SOME NOTES ON ACOUSTICS,
ELECTRO-ACOUSTICS, AND SOUND SYNTHESIS
by Dr. Steve Paxton, Texas Tech University
I. Accoustics
We hear sound because of rapid and relatively subtle
changes in air pressure next to our ear drums. These are pressure
changes can be caused by doors slamming, strings vibrating, children screaming,
bells ringing, jets taking off, columns of air being disturbed (flutes,
trumpets, didjeridus, pop bottles, etc.) and by thousands of other natural
and decidedly un-natural phenomena.
These fluctuations of air pressure travel through
air in waves, just as ripples travel through disturbed water. These
waves can be measured. Such a measurement is often expressed graphically
as a wave-form, a picture or graph of how the air pressure changes in time.
All change is measurable and can be graphed.
We can measure how frequently or rapidly the changes occur, how drastic
the changes are, and how complex the changes are. In terms of the fluctuating
air pressure that causes sound, we call the three measurements frequency,
amplitude, and timbre, and they constitute the three parameters by
which all sound can be analyzed. By observing frequency (pitch),
amplitude (loudness), and timbre (tone quality), we can observe everything
that is observable about sound.
Frequency
Frequency is the measurement of how often a change occurs.
If a vibrating string causes the air pressure surrounding it to change
very rapidly, then the string itself must be changing position (vibrating,
that is) very rapidly. Psychologically, we interpret different frequencies
as different pitches. Slowly vibrating strings, columns of air, drum
heads, etc. create 'low-pitched' sounds, and rapidly vibrating strings,
columns of air, drum heads, etc. create 'high-pitched' sounds. Graphically,
in the picture of a wave-form, frequency corresponds to wave-length.
Amplitude
Amplitude is the measurement of how drastic a change
is. When you play on a swing at the park, you can make a small singing
arc, traveling just a short distance before reversing your direction, or
you can make a big arc, traveling a great distance before heading back
to where you started (and beyond -> your momentum carries you even beyond
where you started). The small-arc swing has a low amplitude, while
the big arc has a high amplitude. Psychologically, we interpret different
amplitudes of sound as having different levels of loudness, because of
how drastically the air pressure is fluctuating. Graphically, in
the picture of a wave-form, amplitude corresponds to wave-depth.
Timbre
Timbre is more difficult to visualize and understand
than frequency and amplitude, partly because it is in many ways a resultant
effect of frequency and amplitude. (In fact, all three of these parameters
of sound are inextricably linked together. It is only in analyzing
them that we even dare to think of them as being separate.)
Change can be simple or complex, in the same way
that a trip to California could be straight down I-40 or a complex combination
of side trips over county and state roads. To carry the analogy even
further, a complex, convoluted trip to California could be analyzed as
being merely the combination of many individual simple trips over these
various county and state roads. In the same way, sounds can be simple
or complex, and the complex ones can be thought of as multi-layered combinations
of the simple ones.
The simplest sound is the result of a perfectly
even and regular increase and decrease of air pressure. A test tone
on radio or TV produces such a sound. The sound of a flute or of
someone singing 'ooh' is almost as simple. If many of these frequencies
and amplitudes occurring simultaneously are mathematically related to each
other in fairly simple ways (whole-number ratios, for instance), the overall
sound will meld together and we will perceive it as a unified, single tone
of a specific fundamental pitch. Devices such as oboes, trumpets,
opera singers (usually), and sawtooth wave oscillators create such sounds.
At other times, the relationships between the many
different frequencies and amplitudes may be so mathematically complex that
we cannot isolate any one fundamental pitch, but rather we hear a wide
blur of pitches that we often describe as noise. Bass drums, jet
engines, rattlesnakes, and spoken consonants are more likely to create
these relatively more complex and noisy types of sounds. In the picture
form of a wave-form, timbre corresponds to wave-shape.
II. Electro-Accoustics
Sounds can be produced electronically because acoustics
and electricity behave in ways that are analogous to each other.
They operate in similar fashion. Levels of electrical current, or
voltage, can increase and decrease much in the same way the air pressure
can increase and decrease. If these fluctuations of voltage are introduced
to a loudspeaker (a flexible membrane attached to an electromagnet), they
can cause the loudspeaker to vibrate, It of course vibrates with
the same frequency, amplitude, and timbre (wave-form) as the fluctuating
voltage driving it. The vibrating loudspeaker causes the air pressure
adjacent to its flexible membrane to change rapidly, with, again, the same
frequency, amplitude, and timbre as the electrical current that made the
whole system start vibrating in the first place. And finally, if
our ear drum is close enough to these changes in air pressure, we hear
sound.
A synthesizer is a device that can create fluctuating
voltage, and that provides the user ways to alter the frequency, amplitude,
and timbre of those voltage fluctuations. Said another way, it generates
sound waves, and allows the user to alter the wave-forms. (Similarly,
a violin is a device that can create fluctuating air pressure, and that
provides the user with ways to alter the frequency, amplitude, and timbre
of those air pressure fluctuations.
