Equalizer is basically audio effect device to emphasize or deemphasize certain specific frequency components of a signal. The main usage of equalizer is to flatten the system spectral response [1]. In order to have our desired response of system for given input audio signal we used filtering device. Using filtering devices different methods had been developed. Graphical Equalizer is one of method used which consist of audio equipment for flattening the system spectral response in the audio signal band or produce other desirable effects. The report demonstrates the implementation of graphic equalizer by using DSK(DSP Starter Kit) C6713 kit interfaced with lab view software tool to analyze the response of sound and frequency in time domain as well as in frequency domain.
Key Words - Equalizers, Band pass filters, low pass filters, high pass filters, band stop filters, and Lab view.
INTRODUCTION
Equalizers are basically electrical filter networks which generate their output behavior change with frequencies. The most well known equalizers which are available commonly are amplitude equalizers, their basic attribute is that output signal changes with respect to amplitude [2].Equalizers are of many types for instance parametric equalizer, shelving equalizer, active etc, but here we used graphic equalizer. Equalizers are used in many places: in broadcast transmitters and receivers, sound systems, tape recording and reproduction etc. Equalizers have lot of applications, they are used to overcome loss in storage systems and many more applications [2]. The specialty of graphic equalizer is to eliminate the noise with great output of audio sound quality. In this project we focus on graphic equalizer. The main purpose of the project is to deploy digital signal processor kit C613 for processing and analyzing different frequency signals in the sound, and then finally DSP kit is interfaced with lab view to check the characteristic of signals in both the scenarios in time domain and frequency domain.
The tool which was used to design graphic equalizer is code composer studio. This tool is used for debug and develops embedded applications.
Lab view is basically designed to simulate different communication systems, here in this project the motive was to interlink code composer studio with lab view by using C code programming to produce the output frequency response in real time data exchange environment.lab view is the environment to create suitable scenario for programming languages by using different instruments. One of the main advantages of lab view over other scenarios is to help the users to access hardware easily. Lab view has the compiler that generates code for CPU. Lab view has lot of applications, for instance data acquiring, signal generating, mathematical operations, statistical operations and signal analysis etc[3].
2. PRE STUDY
Before starting the implementation of graphic equalizer by using different techniques, it would be necessary to have better understanding of digital signal processing techniques. The most common techniques used are: FIR (Finite Impulse Response), IIR (Infinite Impulse Response), FFT (Fast Fourier Transform) and IFFT (Inverse Fast Fourier Transform).Different examples illustrate the implementation of these digital signal processing techniques. Every technique has its own attributes and consequences. For instance IIR is more feasible in terms of algorithms, requires less time to be executed in terms of cycles but needs complex calculation for filter coefficients. On the other hand the demand of FFT is high with respect to sample points and requires complex execution to eliminate distortion from input signal. The basic idea which works behind FFT is the algorithm that is responsible for converting time domain signal into frequency domain signal that depends upon discrete Fourier transform. From keen study and the implementation of these digital signal processing techniques we came to know that FIR technique is the most suitable and reliable technique for the implementation of equalizer. Because FIR filters are linear phase and they are more compatible with digital signal processors. This technique is the simplest and feasible with hardware as compared to IIR and FFT [4].
Moreover there should be sufficient knowledge about matlab FDA tool implementation on proposed filters like high pass, low pass, band pass and band stop. Additionally the final phase of the project demands precise study of interfacing DSP kit with lab view. It is necessary to know about the operation and functionality of LAB View tool[3].
3. Experimental setup
The architecture TMS320C6713 (C6713) consist of the C6713 floating-point digital signal processor and a 32-bit stereo codec TLV320AIC23 (AIC23) for input and output. The onboard codec AIC23 [1] uses a sigma-delta technology that provides ADC and DAC. It connects to a 12-MHz system clock. Variable sampling rates from 8 to 96 kHz can be set readily.
The DSK board includes 16MB (megabytes) of synchronous dynamic random access memory(SDRAM) and 256kB (kilobytes) of flash memory.
