Lecture Notes

09PC602
DIGITAL SIGNAL PROCESSING

VI Sem B.E (I.T)

Dr. J. SASIKALA
Associate Professor
Department of Information Technology
 Dr. J. Sasikala
Associate Professor

IIR Filters Unit-3

 
 
 

1 
 
  

2 
 
  

3 
 
  

4 
 
  

5 
 
  

6 
 
  

7 
 
  

8 
 
  

9 
 
  

10 
 
  

11 
 
  

12 
 
  

13 
 
  
14 
 
  

15 
 
  

16 
 
  

17 
 
  

18 
 
  

19 
 
  

20 
 
  

21 
 
  

22 
 
  

23 
 
  

24 
 
  

25 
 
  

26 
 
  

27 
 
  

28 
 
  

29 
 
  

30 
 
  

31 
 
  

32 
 
  

33 
 
  

34 
 
  

35 
 
  

36 
 
  

37 
 
  

38 
 
  

39 
 
  

40 
 
  

41 
 
  

42 
 
  

43 
 
  

44 
 
  

45 
 
  

46 
 
  

47 
 
  

48 
 
  

49 
 
  

50 
 
  

51 
 
  

52 
 
  

53 
 
  

54 
 
  

55 
 
  

56 
 
  

57 
 
  

58 
 
  

59 
 
  
60 
 
  

61 
 
  

62 
 
  

63 
 
  

64 
 
  

65 
 
  

66 
 
  

67 
 
  

68 
 
  

69 
 
  

70 
 
  

71 
 
  

72 
 
  

73 
 
  

74 
 
  

75 
 
  

76 
 
  

77 
 
  

78 
 
  

79 
 
  

80 
 
  

81 
 
  

82 
 
  

83 
 
  

84 
 
  

85 
 
 

 
 

 
 
Unit-4 - Finite Word Length Registers
UNIT‐V  Multirate Digital Signal Processing 

 
 

 
  

Fig. Spectra of the inte
ermediate aand decimaated sequen
nce 

 
  

 
  
Fig. Spectra of the Original, intermediate and other sequences 

 
 
  

 
  
Fig. An illustration of interpolation by a rational factor (L=3, M=2) 

 
  

 
 
  

        
position of (a) a decim
Fig.  Polyphaase decomp mation filter   (b) an in
nterpolation
n 
fiilter 

      
  

 
 

 
  

 
 
  
 

 
Fig. Digital to Analogue conversion for a CD player using x8 oversampling 

 
 

   
  

 
Fig. Simplified block diagram of single‐bit ADC scheme 

 
Speech Compression 

 
  

 
  
 

 
  

 
  

 
   
  

Filter into two bands: 
50 Hz‐3.5 KHz: sample at 8 kHz, encode at 48 KB/s 
3.5 KHz‐ 7 KHz: sample at 16 kHz, encode at 16 KB/s 
 
4 KHz‐7 KHz 

 
 

 
  

 
 

 
  
                                Vocal
tract excitation parameters (LPC
Coefficients)
Adaptive Filters 
Adaptive Filters ‐ Applications 
New Trends in Adaptive Filtering
 Musical Sound Processing 
The audio effects are artificially generated using various signal
processing circuits and devices, and increasingly by digital signal
processing techniques, often referred as Musical Sound processing.
All musical programs are produced in basically two stages:

The signals from each track are

Sound from each individual manipulated by the sound engineer
instrument is recorded in an to add special audio effects and are
acoustically inert studio on a combined in a mix-down system to
single track of a multi-track finally generate the stereo
tape recorder recording on a two-track tape

Time Domain Operation

Commonly used time-domain operations are:

   
 Singlle Echo
o Filter

Echo arre simply generated

d by delayy units. Be ecause off the com like shape of the
magnitu
ude respo onse, such a comb filter.
h a filter iss known as

For exaample, the

e direct so
ound and a single echo
e appe
earing R sampling
s p
periods
later can be simp
ply generaated by the FIR filte
er shown in
i Fig., whhich is
characterized by the differrence equ
uation:

Multip
ple Echo
o Filter

To geneerate a fix
xed numbeer of multtiple echoe
es spaced
d R samp
pling perio
ods with
the exponentiallyy decaying
g amplituddes.

