Measured Period VOICE SIGNAL

Slides:



Advertisements
Similar presentations
Digital Signal Processing
Advertisements

Review. 2 TEST A 1.What is the single most important property of vibrating structure? 2. What happens when structure is vibrating in resonance? 3.What.
Image Enhancement in the Frequency Domain Part III
Removing the deflation from the recorded NIBP data Fig. 2: a) Filtered pulses (right scale) obtained by band pass ( Hz) filtering and b) Oscillometric.
14.1 ARITHMETIC MEAN Experimental Readings are scattered around amean value Fig Scatter of the readings around the mean value.
Freeman Algorithms and Applications in Computer Vision Lihi Zelnik-Manor Lecture 5: Pyramids.
Adaptive algorithm for determining limits of T wave in ECG waveform By Ashwin Panchbhai Sudheer Poojary.
Chapter 6 Random Error The Nature of Random Errors
Digital to Analogue Conversion Natural signals tend to be analogue Need to convert to digital.
The Wavelet Tutorial: Part3 The Discrete Wavelet Transform
The Normal Distribution Chapter 3 College of Education and Health Professions.
CECOM Bottom Line: THE WARFIGHTER File Name.PPT - Briefer’s Name - Briefed10/12/20151 Wei Su and John Kosinski U.S. ARMY CECOM RDEC Intelligence and Information.
1 Analysis and Synthesis of Pathological Vowels Prospectus Brian C. Gabelman 6/13/2003.
- 1 - YLD 10/2/99ESINSA Tools YLD 10/2/99ESINSA Filters Performances A filter should maintain the signal integrity. A signal does not exist alone.
Resonance Enhancement of the intensity of a particular frequency component(s) with respect to the intensity of the other components that occurs when its.
GG313 Lecture 24 11/17/05 Power Spectrum, Phase Spectrum, and Aliasing.
Topic: Pitch Extraction
FREQUENCY (HZ) LOG10 MAGNITUDE Figure 2.30a. PSD of original voice
Signal acquisition A/D conversion Sampling rate  Nyquist-Shannon sampling theorem: If bandlimited signal x(f) holds in [-B;B], then if f s = 1 / T.
146 Figure 5.1. Measured synthetic fundamental period combines portions from two separately synthesized pulses; time between minima is measured: the last.
Instrumental Analysis Elementary Statistics. I. Significant Figures The digits in a measured quantity that are known exactly plus one uncertain digit.
Multirate Signal Processing* Tutorial using MATLAB**
Part 5 - Chapter 17.
From: Variational Integrators for Structure-Preserving Filtering
Automatic Speech Processing Project
MECH 373 Instrumentation and Measurements
Volume 94, Issue 1, Pages (January 2008)
Simulation of surface EMG signals at different contraction intensities with the model used in Fig. 1. Simulation of surface EMG signals at different contraction.
Chapter 3 Sampling.
Skewness Skewness literaly means lack of symmetary. Thus it helps us to understand the shape of distributions.
Copyright © American Speech-Language-Hearing Association
Date of download: 12/27/2017 Copyright © ASME. All rights reserved.
Fundamentals Data.
Digital Control Systems Waseem Gulsher
Multimedia Fundamentals(continued)
Frequency Domain Analysis
Chapter 12 Curve Fitting : Fitting a Straight Line Gab-Byung Chae
Part 5 - Chapter 17.
Pyramid Coder with Nonlinear Prediction
Component: E Local quakes Broad frequency Two tremor bursts
MMSE Optimal Design: The Least Squares method
Unit 44 Vibrationdata Sine Filtering.
Volume 105, Issue 3, Pages (August 2013)
CHAPTER 5 Fundamentals of Statistics
Volume 58, Issue 3, Pages (May 2008)
FUNDAMENTAL PERIOD TRACK FEATURES
Chapter 3: Averages and Variation
Joseph M. Johnson, William J. Betz  Biophysical Journal 
Explaining Entropy responses after a noxious stimulus, with or without neuromuscular blocking agents, by means of the raw electroencephalographic and.
S2 Bilinear Couplings Study
Cortical Mechanisms of Smooth Eye Movements Revealed by Dynamic Covariations of Neural and Behavioral Responses  David Schoppik, Katherine I. Nagel, Stephen.
Tracking Phospholipid Populations in Polymorphism by Sideband Analyses of 31P Magic Angle Spinning NMR  Liam Moran, Nathan Janes  Biophysical Journal 
Indicator Variables Response: Highway MPG
Lab 2 Sampling and Quantization
Singular Behavior of Slow Dynamics of Single Excitable Cells
Hippocampal “Time Cells”: Time versus Path Integration
Marian Stewart Bartlett, Gwen C. Littlewort, Mark G. Frank, Kang Lee 
Xiao-Han Li, Elizabeth Rhoades  Biophysical Journal 
Analysis of Dynamic Brain Imaging Data
Strength and Orientation Tuning of the Thalamic Input to Simple Cells Revealed by Electrically Evoked Cortical Suppression  Sooyoung Chung, David Ferster 
Chapter 9 Advanced Topics in DSP
CHAPTER – 1.2 UNCERTAINTIES IN MEASUREMENTS.
Blockers of VacA Provide Insights into the Structure of the Pore
Relationship between approximate entropy and visual inspection of irregularity in the EEG signal, a comparison with spectral entropy  A. Anier, T. Lipping,
Lec.6:Discrete Fourier Transform and Signal Spectrum
Dayashankar Dubey (MTech)
Copyright © 2014 Elsevier Inc. All rights reserved.
Spatial Representation along the Proximodistal Axis of CA1
Reliability of Assessment of Protein Structure Prediction Methods
Yufang Wang, Ling Guo, Ido Golding, Edward C. Cox, N.P. Ong 
Presentation transcript:

Measured Period 8 6 4 2 VOICE SIGNAL -2 -4 -6 -8 0.018 0.02 0.022 VOICE SIGNAL -2 -4 -6 Measured Period -8 0.018 0.02 0.022 0.024 0.026 0.028 0.03 0.032 0.034 0.036 TIME (SEC) Figure 2.19. Identification of negative peaks for fundamental frequency tracking on a typical original voice signal. The tracking algorithm automatically selects the minima indicated by X and performs a parabolic interpolation to determine the time of minima to a precision of less than one sample. Circles indicate sample points.

275 270 FREQ (HZ) 265 260 255 0.2 0.4 0.6 0.8 1 TIME (SEC) Figure 2.20. A high resolution 1 sec. fundamental frequency track resulting from subsample interpolation. Note lack of fundamental frequency quantization (flat spots) that would occur without interpolation.

275 270 FREQ (HZ) 265 260 0.2 0.4 0.6 0.8 1 TIME (SEC) 4 2 FREQ (HZ) -2 -4 0.2 0.4 0.6 0.8 1 TIME (SEC) Figure 2.21. The fundamental frequency track of Fig 2.20 is low pass filtered (top part A curve) and high pass filtered (bottom part B curve) yielding the tremor and HFPV time series respectively. A cutoff frequency of 10 Hz is selected.

45 40 35 30 25 #OCCURRANCES 20 15 10 5 -1.5 -1 -0.5 0.5 1 1.5 DELT FREQ (%) TOTSKIP=0 FCUT=10 TOTMAN=0 Figure 2.22. Histogram of frequency deviations of the high pass filtered fundamental frequency time series of Fig 2.20. Successful fundamental frequency tracking yields a Gaussian form distribution.

ORIGINAL FUNDAMENTAL FREQ. 250 ORIGINAL FUNDAMENTAL FREQ. 240 RESAMPLED FUNDAMENTAL FREQ. 230 FUNDAMENTAL FREQUENCY (HZ) 220 210 200 190 0.2 0.4 0.6 0.8 1 TIME (SEC) Figure 2.23. Fundamental frequency time series of original voice and voice re-sampled to remove tremor (low frequency variations). Most of the significant fundamental frequency variation is removed

FREQ (Hz) TIME (SEC) DELTA FREQ PERCENT 0.2 0.4 0.6 0.8 1 266.2 266.4 0.2 0.4 0.6 0.8 1 266.2 266.4 266.6 FREQ (Hz) TIME (SEC) -0.1 -0.05 0.05 0.1 DELTA FREQ PERCENT Figure 2.24. Fundamental frequency time series of voice re-sampled to remove all fundamental period variations: both HFPV and tremor. In the upper plot the residual variation after removal is shown in Hz; it is less than 0.4 Hz. In the lower plot the residual is shown in percent.