Project report

15



Audio Processing and Noise Reduction With help of Digital filters



EE450 System analysis and simulation
By

Student ID# 18091 Patel Ankitkumar v.
Student ID#16221 shivangkumar naik
TABLE OF CONTENTS
ABSTRACT……………………………………………………………….3
INTRODUCTION………………………………………………………...3
BACKGROUND……………………………………………………….....4
PROCEDURE……………………………………………………………..5
APPLICATION……………………………………………………………6
MERITS……………………………………………………………………7
DEMERITS………………………………………………………………..7
MATLAB………………………………………………………………….8
FILTER DESIGN………………………………………………………….8
CONCLUSION…………………………………………………………....14
REFERENCES…………………………………………………………….15







ABSTRACT
In its most general form, a digital filter is a system that will receive an input in the form of a discrete-time signal and produce an output again in the form of a discrete-time signal. There are a lot of types of digital filters available to done a process of removing unwanted sound from transmitted audio signals. This project report emphasize on filter process on audio file and obtain desired output by using digital filter(Butterworth Low-pass Filter) and also observe magnitude response and phase response. In this paper, filter have been designed by using MATLAB software.

Introduction
Audio noise
Audio noise reduction system is the system that is used to remove the noise from the audio signals. Audio noise reduction systems can be divided into two basic approaches. The first approach is the complementary type which involves compressing the audio signal in some well-defined manner before it is recorded. The second approach is the non-complementary type which use techniques to reduce the noise level already present in the source material.
Digital filter (Butterworth)
Filters are very important operation of most electronics circuits. At the time when people start to communicate through electronics media. For examples, radio, cell phone etc. they need to remove unwanted noises to hear clearly desire sound and that is how idea of filter came. Filter is a device or process that eliminate undesired signal from transmitted signal which means remove unwanted noise or reduce background noise. Filters are categorized as per it works in range of frequencies. Filter neither add new frequencies to the transmitted signal nor change the components of that signal, but it only change the relative amplitude of the frequency components and their phase relationships. Filter is being use in circuit to concentrate signal in particular frequency range and reject signals on outer side frequency ranges.

Background
Butterworth filters are the most commonly used digital filer in audio circuit. This filter designed to get a flat frequency response as possible in pass-band and zero roll of response in stop-band. Butterworth filters are frequency based, so the effect of filtering can be easily understood and predicted also, they are fast and easy to use. Nevertheless, biggest drawback of the Butterworth filter is that it obtains this pass-band flatness by taking away from large transition band as the filter change from pass-band to stop-band. In addition, it has poor phase characteristics as well.
Idle frequency response referred as 90o corner.


As per the graph, the higher the Butterworth filter order, the higher the number of cascaded stages there are within the filter design, and the closer the filter becomes to the ideal response. However, in practice this "ideal" frequency response is impossible because of it produces excessive pass-band ripple.

Procedure
A basic digital filter system is composed of an ADC providing digital samples of an analog input then digital Processing system (μP/ASIC/FPGA)followed by DAC converting processed samples to analog output. Real-time signal processing: All processing operation must be complete between two consecutive samples.

When digital filter used for noise reduction, first it needed to input analog signals are must be converted into digital form by an analog to digital converter (ADC). This process is necessary because the digital filter cannot directly process an analog signal, so first the signal must be sampled at discrete points in time and then the samples must be digitized. The time interval between successive samples is referred to as the sampling period. As the sampling period less, more accurate output signal. Furthermore, filtering process done on discrete time signal then finally this filtered digital signals converted to analog signal to get desired output by using digital to analog converter (DAC).

APPLICATION
There are many applications of filters in the field of Telecommunication, Digital Signal Processing, Image Processing and Wireless communication etc.
They may also be used to "boost" and "cut" specific frequency ranges in an audio file to make it more pleasant to the ear.
Digital filters may be used to scrub audio data streams to increase the reliability of RADAR for defense systems.
It may also use to remove noise from digital pictures and television images.

MERITS
Digital storage is cheap.
Digital information can be encrypted for security.
Flexibility can be achieved with software implementations.
Accuracy can be controlled by choosing word length.
Physical size is small.
Reliability is high.

DEMERITS
Sampling causes loss of information.
Limited speed of processors.

MATLAB
MATLAB use to design and implement digital audio filter on sound files which we created. MATLAB is engineering mathematics program which focused on numerical computation. It is used to manipulating data and also offers a variety of functions for analyzing. To add, it is much easier to design and implement our digital filters than other languages such as C or JAVA.

FILTER DESIGN
Input audio signal

MATLAB software is used to design Butterworth Low-pass Filter to remove unwanted components from input signal and to produce desired output. Here is the MATLAB code to design Butterworth filter for audio processing and noise reduction.
MATLAB code:
clc;
clear all;
rp = input('Enter the passband ripple = ');
rs = input('Enter the stopband ripple = ');
wp = input('Enter the passband frequency = ');
ws = input('Enter the stopband frequency = ');
%fs = input('Enter the sampling frequency = ');
fs = 8000;
y = wavread('C:\Users\Ankit Patel\Desktop\train.wav'); %open the message
y=y(1:28000); % This simply trims the file to the desired length

w1 = 2*wp/fs;
w2 = 2*ws/fs;
[n,wn] = buttord(w1,w2,rp,rs);
[b,a] = butter(n,wn);
w = 0:0.01:pi;
[h,om] = freqz(b,a,w);
m = 20*log10(abs(h));
an = angle(h);
subplot(2,1,1);
plot(om/pi,m);
subplot(2,1,1);
plot(om/pi,m);
title('Magnitude Response');
ylabel('Gain in dB ---->');
xlabel('Normalised Frequency ---->');
grid on;
subplot(2,1,2);
plot(om/pi,an);
title('Phase Response');
xlabel('Normalised Frequency ---->');
ylabel('Phase in radians ---->');
grid on;

Ts=1/fs; %sample period
n=0:3.5*fs-1; %sample index (fs=samples per second), 3.5 seconds
tn=n*Ts;
filterLow_y=filtfilt(b,a,y);
freqz(b,a,128,fs);
figure('Color',[1 1 1]);
plot(tn,y);
axis('tight');
title('Audio file train.wav');
figure('Color',[1 1 1]);
plot(tn,filterLow_y);
axis('tight');
title('Audio file train.wav Filtered');
soundsc(filterLow_y);











INPUTS





MAGNITUDE RESPONSE AND PHASE RESPONSE GRAPH





OUTPUT


CONCLUSION
In this paper, we described that the low pass Butterworth filter are maximally flat as an amplitude response in the pass-band, and there is almost no ripple in pass-band. The transition time is controlled when you increase the filter order otherwise to obtain the accuracy of filter. In output audio more noise unwanted frequencies are eliminated from the input signal and produce output is more clearly to hear than input audio. The Butterworth response has better magnitude and phase response. Finally, the Butterworth filter is very simple and takes low order for accuracy. Output signals are frequency-based and the effect of filtering can be easily understood and predicted.

REFERENCES
"Butterworth Low-Pass Filter", Ece.uic.edu, 2016. [Online]. Available: http://www.ece.uic.edu/~jmorisak/blpf.html.
(1997) MATLAB The Language of Technical Computing. The Math Works Inc., Natick
D. Manolakis and V. Ingle, Applied digital signal processing. Cambridge: Cambridge University Press, 2011.
"Learn MATLAB", tutorial points, 2016. [Online]. Available: http://www.tutorialspoint.com/matlab/. [Accessed: 23- Apr- 2016].