Analog synthesizers
Analog synthesizers contain electronic circuits that
actually generate fluctuation voltage, and thereby generate sounds, as
described above. These are called oscillators. Analog synthesizers
also contain filters, devices that allow the user to alter the timbre of
sounds generated by the oscillators. Finally, they contain amplifiers,
devices used to alter the amplitude of the sounds generated
by the oscillators.
Digital Synthesizers
Digital synthesizers do not normally contain actual
oscillators, filters, and amplifiers. Instead, they contain computers
that calculate mathematical representations of sound waves. In much
the same way that a graphing calculator or a computer program can create
a wave-form on its display. These mathematically represented wave-forms
can be altered as to their frequency, amplitude, and timbre to create more
wave-forms. In fact, whole tables of these "imaginary" sound waves,
expressed only as numbers, can be stored in the memory of a digital synthesizer.
However, we can't really hear these so called waves. We can just
see the lists of numbers or the graphic diagrams that represent them.
And lists of numbers and pictures of wave-forms cannot drive loudspeakers.
Therefore, every digital synthesizer contains a rather sophisticated device
that converts numbers into actual fluctuations of voltage. This device
is called a digital-to-analog converter (DAC), and is usually the last
stage of a digital synthesizer.
III. Sound Synthesis
Most musicians, whether they work with acoustic music,
electro-acoustic music, or both, spend most of their time dealing with
frequency and amplitude. However, it is the parameter of timbre that
more subtly and pervasively inhabits everything we do as musicians.
Perhaps it is because timbre is so often beyond our control that we tend
to concentrate on it less.
The most interesting things about timbre are the
natural acoustical phenomena associated with it, specifically overtone
and the harmonic-series.
Overtones
The simplest possible sounds discussed above (test-tones,
some flute notes, etc.) exist singularly as very gradual and regular increases
and decreases of energy, whether the energy be in the form of changing
air pressure or changing voltage. We hear these very pure sound waves
as fundamental sounds, the frequency of which is very easy to focus on.
Because of this singularity, these pure tones can even be said to have
no timbre, or no color, or to be wholly lacking in timbre and color.
A graphic representation of such a pure tone, its wave-form, would be in
the shape of a sine-wave.
A more complex sound, on the other hand, like a
bass drum or a spoken consonant, consists of many simultaneously heard
sound waves, each of which is in fact a sine-wave with different frequency.
These individual components of a complex sound are called overtones or
partials, and when piled up on top of one another they form complex wave-forms.
The overtone sometimes obscure the frequency of the lowest, or fundamental
sine-wave (the frequency which the overtones are "over"), and sometimes
they reinforce it.
The Harmonic Series
When a complex sound has overtones that strongly reinforce
the fundamental sine-wave over which they are heard, they constitute a
harmonic series of overtones. The reason they reinforce the
fundamental are related to the very specific and mathematically precise
relationships that exist between their frequencies and amplitudes:
their frequencies and amplitudes relate to the frequency and amplitude
of the fundamental in whole-number ratios.
The frequency relationships are the most consistent
and easy to understand. The first overtone (also called the second
partial) is twice the frequency of the fundamental, the second overtone
(third partial) is three times the frequency of the fundamental, the third
overtone (fourth partial) is four times the frequency of the fundamental,
and son on. That means that the frequency ratios between the partials
consist of ratios such as 2:1, 3:2, 4:2, 3:1, etc.
-> all whole number ratios.
Unlike frequency, the amplitudes of the overtones
in a harmonic series differ from wave-form to wave-form. In some
instances the second partial may be one-half of the amplitude of the fundamental,
the third partial one-third the amplitude of the fundamental, the fourth
partial one-fourth the amplitude of the fundamental, and son on.
For other wave-forms, the second partial may be one-fourth, the third partial
one-ninth the amplitude of the fundamental, the fourth partial one-sixteenth
the amplitude of the fundamental, the fifth partial one-twenty-fifth the
amplitude of the fundamental, and so on. Different wave-forms contain
differing amounts of energy, or amplitude, in the upper partials, but these
amplitudes are always related to each other in whole-number ratios, in
inverse proportion to the amplitude of the fundamental: 1,2,
1:3, 1:4 in the first example; 1:4, 1:9, 1:16,
1:25 in the second.
Certain types of physical objects produce this harmonic
series of overtones when they are "excited" or disturbed, and air pressure
around them is subsequently disturbed. Such objects include strings
and columns of air. Most of the sounds traditionally used in music are
produced by instruments that use strings or columns of air, and these instruments
therefore generate a harmonic series of overtones when played.
Certain other types of physical objects do not produce
a harmonic series of overtones when they are "excited." This group
includes struck membranes and struck wooden or metal plates. Many
of these sounds also find their way into traditional music, usually by
way of instruments in the percussion family. These instruments do
produce overtones, but they are usually non-harmonic overtones. Electronic
sound synthesis, of both the digital and analog variety, produces opportunities
for musicians and sound designers to directly manipulate the frequencies
and amplitudes of overtones. This makes it possible to emulate the
sounds of acoustic instruments by duplicating the frequency and amplitude
ratios in nature, but also to create sounds that could not exist in the
natural world of acoustics: sounds whose frequency and amplitude
relationships have their genesis in the laws of mathematics or in the imaginations
if musicians or both.