This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
THE Graphic equalizer is an audio controller in the stereophonic system, which makes us to see graphically and manages different frequency bands individually. These equalizers have different sets of identical amplifiers for each channel in the audio system. They are two types of equalizer techniques graphic equalization and parametric equalization. In our project we use graphic equalizer called constant Q graphic equalizer. The center frequencies are corresponding to the ISO (international standard organization) standard for graphic equalizer center frequencies. We analyze the project with digital signal processor starter kit (DSK) using TMS320C6713 floating point processor on board with the support for input and output by software using the code composer studio. In this project we spotlight on the design and implementation of the audio equalizer with the digital signal processor. The next part of this paper explains the design techniques designed for digital system execution on the processor. Finally we design the model of the equalizer in the code composer studio and executed with the DSK6713 processor by taking real audio signal as input and observing the output by varying the characteristics of the equalizer. In order to use the DSP kit we make use of the DSK support tools. In the DSP starter kit we have three components such as Code
Composer Studio (CCS), an oscilloscope along with signal generator. CCS gives software supporting tools and also
Present's common platform for c compiler, assembler, math code, linker and debugging etc. It also contain the 32-bit codec for input, output communication, universal synchronous bus cable, and a power supply. In order to work with DSP we have to first execute some operations with the DSP. That is Quick test. The Quick test is for testing the DSP board for proper functioning. In Quick test select the GEL in the MENU and select the CHECK DSK. If any GEL files are required, it can be added by GOTO FILE load Gel. In order to create the projects add the all supporting files to the project, build the project with the help of BUILD option in PROJECT MENU option. It will produce the BINARY OUT file. The generated BINARY
OUT file can be used to load into the DSP processors.
Dsp Starter Kit:
TMS320C6713 DSP Starter Kit (DSK) gives users a opportune, low cost means of evaluating the features and architecture of theTMS320C6713 Digital Signal Processor from Texas Instruments. Various algorithms are well suited and based on VLIW architecture, on which the TMS320C6713(C6713) is also based on. The C6713 is capable of both ﬁxed and ﬂoating-point processing. In order to improve the various applications of signal processing, the stand alone C6713 DSK has introduced a wide range of on board peripherals and interfaces.
Fir Filter Fundamentals And Design
The instruction set and architecture of TMS320C6x makes it well appropriate for FIR filtering operations. Digital filter, such as an FIR filter, operates on discrete-time signals and can be implemented with a DSP such as the TMS320C6713. There are wide methods including a number of tools available to design and implement within a few minutes an FIR filter in real time using the TMS320C6x-based DSK. The filter design consists of the estimation of a transfer function with a resulting set of coefficients. Different techniques are available for the design of FIR filters.Fourier series is mostly utilized as computer aided design technique for the designment of FIR filter.The important charactersistic feature of FIR filter is that it can guarantee linear phase. The linear phase feature can be very helpful in applications such as speech analysis, where
phase distortion can be dangerous. Hence Linear phase filters are FIR filters. However, not all FIR filters have linear phase. Many applications in adaptive filtering and speech processing , such as in a linear predictive coding (LPC) application uses Lattice structure of FIR filters. Approximating the magnitude response of the transfer function to a desired magnitude response is done by making use of Fourier series design method of an FIR filter.
They are very excellent equalizers and are much suited to situations where the actual "modification" is required more. Achieving the equalization with four different filters which had fixed frequencies is called the audio graphic equalizer, which is very efficient. The variation of the magnitude response leads to high quality output sounds. From tone controls in terms of flexibility and control of Graphic equalizers, and the operation is still quite simple. A set of filters together is simply called as Graphic Equalizer, with a fixed center frequency for each filter that cannot be changed. The only control you have is the amount of boost in each frequency band. Sliders are those used for controlling the boost or cut. This interface is pretty perceptive because the positions of the sliders themselves resembles the frequency response of the equalizer. The sliders are a graphic representation of the frequency response, hence the name 'graphic' equalizer. The actual graphic equalizer implementation is different than the common tone controls. certain frequency bands are attenuated on stereo boost which acts as bass and treble tone knobs while letting everything thing else pass unaffected, so we can sequence them in series.
