MBE Vocoder
Page 0 of 34
Outline

Introduction to vocoders
 MBE vocoder
– MBE Parameters
– Parameter estimation
– Analysis and synthesis algorithm

AMBE
 IMBE
Page 1 of 34
Vocoders - analyzer
1.
2.
Speech analyzed first by segmenting speech using a
window (e.g. Hamming window)
Excitation and system parameters are calculated for
each segment
1.
2.
3.
Excitation parameters : voiced/unvoiced, pitch period
System parameters: spectral envelope / system impulse
response
Sending this parameters
Page 2 of 34
Vocoders - Synthesizer
Excitation Signal
White noise/ unvoiced
Pulse train/voiced
Page 3 of 34
System
parameters
Synthesized
voice
Vocoders

But usually vocoders have poor quality
– Fundamental limitation in speech models
– Inaccurate parameter estimation
– Incapability of pulse train/ white noise to produce all voice
• speech synthesized entirely with a periodic source exhibits a
“buzzy” quality, and speech synthesized entirely with a noise
source exhibits a “hoarse” quality

Potential solution to buzziness of vocoders is to use of mixed
excitation models
 In these vocoders periodic and noise like excitations are mixed
with a calculated ratio and this ration will be sent along the
parameters
Page 4 of 34
Multi Band Excitation Speech
Model

Due to stationary nature of a speech signal, a window w(n) is usually applied to signal
sw (n)  w(n) s(n)

The Fourier transform of a windowed segment sw ( )can be modeled as the product
of a spectral envelope H w ( ) and an excitation spectrum | Ew ( ) |
sˆw ( )  H w ( ) | Ew ( ) |

In most models H w ( ) is a smoothed version of the original speech spectrum sw ( )
Page 5 of 34
MBE model (Cont’d)

the spectral envelope must be represented accurately enough
to prevent degradations in the spectral envelope from
dominating.
– quality improvements achieved by the addition of a frequency
dependent voiced/unvoiced mixture function.


In previous simple models, the excitation spectrum is totally
specified by the fundamental frequency w0 and a
voiced/unvoiced decision for the entire spectrum.
In MBE model, the excitation spectrum is specified by the
fundamental frequency w0 and a frequency dependent
voiced/unvoiced mixture function.
Page 6 of 34
Multi Banding



In general, a continuously varying frequency dependent
voiced/unvoiced mixture function would require a large number of
parameters to represent it accurately. The addition of a large number of
parameters would severely decrease the utility of this model in such
applications as bit-rate reduction.
To further reduce the number of these binary parameters, the spectrum
is divided into multiple frequency bands and a binary voiced/unvoiced
parameter is allocated to each band.
MBE model differs from previous models in that the spectrum is
divided into a large number of frequency bands (typically 20 or more),
whereas previous models used three frequency bands at most .
Page 7 of 34
Multi Banding
V/UV
Original
information
spectrum
Noise
Spectral
spectrum
envelope
Excitation
spectrum
Periodic
spectrum
Synthetic
Page 8 of 34
spectrum
MBE Parameters

The parameters used in MBE model are:
1.
2.
3.
4.
Page 9 of 34
spectral envelope
the fundamental frequency
the V/UV information for each harmonic
and the phase of each harmonic declared
voiced. The phases of harmonics in frequency
bands declared unvoiced are not included
since they are not required by the synthesis
algorithm
Parameter Estimation



In many approaches (LPC based algorithms) the algorithms for
estimation of excitation parameters and estimation of spectral envelope
parameters operate independently.
These parameters are usually estimated based on heuristic criterion
without explicit consideration of how close the synthesized speech will
be to the original speech.
– This can result in a synthetic spectrum quite different from the
original spectrum.
In MBE the excitation and spectral envelope parameters are estimated
simultaneously so that the synthesized spectrum is closest in the least
squares sense to the spectrum of the original speech “analysis by
synthesis”
Page 10 of 34
Parameter Estimation (Cont’d)

the estimation process has been divided into two
major steps.
1. In the first step, the pitch period and spectral
envelope parameters are estimated to minimize the
error between the original spectrum and the synthetic
spectrum.
2. Then, the V/UV decisions are made based on the
closeness of fit between the original and the synthetic
spectrum at each harmonic of the estimated
fundamental.
Page 11 of 34
Parameter Estimation (cont’d)

The parameters estimated by
minimizing the following error
criterion:
–

Where
The error in an interval
1

2

  s
2
w
( )  sˆw ( )  d

sˆw ( )  H w ( ) | Ew ( ) |
m 
1
2
2
bm
 s
w
( )  Am EW ( )  d
am
bm
is minimized at:
Am 
 Sw( ) Ew( ) d
am
bm
 Ew( )
am
Page 12 of 34
2
d
Pitch Estimation and Spectral
Envelope

