b) Musical pitch is one aspect. The categories are the named pitches,
and, in the context of a musical performance, pitches may be
assimilated to these categories. Another aspect is the series of
temporal categories produced by simple subdivisions of an
interval of time.
Activity 28
Synchronism is achieved through matched counting among the
players, for which a score is not needed. In addition, in many styles of
music, notation is not used. Even in those musical styles where notation
is used, the performers play from a part (showing their own part in the
music) rather than from a score, which shows all the parts together.
139 TA225 BLOCK 1 INVESTIGATING SOUND CHAPTER 3 SOUND AND TIME 139
LEARNING OUTCOMES
After studying this chapter, you should be able to:
1 Explain correctly the meaning of the emboldened terms in the main
text and use them correctly in context.
2 Perform simple calculations based on the relative durations of
minims (half notes), crotchets (quarter notes), quavers (eighth
notes), semiquavers (sixteenth notes), and their dotted forms, given
a table of standard note values. (Activity 1)
3 Interpret a metronome mark and perform simple calculations based
on it, given a table of standard note values. (Activity 2)
4 Explain how durations of notes can be related to beats and their
spacings. (Activity 7)
5 Explain what a bar is and show the patterns of accentuation in bars
of two beats, three beats and four beats. (Activities 9 and 19)
6 Show that the note values (and rests) in a bar add to the correct
amount in relation to the time signature, given a table of standard
note values, and be able to relate the duration of a bar to its
constituent notes (and rests). (Activities 11 and 12)
7 Interpret the duration of a note (dotted and undotted), or a series of
notes, in terms of a number of beats, given a time signature and a
table of standard note values. (Activity 17)
8 Explain the use of a time signature (Activities 20 and 21).
9 Interpret common time signatures (2/4, 3/4, 4/4, 6/8, 9/8 and 12/8)
in terms of the number of beats in the bar and the note value used
to represent the beat. (Activities 15, 17 and 20)
10 Discuss some of the findings of Carl Seashore regarding the
relationship between a notated piece of music and a performance of
the same piece. (Activity 24)
11 Describe simply the concept of categorical perception and to relate
it to aspects of music. (Activity 27)
12 Explain the relationship of a musical score to a musical part, and
explain how ensembles of performers maintain synchronism.
(Activity 28)
140 TA225 BLOCK 1 INVESTIGATING SOUND INDEX 140
141 TA225 BLOCK 1 INVESTIGATING SOUND INDEX 141
INDEX
Notes 1 This index covers Block 1, Chapters 1, 2 and 3 only.
2 Where terms are referenced in two or more places, the page number is only given in one place,
cross references are given for the other entries.
3 Page numbers in bold refer to places where the term appears emboldened in the main text.
4 The index does not cover the aims, chapter summaries, answers to self-assessment activities or
learning outcomes.
