You are on page 1of 19

DIGITAL SIGNAL

PROCESSING
- GARUDA SUMA PRANAVI
-SRI TEJA GADAMSETTI
- ADITYA KP

SIGNAL SAMPLING AND


RECONSTRUCTION
Insignal
processing,samplingis the
reduction of a
continuoussignalto a discrete
signal. A common example is
the conversion of a sound wave
(a continuoussignal) to a
sequence ofsamples(a
discrete-timesignal).

Suppose the highest frequency


component, for a given analog
signal isfmax. According to the
Nyquist Theorem, the sampling rate
must be at least 2fmax or twice the
highest analog frequency
component. If the sampling rate is
less than 2fmax, some of the highest
frequency components in the
analog input signal will not be
correctly represented in the
digitized output. When such a
digital signal is converted back to
analog form by a digital-to-analog
converter, false frequency
components appear that were not in
the original analog signal. This

SAMPLING ACCORDING TO NYQUIST


CRITERIA
f=1000;
fs=8000;
t=0:1/fs:1;
x=sin(2*pi*f*t);
plot(t,x);
grid on;
axis([0 6e-03 -1.2 1.2]);

ALIASING (1000HZ SIGNAL APPEARS TO BE A 500HZ SIGNAL)


f=1000;
fs=1500;
t=0:1/fs:1;
x=sin(2*pi*f*t);
plot(t,x);
grid on;
axis([0 6e-03 -1.2 1.2]);
(also has a phase shift of pi
radians)

ALIASING

(2000HZ SIGNAL LOOKS LIKE A 500HZ SIGNAL)

f=2000;
fs=1500;
t=0:1/fs:1;
x=sin(2*pi*f*t);
plot(t,x);
grid on;
axis([0 6e-03 -1.2 1.2]);
(no phase shift)

SIGNAL SAMPLING AND


RECONSTRUCTION
Downsampling (decimation) is
the process of subsampling a
discrete signal. If D is the
downsampling factor then one
out of D samples are chosen
from the original signal to
construct the sampled signal.
Upsampling is the process of
introducing zeros between
samples to create a longer
signal.

ORIGINAL SIGNAL
The original signal
speech_dft in mp3
format has been
sampled at 22050hz
(therefore signals in
the -11 to 11khz range
are depicted), is 16 bit
and is in mono formatto reduce processing
on the output side.
(taken from multimedia
file- Simulink library).
However other audio
files or real time
speech signals can also

The input to the multiport


switch should be signals
sampled at the same
frequency for it to function
properly, hence we must add
redundancies to obtain the
desired sampling rate of
22050Hz which is done
through an upsampler block.
The consequent components
are just for amplification and
smooth passing of the signal
(mitigation for aliasing) as
there is loss of signal energy
while propagating through the

SIGNAL DOWNSAMPLED BY A FACTOR 4


The original signal has been
sampled at a rate 22050/4=
sampled at 5.5khz. Hence,
according to nyquist criteria
the max. frequency component
should be 2500hz in the speech
signal, which is not the case.
This results to aliasing
(distortion of signal- string like
sounds) where the resulting
signal seems to be of a
different frequency than the
original signal when
reconstructed.

SIGNAL THROUGH A LOWPASS FILTER


We know that most of the
information in any signal is
contained in the low frequency
components. Hence, we pass
the signal through a lowpass
filter of 2.4khz cutoff
frequency. Hence, there is no
noise or distortion but a small
amount of information may be
lost or attenuated.

SIGNAL PASSED THROUGH A LOWPASS


FILTER AND THEN DOWNSAMPLED
The signal passed through a
lowpass filter and then
downsampled by a factor of 4
sounds very similar to the signal
passed only through a lowpass
filter. The lowpass filter just
mitigates the distortion caused
by aliasing. By downsampling the
signal we have increased the
efficiency by avoiding further
unnecessary computations for
other complex applications and
also reduced the amount of
space required to store these

SAMPLING AN ANALOG SIGNAL


In practice, the sampling is accomplished by means of high speed
switching transistor circuits. Accordingly, we find that the resulting
sampled waveform deviates from the ideal form because a physical
switching circuit however fast still requires a non zero interval of time.
The sampling takes the form of natural sampling and flat top sampling. In
both cases, the spectrum of the sampled signal is scaled by the ratio of
T/Ts, where T is the sampling pulse duration and Ts is the sampling period.
Typically, this ratio is quite small, with the result that the signal power at
the output of the low pass filter in the receiver is correspondingly small.
However, we can remedy this situation by the use of amplification

SAMPLE AND HOLD AMPLIFIER


In a sample and hold amplifier, the
switch is timed to close only for a
small duration T of each sampling
pulse, during which time the
capacitor rapidly charges up to a
voltage level to that of the input
sample. When the switch is open,
the capacitor retains its voltage
level until the next closure of the
switch. Thus, the sample and hold
circuit represents a staircase
interpretation of the original signal.

The waveform to the left


represents time scope i.e.
the random signal applied
to the circuit. The
waveform in the middle
represents the clock and
the waveform at the left
shows the output of the
sample and hold circuit

The signal is reconstructed more faithfully in the case of


sample hold amplifier rather than the case of sampling
amplifier, where the reconstructed signal suffers from an
amplitude distortion.

You might also like