A graphic equalizer make use of a set of bandpass filters that are intended to completely isolate certain frequency bands. Equalization is the method of varying the frequency response characteristics of a signal. A music equalizer is used to amplify or attenuate a particular band of frequencies of a given signal in order to get better sound effects. The input signal is divided into different frequency bands by a series of bandpass filters (BPF) and then each band is attenuated or boosted by different frequency bands which inturn have independent gain control. All those sliders on the front of the equalizer are the gain controls in each band. The parallel arrangement of Gain factors (G) and all the bands are added finally to generate a composite signal.
Once the signal passes through the bandpass filters, you can now manipulate each of the filters, compared to the series connection with the tone controls, is used to diminish the more harmful effects of the filters. The magnitude frequency responses shown above do not tell us everything about the filter. The filter has a phase response as well. While phase distortion is desirable in some cases , in most sound reinforcement applications, we want to ensure that the sound isn't colored at all if possible. For each filter you add in series, its phase response is added to the phase response of the other filters. The phase response also reveals how the filter actually delays the signal. If you have two or three filters in your tone control, chaining them in series might be fine, but with a 15 or 31 band graphic equalizer, distortion begins to add up and make a difference.
Equalization Method Using Discrete Filters:
The audio graphic equalizer has evolved into a set of around thirty filters at fixed frequencies, covering the audio range. The operator has adjusted the level of each individually, either to correct a magnitude response variation, or to create one intentionally. This degree of control has remained popular despite the recent advances of technology.Digital Signal Processing (DSP) has made it practical to provide many times this resolution to the point where the term arbitrary magnitude response is considered applicable. As is well known, equalization has phase shift as a mathematical requirement. Phase shift has not always been considered as important as magnitude response because it was less audible. Still, studies have confirmed that it is audible in some situations.
Minimum phase is often chosen for its economy and because it is appropriate for correcting a system with minimum phase characteristics, which may be cancelled by minimum phase filtering without adding time delay . Minimum phase filters might also be appropriate for magnitude correction of system responses with unknown phase characteristics. Although discrete filters can be designed with narrow bandwidths, even approaching a filter shape with a flat top and steep cutoff is expensive, so in practice each filter has an effect over a wider range of frequencies than it is intended to affect. The filter response curves still have significant magnitude at neighboring filter band frequencies. Each band becomes independent, or very nearly so. Otherwise, to be effective, an operator must be very accustomed to the product's particular filter and combining behavior. Equalizer filters began as analog second-order filters, and have since been implemented as digital IIR filters. Equalization curves with complex shapes can also be accomplished with single large FIR filters, allowing multiple frequency band settings to be filter significantly affects no more than the first few neighboring frequency bands.
An FIR filter may be designed to approximate the impulse response of many IIR filters, and provides the ultimate in flexibility, where magnitude and phase may be adjusted more specifically and semi-independently. In order to support
low frequencies, many thousands of taps are required. In order to economically implement this, complicated methods like multirate processing or the use of FFT for fast circular convolution are used. IIR filters have some advantages over single FIR filters, namely simplicity and speed of design, and better efficiency for a limited number of bands that include low design frequencies. Analog filters are designed once, while a digital filter may need to be redesigned when any parameter changes. Small IIR filters can be adjusted (redesigned) quickly, compared to large FIR filters. Equalizers are usually adjusted manually, but in theory an automated analysis of the sound system may be
used.Care then must be taken with correcting nonminimum phase responses, where matching the phase requires added delay. Also, equalization of deep notches in the response should not be attempted because of potential adverse effects in other areas of a room, as well as the possibility of amplifier overload. These techniques are also beyond the scope of this paper.
Lab VIEW is a full-featured graphical programming language and development environment for embedded system design. LabVIEW provides a single graphical design tool for algorithm development, embedded system design, prototyping, and interfacing with real-world hardware. Additional modules have been designed to expand the core functionalities of LabVIEW to real-time operating systems, DSP, and FPGA programming making LabVIEW an ideal platform for signal processing algorithm design and implementation.This section briefly introduces the LabVIEW development environment from installation and basic programming to system development. The LabVIEW development system is the core software with which other modules, toolkits, and drivers can be installed to expand functionality and capability.
