The Digital Representation of

Sound
Engineering Report
Written by Richie Ung Trang
1. Abstract
The purpose of this document is to inform individuals about the
representation of sound in computing devices. Initially, the report
frames the role of software and computer engineers involved in
designing audio interfaces, as well as the academic career paths
involved. The report then examines the various conversion
processes which allows sound to be stored in computer memory.
2. Introduction
Technological development and the need for audio sharing has led
to music and sound being stored in almost every computing device
found today. The digital representation of sound has led to the music
industry becoming both a renown and established discipline. Digital
audio recordings are also required in our current telecommunication
devices such as mobile phones. However, these progressions are
not possible without first being able to condense sound in computer
memory (in binary).
The nature of work required by engineers who design audio
interfaces is a proficient understanding in mechanical waves and
computer programming. They require at a minumum, a Bachelors
degree in software or computer engineering which can be attained
at a university. The report outlines how sound is recorded and
condensed through an Analogue to Digital Converter (ADC) and the
computer interfaces involved.
3. Scope of the Profession
Software engineers employed to design computer interfaces would
work in a variety of settings including software firms, private
companies or in some cases work through entrepenuership. They
are responsible for communicating where sound is stored in local
memory through simulated programs or operating systems. They
must coordinate controls into a responsive program and convert
instructions into computer language (a process called programming
or coding). They can be expected to investigate any failures in the
program and undergo updates.
Computer engineers who design audio interfaces (the technology to
convert analogue to digital and visa versa) are often employed in
computer firms, technology firms or private companies. They are
required to build electrical products and computer hardware

interfaces. Computer engineers can be expected to create models

which will be manufactured for consumers.
4. Training for the profession
All countries require a tertiary level of academic achievement and a
set minimum credential to enroll in a B Engineering (Software) or a
B Engineering (Computer) Degree. Bachelor degrees can be attained
at universities or training at vocational schools. Predominantly,
engineers in the computer sector must have an intricate
understanding of all computer components. This course requires
extensive competency in electrical engineering and a high level a
high level of achievement to attain a MEng (Masters in Engineering).
Software engineers major in programming. They learn debugging
skills and how to develop high quality and reliable systems.
Graduates often undergo a wide range of work experience in the
field involving communication and team cooperation. This
experience is imperative before they can be employed in esteemed
corporations such as Gigabyte, HP or BOSE Australia. A small
proportion of these engineers work towards a PhD, which allows
them to coordinate major projects and work in managerial positions.
5. Investigation
Unlike numbers and images, sound as a relative quantity is difficult
to store in computers. Sound is audible by humans due to pressure
variations in the ear drums.1.1. It is then questionable how a finite
collection of values for sound can be stored and used in computing
devices. However sound is a form of wave and oscillation. Methods
to store the amplitude, frequency, wave length and other
characteristics of sound waves have been created. However, there
are still numerous processes required to condense sound into
computer memory (in binary).
Utilising an abstract approach, the general process of conversion is
as follows
1. Direct the sound at a microphone. Microphones convert
acoustical energy (pressure variations) into electrical energy
(electrical variations). Provides continuous values for current,
voltage and frequency.
2. Use a Analogue to Digital Converter (ADC). Works by recording
the instaneous voltage at a fast predetermined rate. The
device converts continuous values for voltage into discrete
binary values. Process is called quantisation
3. Store the discrete values in binary and computer memory
(bits).

1.
Microphone

2.
Analogue
to Digital
Converter
Quantisati
on

Abstract view of sound to digital conversion

3. Store
discrete
values

1. Audio recorded using a microphone

First, a microphone is necessary to convert sound5.1 into quantifiable
values. A microphone is a transducer that converts sound energy
(pressure variations) into electrical energy (electrical variations).
A microphone uses a thin rubber diaphragm5.2 that oscillates at the
frequency of the sound wave. The diaphragm also moves back and
forth a distance proportional to the loudness of the sound wave.
Cross Section of a microphones Diaphragm

Illustration of pressure waves causing the diaphragm and coil

to move
The diaphragm is connected to a coil that moves back and forth.
The coil is surrounded by electromagnets which induces a current.
As a result, an electrical audio signal is generated. Continuous
values of voltage is emitted through the audio signal.

2. Converting continuous voltage waveform into

discrete values using a Analogue to Digital
Converter

An Analogue to Digital Converter (ADC) takes the smooth voltage

waveform from the microphone and returns a list of discrete values.
This process is called quantisaton.
The electrical signal from the microphone is directed into an ADC
and the voltage is recorded at a high frequency (a process called
sampling). However, quanitsation is always an approximatation of
the original wave source. A level of accuracy however can be
achieved such that the human ear cannot tell the difference. For
accurate sampling frequencies for music, the Nyquest5.3 sampling
theorem states to sample at a rate doube the wave frequency.
Analogue waveform and its digital cousin

The analogue waveform has smooth and continuous changes. The

digital version however has the same shape but a staircase look.
This is a direct result from quantising the wave into discrete values.

