EXP 10: SIGNAL PROCESSING APPLICATION

The given experiment is a Group Experiment performed by Umesh Gawale, Tejashree Gore, Pranav Ghaisas and Natasha Choudhary.

Our applications was based on Comb filters.
Patent Review :
Speech Signal Processor using Comb Filter

Application No.: 792,405
Patent No. : US4099030
Inventors : Yoshimutsu Hirata, No. 10-307, 842, Arai, Hino, Tokyo, Japan
Publication : 4^th July, 1978
Summary : Disruptive discontinuities resulting from sampling in a time compression and expansion FT system are minimized by use of comb filters.

https://drive.google.com/open?id=0ByBqyA3dBdX7T3ROQ3VXRy04Vnc

IEEE Paper Review:
Speech Enhancement Using Harmonic Emphasis and Adaptive Comb Filtering
Publisher : Wen Jin, Xin Liu, Michael S. Scordilis and Lu Han
Summary :
An enhancement method for single-channel speech degraded by additive noise is proposed. A spectral weighting function is derived by constrained optimization to suppress noise in the frequency domain. Two design parameters are included in the suppression gain. Further enhancement of the harmonics is achieved by adaptive comb filtering derived using the gain factor with a peak-picking algorithm. The performance of the enhancement method was evaluated by the modified bark spectral distance, composite objective measures and listening tests. Experimental results indicate that the proposed method outperforms spectral subtraction; a main signal subspace method applicable to both white and colored noise conditions and a perceptually based enhancement method with a constant noise-flooring parameter, particularly at lower signal-to-noise ratio conditions. Our listening test indicated that 16 listeners on average preferred the proposed approach over any of the other three approaches about 73% of the time.

https://drive.google.com/open?id=0ByBqyA3dBdX7RTVZcVJQMVlqN00

EXP 9: BASIC OPERATIONS USING DSP PROCESSOR

We studied how to program DSP Processors using C language and Assembly language. The kit used was TMS320F28375. The kit was demonstrated by a student from B.E ETRX. Instructions required for carrying out basic Arithmetic, Logical, Shift and Rotate Operations were demonstrated by observing values of registers before and after execution. The Software used was Code Composer Studio by Texas Instruments.

EXP 8: FIR FILTER DESIGN (FREQUENCY SAMPLING)

Design of Linear Phase FIR Filter using Frequency Sampling Method. The programming technique was same as that for Window Function Method. It was designed for a LPF and HPF instead of BPF.
The input specifications were:
(1) Pass band Attenuation (Ap) (2) Stop band Attenuation (As )
(3) Pass band Frequency (Fp) in Hz (4) Stop band Frequency (Fs) in Hz
(5) Sampling Frequency in Hz
The phase spectrum was linear. Hence it is a FIR Filter.

https://drive.google.com/open?id=0Bzfvoo_rjoa8X2JPTHpLdHNVa3M

EXP 7: DESIGN OF FIR FILTER (Window Function Method)

Design of Linear Phase FIR Filter using Window Function. It was designed for two cases: LPF and BPF using Scilab and C programming. The C program accepted input values and gave output as h(n). The scilab code accepted h(n) and gave magnitude and phase spectrum.
The input specifications were:
For LPF  filter Design :
(1) Pass band Attenuation (Ap) (2) Stop band Attenuation (As )
(3) Pass band Frequency (Fp) in Hz (4) Stop band Frequency (Fs) in Hz
(5) Sampling Frequency in Hz
 For BPF  filter Design :
(1) Pass band Attenuation (Ap) (2) Stop band Attenuation (As )
(3) Pass band Frequency (Fp1, Fp2) in Hz (4) Stop band Frequency (Fs) in Hz
(5) Sampling Frequency in Hz
The phase spectrum showed linear phase. Hence it is a FIR Filter.

https://drive.google.com/open?id=0Bzfvoo_rjoa8VHVVYVpLS3lpdjA

EXP 6: DESIGN OF CHEBYSHEV IIR FILTER

Design of Low Pass Chebyshev and High Pass Chebyshev Filter is performed using Scilab. The input specifications are specified by the user:
(1) Pass band Attenuation (Ap) (2) Stop band Attenuation (As )
(3) Pass band Frequency (Fp) in Hz (4) Stop band Frequency (Fs) in Hz
(5) Sampling Frequency in Hz
The poles for both the filters lie inside the unit circle, hence both are stable filters.
The order of the filters can always be increased for higher accuracy.
Ripples are observed in magnitude spectrum at Pass Band. Hence its Chebyshev-I filter.
The observed and calculated values of Ap and As  were found approximately same.

https://drive.google.com/open?id=0Bzfvoo_rjoa8aDVISHdZY2dvdlE

EXP 5: DESIGN OF BUTTERWORTH FILTER

Designing of  Low Pass Butterworth Filter and High Pass Butterworth Filter using Scilab is performed. Here the Input Specifications were specified by the user
(1) Pass band Attenuation (Ap< 3 dB) (2) Stop band Attenuation (As> 40 dB )
(3) Pass band Frequency (Fp) in Hz (4) Stop band Frequency (Fs) in Hz
(5) Sampling Frequency (F) in Hz
The poles for both the filters lie inside the unit circle, hence they are stable.
In the magnitude there are no ripples observed in either of the bands. A smooth magnitude curve determines a Butterworth Filter.
The order of the filter can always be increaesed for greater accuracy.
The observed and calculated values of Ap and As were found approximately same.

https://drive.google.com/open?id=0Bzfvoo_rjoa8Nm0yVGlJaWFFQkU

EXP 4: FILTERING OF LONG DATA SEQUENCE

We performed Filtering of Long Data Sequence using Overlap Add Method(OAM) and Overlap Save Method(OSM).. The algorithms decrease the delay in getting the output. In OAM the signal is decomposed into N sequences and zero padding is added. In OSM the decomposed signal of the first short length consists of zero, since the second signal consists the first signal as well, it saves the data. OAM and OSM are used in processing real time causal signals

https://drive.google.com/open?id=0Bzfvoo_rjoa8V25BdExPZmxzQzQ

https://drive.google.com/open?id=0Bzfvoo_rjoa8R2hNOEpNaHduUXc