The LabVIEW development environment allows you to develop graphical programs and graphical user interfaces easily and effectively. A program in LabVIEW is called a Virtual Instrument (VI), and acts as an individual function similar to a function defined in the C programming language.
Main Description Of Our Work:
We were able to achieve the creation of the equalizer using a very nice concept. We did a serious pre-lab study for us to generate this concept. We stared first by clicking on the software (CCStudio V3.1). This now gave us the access to the 6713 chip. We get the chip connected and created a new project. The necessary files which make the minimum requirement to setup the program were added to the different folders such as the source and library folders etc. We use the scan all dependencies command, this search for all the header files in the system and them to the project the building of the program is done so that the .out file can be generated. If they happen to be an error during the building then the .out file will not be generated. The .out file will then be loaded into the DSP board chips in order to achieve its desired purpose. The concept we used was to create four filters which comprises of a low-pass, two band-pass and a high-pass filters. The low-pass filter has a frequency range between 100-900 hertz at an interval of 100 hertz. The first band-pass filter has the range from 900-3000 hertz with an interval of 300 hertz. The second band-pass filter has a range between 3000-4200 hertz at an interval of 200 hertz. The four filters have a range between 4200-6200 hertz at an interval of 400 hertz. A window size of 80 was used at a sampling frequency of 48 kHz. Since we have the order of 80, the filter coefficient will be 81. Too high order will lead to aliasing effect. The generation of the filter co efficiencies were done using FDA tool in the mat lab. FIR filter was considered with window method. This was used to generate all the filter co efficiency of the individual frequency of the filter. A C-header file was generated which was loaded into the software. The same procedure was used in generating the filter co efficiency from 100-6200 hertz.
The main aim of the graphic equalizer is to allow the user to view graphically, control the different frequency bands and reduce the distortion in order to provide all frequencies in the signal. So we start with the simple filters to make an appropriate equalizer. Usually the speech sampling rate is at 8 samples/per sec. and the sampling frequency of compact disc digital audio player nearly at 44K samples/sec. so we have concerned to this project the sampling frequency rate at 48 KHz so that no aliasing takes place. The aliasing mean disturbance due to the improper sampling. If the sampling frequency is less than twice of the input frequency than aliasing takes place. So we take 48 kHz as a sampling frequency, because our graphic equalizer output is summation of the all frequencies.
The idea of this project is very simple to get graphical equalization by taking four-band filters one is low pass filter and two band pass filters and one high pass filter. For our convenience the audible range between 20Hz - 20K Hz and most of the human voice frequency is in between 300Hz- 4K Hz.
So we have concentrated on this range of frequencies. For that we have taken four-band filters are low-pass filter from 100-900Hz with step size of 100Hz (Because audible ranges are in the low frequencies, so we have concentrated on the low frequency. For this we assume that low pass filter is healthier) and after that we introduced band pass filter. To increase the resolution of the band pass filter that can be alienated into two filters. So we have separated one band pass into two band pass filters. These are bandpass1, band pass filter2.the first one is band-pass filter1 from 900-3000Hz with step size of 300Hz, and second one is band-pass filter2 from 3200-4200Hz with the step size of 300Hz and after that other frequencies high-pass filter from 4200 to 6000Hz with step size of 400Hz. For each step we have to calculate the filter coefficients in order to get the desired transfer function. The equalizer coefficients file consists of total filter coefficients of the graphic equalizer. In the coefficient file each set contains 81(N) coefficient, these coefficients are designed by using MATLAB graphical user interface (GUI) filter designer SPTOOL or MATLAB FDATOOL. For our convenience we are considering the order of filter is 80 to get the desired filter. So we can get 81(N+1) coefficient.
This paper proposed a concept of a graphical equalizer
Based on c programming and with simple filters. The project of graphic equalizer is made easier to run-on the DSP because of its compatibility and also successful one. The output of the graphic equalizer program on DSP processors understands the
concept of the graphic equalizer, and in similar way we can also design different techniques in designing an equalizer for e.g. adaptive channel equalizer and etc.
The authors would like to thank Kristen Nilsson for his valuable guidance and discussions in the accomplishment of the project.
Embedded Signal Processing with the Micro
By Dr. Woon-Seng S. Gan, Dr. Sen M. Kuo
 © 2006 John Wiley and Sons, Inc.