One can use an FIR filter with

w a tran
nsfer funcction of the form:

H(z) =

   
 Rev
verbera
ation

 The reverbera
ation is co
onsidered o denselyy packed echoes.

 The IIR comb b filter by itself doe

es not pro
ovide natu
ural-sound
ding reverberation
for tw
wo reason
ns, which are:

 Its magnnitude ressponse is not consttant for all frequenccies, resu

ulting in a
"coloratiion" of many
m muusical souund that are ofte en unplea asant for
listening
g.

 The outpput echo density

d given by nu
umber of echoes
e pe
er second
d
generate
ed by a unnit impulsse at the in
nput is muuch lower than thatt
observe
ed in a rea
al room thus causin ng "flutterin e composite
ng" of the
sound.
F
Flangin g

 A nu
umber of special so ound effe
ects are often
o used
d in the mix-down
m process.
One such effe
ect is flang
ging.

 It wa
as createdd by feeding the sa ecorders and then
ame musiccal piece to tape re
comb bining the
eir delaye
ed outputss while varying
v the differen
nce betweeen their
delayy.

 One way of vaarying tim ow down one of the

me is to slo e tape reccorders byy placing
the operators
o thumb on n the flan
nge of the el, which is led to the name
e feed ree
flang
ging.

 Flang
ging Effec
ct:

Chorrus Gen
nerator and Phasing

 The chorus efffect is acchieved when

w seveeral musiccians are playing the same
musiical piecee at the saame time but with small chaanges in the
t amplittude and
smalll timing differencess between
n their sou
unds.

 The phasing effect is producced by processin

ng the signal
s through a
 narro
owband notch
n filte
er with variable
v b
botch chaaracteristiccs and adding
a a
scale
ed portion
n of the no otch filter output
o the
e original signal.

 Chorrus Effect:

 Phassing Effec
ct:

Freq
quency--Domain
n Opera
ations

 Thesse effects s are acchieved byb passin ng the orriginal siggnals throough an
equaalizer, the
e purpose of equallizer is to provide "presence e" by pea
aking the
mid-ffrequencyy compon he range of 1.5 GH
nents in th Hz to 3 GHz
G and to
t modify
the bass-trebl
b nships by providing boost or cut to co
e relation omponnetss outside
this range.
r

 Adva
antages:
 an inform
mation can be conveyed, displa ayed or ma anipulatedd
 perfect re
eproducibiility-identiccal perform
mance fromm unit to un
nit.
 Guarante eed accura
acy is onlyy determine ed by the number
n of bits used..
 Stored almost inde
efinitely witthout loss of information

 Disadvantages:
 Speed and cost- bee expensivve with larg
ge bandwiidth signalss
 DSP des signs can be time consuming g plus neeed the ne ecessary resources
r
(software
e etc.)
 Finite wo
ord-length problems - if only a limited number
n of bits is use
ed due to
economic conside erations, serious degradatio on may result in n system
performaance.

   
 Ou
utput fro
om the Algorithm
Singlle Echo
o Filter

Filterr Structure

Impulse Respoonse with

A==0.8, N=6 & R=4
Multip
ple Echo
o Filter

Impulse Respo
onse with A=0.8 for R=4
R Magniitude Resp
ponse for R
R=7
Rev
verbera
ation

Block Diagrram
Imp
pulse Resp
ponse with
h A=0.8 forr R=4

Im
mage EEnhan
nceme
ent
Originall Image

Compre
essed Image
   
 Imagge Enhanccement

E
Example
es

   
Spatial Domain Enhancement

Enhancement Methods
Nois
se Redu
uction
Intensity Adjustment
 Histogram Equ
ualization
n

   
 Image Thresholding

   
 Gray Level Slicing

Image Rotation

   
 Conversion M
Methods

Examp
ples:
 Resources Required

   
 Examples
Original Image Image with Salt & Pepper Noise

Filtered Image Histogram Equalization

Contrast Stretched Image Thresholded Image

   
  

Grayscale slicing background Grayscale slicing without

background

180 degree rotation

   
 APPLICATIONS
Biology Astronomy

Medicines Security, Biometrics

Satellite Imaging Personal Imagery

 APPLICATIONS

Example : Audio Sample Rate Conversion

Example : Oversampling ADC