> (in music notation) see accent sign decay (of a note) 69, 72
accent (in music) 117 decibel 40, 45
strong 123 adding 42, 45
weak 123 mathematical definition (non-
accent sign (in music notation) 122 assessable) 43
air pressure 10, 18 table of equivalent ratios 41
amplification 42, 45 dissonance 82
amplifier 42 domain
amplitude 27, 29, 31 frequency see frequency domain
peak to peak see peak-to-peak amplitude time see time domain
root-mean-square see r.m.s. amplitude dotted note 108, 123
amplitude spectrum 65, 67 dynamic range 39
antiphonal 21 of human hearing 39, 40
atmospheric pressure 10
ear
attack (of a note) 69, 72
frequency response see frequency
response of human hearing
ear–brain combination 75
bandwidth 66, 67 eighth note see quaver
critical see critical bandwidth
electronic synthesis 67
bar (in music) 118
enharmonic pairs 89, 91
bar line 118
equal temperament 80, 90, 92, 94
bassoon formant 78 expression (in music) 129
bassoon spectrum 75
beam (in music notation) 109 f see frequency
beat (musical) 111 fifth (musical interval) see perfect fifth
as a time interval 112 first harmonic 64
divided 133 flanging 26
equal subdivisions of 112 flat (in music notation) 119
notational subdivisions of 133
strong 117 flute waveform 58
unequal subdivisions of 115 formant 77, 78
weak 117 bassoon see bassoon formant
beating 91, 93, 95, 97 Fourier analysis 64
Fourier synthesis 64
Fourier’s theorem 59, 64
cancellation (phase) 24 Fourier, Joseph 59, 79
categorical perception 128 fourth (musical interval) see perfect fourth
chord frequency 19, 22
major see major chord bandwidth see bandwidth
chromatic scale 89, 91 fundamental see fundamental frequency
clef (in music notation) 119 harmonically related see harmonically
related frequencies
click track 134
frequency domain 66
common time 120
frequency domain representation 67
compound time signature 124
frequency range
concert pitch 33
human hearing see frequency response –
consonance 82 human hearing
critical bandwidth 96 frequency response
crotchet 108, 112, 120, 124, 125, 132 of human hearing 33, 38, 40
cycle 12, 18 frequency spectrum 65, 67
cyclic (or cyclical) motion 12, 13 flat 77
full score 131
142 TA225 BLOCK 1 INVESTIGATING SOUND INDEX 142
fundamental metaphor 7
missing see missing fundamental metre see time signature
fundamental frequency 61, 64 metronome 109
fused tone 76 metronome mark 108
microvolt 29
Gabrieli, Andrea 21 MIDI 133
Gabrieli, Giovanni 21 millisecond 18
gain (of an amplifier) see amplification millivolt 29
graph minim 108, 112, 121
frequency spectrum 67 missing fundamental 75
groove see swing modes of vibration 67
modulation (between musical keys) 92
music
half note see minim temporal aspect see temporal aspect of
harmonic 39, 57 music
first see first harmonic musical instrument digital interface see
second see second harmonic MIDI
harmonic series 57, 60, 61, 79, 88 musique concrete 107
incomplete 75
relationship to pitch 79
table of frequency ratios and pitches 85 noise reduction 25
harmonically related frequencies 61, 64 notation (musical) 108
harmonics 61, 64 interpretation by computer 130
Helmholtz system 37 note (musical)
hertz 19 dotted see dotted note
relative duration 108
home note see tonic
note allocation 110
human ear see ear
human hearing note values 109
dynamic range see dynamic range –
human hearing oboe waveform 58, 69
frequency response see frequency octave 34, 35, 37, 79
response – human hearing octave ranges
threshold of audibility see threshold of typographical notation 37
audibility offset (of a note) see decay
onset (of a note) see attack
in phase 23 oscillation 12, 18
interference (phase) 26 oscillatory motion 12
interval (musical) 86, 88 out of phase 23
overtone 39
jazz 129, 130
part (of a full score) 132
partial 39
key colour 93 pascal 29
key signature 119 peak-to-peak amplitude 28, 29, 31
keys (musical) perception
and timbre 79 categorical, of music see categorical
kilohertz 19 perception
perception of sound 8
λ see wavelength perfect fifth 86, 88, 89, 91, 92
lag (phase) 23, 26 perfect fourth 87
lead (phase) 23, 26 period (of a waveform) 72
lift (in music) see swing period 13, 17, 19
line spectrum 65, 67 periodic motion 13, 17
longitudinal wave 12 periodic wave 59
non-sinusoidal 60
loudness 32
phase 23, 26
lyrics 126
effect of when adding harmonically
related sine waves 62
Maelzel, Johan 109 phase angle 24
major chord 83 phase difference 23, 24, 26
major scale 81, 88 phase shift 24
major third 87, 88, 90, 91, 92 phasing see flanging
major triad 83 piano waveform 58, 68, 71
143 TA225 BLOCK 1 INVESTIGATING SOUND INDEX 143
Acknowledgement
Cover images: © 1997 Photodisc, Inc., and Ingram Publishing