4. Project Description
A. Filter Design
The FIR filtering method is used for implementation of various of DSP Design circuits as being the most efficient. For our course project we use 4 separate FIR filters with different frequency specification are designed. The detail specification can be found in C code of Equalizer. We can adjust the gain of the output of every filter such that it provides amplification for the specific frequency component of the audio signal. This way the signal can be compensated which is distorted by some noise hence making it sound better. The equalizer designed here covers the frequency range from 0 to 4.4 KHz. Most of physical resonances of aduio instruments usually fall between, say 1 or 2 KHz these frequencies are likely starting points. Thus different frequency components can be altered effectively. We used four different types of filter which are Low pass, High Pass, Band Pass and Band Stop.
Table
Equalizer filter bands
Filter Name
Filter
Band Frequency (KHz)
Low Pass filter
hlp
0 to1
High Pass filter
hhp
1 to 3.3
Band Pass Filter
hbsp
3.3 to 4
Band Stop Filter
hbss
4 to 4.4
The input voice is split into 4 different frequency components by passing it through 4 FIR filters with different frequency characteristics. Such that we have had flat output. The components are multiplied with their respective user configurable gain to increase or decrease their magnitude. They are added all together and we had modified output signal. We connect all the FIR filters of equal length in parallel, multiply them with their respective gain and sum them into a single filter. In this way we increasing the efficiency and the workload of the equalizer is reduced workload to a greater extent.
MATLAB FDATool was used to design the FIR filters and we can also use SPTool. It can be accessed by typing the 'fdatool' command in the MATLAB command window. The coefficients obtained with matlab are in floating point format so we changed it to the integer format by round command of Matlab. The coefficients obtained are in floating point format. We change it to fixed-point number representation improved the performance and accuracy of the Microprocessor. The Hamming Window method is used to derive the filter coefficients. Such that we used filter order of 40 The filter size is kept at 40 which was the minimum size at which output was of good quality. When I used higher order filters the output was distort. Due to reason that higher order filters utilize great resources of DSP Processor and computation take long time. So we had distorted output.
We had showed in Figure 1 how we can used FDATool of Matlab. We also then write C code for FIR filters such that we connect it properly.
Figure.1
In order to change Gain value with slider function in C code Composer multiplied with their respective filter coefficients and added together to make a filter to reduce computational power. But in our project we used Lab View for implementing slider function. Such that we place sliders on front panel of Lab View and name it with volume 1, volume 2, volume 3, and volume 4.
We know that two section of our project assembly code running in DSP starter Kit or DSK and Lab View code, providing user interface must communicate each other.
So both can communicate onboard Joint Test Action Group (JTAG) interface through Real-time data exchange (RTDX) channels.
We know that lab view is data acquisition, control systems, frequency responses and mathematical graphs. So we place Spectral measurement Express VI (virtual instrument ) for measurement of FFT RMS magnitude Figure 2.we had output in time domain.
Figure .2
From picture it appear that we have no output in waveform graph instrument but as processing speed of DSP is very fast so we have no graphical display. But when we used Probe option on wire connecting Spectral measurement with waveform graph in Block Diagram of Project we can have FFT RMS value shown in Figure 3.
Figure 3
Conclusion
Though output quality is good from our equalizer we designed and implemented on DSP Kit. But we can improved our output in different ways. If use filter of less order the computational efficiency of DSP starter Kit can be increased. We can improve quality of filtering of sound and have pleasant by using Lab View Embedded Edition. It is a special edition of Lab VIEW that installs in a separate directory and does not interfere with Lab VIEW Base, Full, or Professional development systems. The DSP Module is example of Labview Embedded edition which when installed. We can use graphical programming methods to learn DSP fundamentals and to develop applications for DSP hardware without having to write any C, assembly, or script source. So we will had to not write the RTDX code for communicating with DSP starter Kit of C6713.Thus Lab View will communicate directly and we will have optimized output.
acknowledgment
We appreciate the guidance of Mr. Kerstin Nilsson of School Engineering Karlskorna in carrying out this project. Indeed it was priceless opportunity to work in Digital Signals Processor field.
References
Digital Signal Processing and Applications with the C6713 and C6416 DSK By Rulph Chassaing ISBN 0-471-69007-4 Copyright © 2005 by John Wiley & Sons, Inc
TMS320C6000 CPU and Instruction Set Reference Guide, SPRU189F, Texas Instruments,Dallas, TX, 2000.
[3] www.Labivew.com