Sampling in the Computing Context

Schematics of an Analogue to Digital Converter:
Includes a binary register, resistor divider network and comparator.

The purpose of the ADC is to get a value for the voltage. One way
would be to increase the voltage using circuits and comparing until
we get the correct value. However, this would take too long. So a
binary register is set up which contains values of voltage in a wide
range.

The binary register5.5 is set up so at each point of resistance, a value

for voltage is compared with the input in the binary register. If the
input voltage is greater, than a led light switches on. If less than the
binary register, then the LED5.6
switches off. The input voltage completes a cycle until a final value
for the voltage is determined. The comparator compares the
multiple signals from the LEDs and produces a one bit digital signal
to be sent and stored by the computer. The process is repeated until
a digital interpretation of the wave is achieved.5.7

3. Storing memory
In order for sound to be easily accessible, it must be stored in a
finite amount of space. Computer storage space is limited. If 64 bit
numbers were used (immense numerical detail) and sampled at a
rate of 100kHz, the digitally stored values would be huge. A quality
and accuracy balance must be struck to fit storage limitations.

Graph of digital sound wave using different number of

bit width

An example of what a 3-bit sound

file might look like (8 possible values).

An example of what a 6-bit sound file

might look like (64 possible values).

The changes in amplitude cannot be detected by the human ear

well if the bit width is too small. Using low bit widths cause the
timbre of the sound to be less defined. In the music industry The
ideal format is 44,100 Hz at 16 bits which is the standard audio CD
quality.
Real world example:
A 10-second signal sampled at 44.1 kHz with 16-bit (2-byte) precision
requires 882,000 bytes (= 10 sec x 44,100 samples/sec x 2 bytes/
sample), or about 861 Kbytes of storage)

Appendices
5.1
Sounds are waves which require a medium to propogate through.
These waves are transmitted through objects or passed through air.
For example, when you knock on wood a sound wave passes
through and disturbs the particles. It then emitted and the wave
disturbs air particles.
The ear drum simply detects the continuous change in pressure as a
result of the wave propogation and disturbance. The frequency of
the change in pressure reflects the pitch of the sound, and the
amplitude reflects the loudness (measured in decibels).
5.2 Microphone diaphragm

Microphones often use multiple rubber layers which deflect at

different frequencies. These layers enable microphones to capture a
wide range of frequencies.

5.3 Analogue to digital converter

a device that converts a continuous physical quantity (usually voltage) to a digital number
that represents the quantity's amplitude.

5.4 Nyquest Sampling Theorem

Humans can hear from 20Hz -20000Hz and often a sampling rate of
40000Hz is used to record music.
5.5 Binary
In mathematics and digital electronics, a binary number is a number
expressed in the binary numeral system, or base-2 numeral system. It is
expressed in 0s and 1s.
From left to right, each position represents a value for the power of 2.
Eg. 1 = 20
=1
101 = 20 +22
=4
10111 = 20 +22 +23 +24
= 29
5.6 Breadboard image of a simple Analogue to Digital
converter

The bargraph LED lights up according to the reading of voltage.

Starting from the bottom, if the reading was 8V the led would light
up showing:
OFF OFF OFF ON OFF OFF OFF OFF OFF OFF
Since it is in the 4th position 23 = 8V.

References
Microphones: how microphones work; types of microphones;
choosing a microphone.. 2015. Microphones: how microphones
work; types of microphones; choosing a microphone.. [ONLINE]
Available at:https://microphones.audiolinks.com/microphones.shtml.
[Accessed 16 May 2015].
Analog to Digital Conversion - learn.sparkfun.com. 2015. Analog to
Digital Conversion - learn.sparkfun.com. [ONLINE] Available
at:https://learn.sparkfun.com/tutorials/analog-to-digital-conversion.
[Accessed 01 June 2015].
Music and Computers. 2015. Music and Computers. [ONLINE]
Available
at:http://music.columbia.edu/cmc/musicandcomputers/chapter2/02_
01.php. [Accessed 02 June 2015].
Pelgrom, M.J.M.P, 2013. Analog-to-Digital Conversion. 2nd ed.
Netherlands: Springer Science+Business Media.
Pelgrom, M.J.M.P, 2013. Analog-to-Digital Conversion. 2nd ed.
Netherlands: Springer Science+Business Media.