You are on page 1of 17

Brief Review of Fourier Analysis

Elena Punskaya
www-sigproc.eng.cam.ac.uk/~op205
Some material adapted from courses by Prof. Simon Godsill, Dr. Arnaud Doucet, Dr. Malcolm Macleod and Prof. Peter Rayner

22

Time domain Example: speech recognition difficult to differentiate between different sounds in time domain
tiny segment

sound /i/
as in see

sound /a/
as in father
23

How do we hear?
Cochlea spiral of tissue with liquid and thousands of tiny hairs that gradually get smaller Each hair is connected to the nerve The longer hair resonate with lower frequencies, the shorter hair resonate with higher frequencies Thus the time-domain air pressure signal is transformed into frequency spectrum, which is then processed by the brain
24

Inner Ear

www.uptodate.com

Our ear is a Natural Fourier Transform Analyser!

Fouriers Discovery
Jean Baptiste Fourier showed that any signal could be made up by adding together a series of pure tones (sine wave) of appropriate amplitude and phase
(Recall from 1A Maths)

Fourier Series for periodic square wave


infinitely large number of sine waves is required
25

Fourier Transform

The Fourier transform is an equation to calculate the frequency, amplitude and phase of each sine wave needed to make up any given signal :

(recall from 1B Signal and Data Analysis)


26

Prism Analogy

Analogy:
a prism which splits white light into a spectrum of colors white light consists of all frequencies mixed together the prism breaks them apart so we can see the separate frequencies
27

White light

Spectrum of colours

Signal

Fourier Transform

Spectrum

Signal Spectrum

Every signal has a frequency spectrum. ! the signal defines the spectrum ! the spectrum defines the signal We can move back and forth between the time domain and the frequency domain without losing information
28

Time domain / Frequency domain


! Some signals are easier to visualise in the frequency domain ! Some signals are easier to visualise in the time domain ! Some signals are easier to define in the time domain (amount of information needed) ! Some signals are easier to define in the frequency domain (amount of information needed) Fourier Transform is most useful tool for DSP
29

Fourier Transforms Examples


signal
cosine
t

spectrum
!"

added higher frequency component

!"

Back to our sound recognition problem:


t

sound /a/
as in father

!"
in logarithmis units of dB

peaks correspond to the resonances of the vocal tract shape they can be used to differentiate between sounds
30

sound /i/
as in see

t
in logarithmis units of dB

!"

Discrete Time Fourier Transform (DTFT) What about sampled signal?


The DTFT is defined as the Fourier transform of the sampled signal. Define the sampled signal in the usual way:

Take Fourier transform directly

using the sifting property of the !-function to reach the last line
31

Discrete Time Fourier Transform Signal Samples


Note that this expression known as DTFT is a periodic function of the frequency usually written as

The signal sample values may be expressed in terms of DTFT by noting that the equation above has the form of Fourier series (as a function of ") and hence the sampled signal can be obtained directly as

[You can show this for yourself by first noting that (*) is a complex Fourier series with coefficients however it is also covered in one of Part IB Examples Papers]
32

Computing DTFT on Digital Computer

The DTFT

expresses the spectrum of a sampled signal in terms of the signal samples but is not computable on a digital computer for two reasons: 1.! The frequency variable " is continuous. 2.! The summation involves an infinite number of samples.
33

Overcoming problems with computing DTFT


The problems with computing DTFT on a digital computer can be overcome by: Step 1. Evaluating the DTFT at a finite collection of discrete frequencies. no undesirable consequences, any frequency of interest can always be included in the collection Step 2. Performing the summation over a finite number of data points does have consequences since signals are generally not of finite duration
34

The Discrete Fourier Transform (DFT)


The discrete set of frequencies chosen is arbitrary. However, since the DTFT is periodic we generally choose a uniformly spaced grid of N frequencies covering the range "T from 0 to 2#. If the summation is then truncated to just N data points we get the DFT

The inverse DFT can be used to obtain the sampled signal values from the DFT: multiply each side by and sum over p=0 to N-1

Orthogonality property of complex exponentials is N if n=q and 0 otherwise


35

The Discrete Fourier Transform Pair

36

Properties of the Discrete Fourier Transform (DFT)

! !

is periodic, for each p is periodic, for each n

! for real data

[You should check that you can show these results from first principles]

37

DTFT Normalised Frequency


Please also note the DTFT and IDTFT pair is often written as:

The assumption here is that ! is a normalized frequency

!=2#f$ = 2#(f/fs) -

normalized frequency (rad/sample)$

f - cycles per second fs - samples per second f/fs - cycles per sample

We will adopt this notation for majority of the slides.

38

You might also like