An efficient method for obtaining a good approximation
for the periodic transform P ( w ) in this interval is to
precompute samples of the Fourier transform of the
window w (n) and center it around the harmonic frequency
associated with this interval.
 For unvoiced frequency intervals, the envelope parameters
are estimated by substituting idealized white noise (unity
across the band) for |E (a)| in previous formulas which
reduces to averaging the original spectrum in each
frequency interval.
 For unvoiced regions, only the magnitude of A, is
estimated since the phase of A, is not required for speech
synthesis.
Page 13 of 34
More about pitch estimation

Experimentally, the error E tends to vary slowly
with the pitch period P
 the initial estimate is obtained by evaluating the
error for integer pitch periods
 Since integer multiples of the correct pitch period
have spectra with harmonics at the correct
frequencies, the error E will be comparable for the
correct pitch period and its integer multiples.
Page 14 of 34
More about pitch estimation
(Cont’d)
Error/Pitch
Speech
segment
Original
Original and
spectrum
Synthetic
P=42.48
Original and
Synthetic
Page 15 of 34
P=42
V/UV Decision


The voiced/unvoiced decision for each
harmonic is made by comparing the
normalized error over each harmonic
of the estimated fundamental to a
threshold
When the normalized error over mth
harmonic is below the threshold, this
frame will be marked as voiced else
unvoiced
Page 16 of 34
m
m 
1
2
bm
 Sw( )
am
2
d
Analysis Algorithm Flowchart
start
Window
Speech
segment
Compute error vs. pitch period
Autocorrelation approach
Page 17 of 34
Select initial pitch period
(Dynamic programming
Pitch tracker)
Refine initial pitch period
(frequency domain approach)
Make V/UV decision for each
Frequency band
Select V/UV spectral
Envelope parameters
For each freq. band
Stop
Speech Synthesis



The voiced signal can be synthesized as the sum of
sinusoidal oscillators with frequencies at the harmonics of
the fundamental and amplitudes set by the spectral
envelope parameters (The time domain method).
The unvoiced signal can be synthesized as the sum of
bandpass filtered white noise
The frequency domain method was selected for
synthesizing the unvoiced portion of the synthetic speech.
Page 18 of 34
Synthesis algorithm block diagram
V/UV
Decision
Voiced envelope
Separate
samples
Envelope Voiced/Unvoiced
samples Envelope samples Unvoiced envelope
samples
Unvoiced envelope
samples
White noise
sequence
Page 19 of 34
Voiced envelope
samples
Unvoiced envelope
Bank of
Harmonic
oscillators
Voiced
speech
samples
Linear
interpolation
STFT
Replace
envelope
Weighted Unvoiced
Overlap-add speech
MBE Synthesis algorithm



First, the spectral envelope samples are separated into voiced or
unvoiced spectral envelope samples depending on whether they are in
frequency bands declared voiced or unvoiced
Voiced envelope samples include both magnitude and phase, whereas
unvoiced envelope samples include only the magnitude.
Voiced speech is synthesized from the voiced envelope samples by
summing the outputs of a band of sinusoidal oscillators running at the
harmonics of the fundamental frequency
sˆv (t )   Am (t ) cos( m (t ))
m
Page 20 of 34
MBE Synthesis algorithm (Voiced)

The phase function  m is determined by an initial
phase 0 and a frequency track  m (t ) as follows:
t
 m (t )    m ( )d  0
0

The frequency track  m (t ) is linearly interpolated
between the mth harmonic of the current frame and
that of the next frame by:
 m (t )  m 0 (0)
Page 21 of 34
(S  t )
t
 m 0 ( S )   m
S
S
MBE Synthesis algorithm
(Unvoiced)




Unvoiced speech is synthesized from the unvoiced envelope samples
by first synthesizing a white noise sequence.
For each frame, the white noise sequence is windowed and an FFT is
applied to produce samples of the Fourier transform
In each unvoiced frequency band, the noise transform samples are
normalized to have unity magnitude. The unvoiced spectral envelope is
constructed by linearly interpolating between the envelope samples
|Am(t)|.
The normalized noise transform is multiplied by the spectral envelope
to produce the synthetic transform. The synthetic transforms are then
used to synthesize unvoiced speech using the weighted overlap-add
method.
Page 22 of 34
MBE Synthesis (Cont’d)

The final synthesized speech is generated by summing the
voiced and unvoiced synthesized speech signals
Voiced
speech
+
Unvoiced
speech
Page 23 of 34
Synthesized
speech
Bit Allocation
Page 24 of 34
Parameter
Bits
Fundamental
Frequency
9
Harmonic
Magnitude
139-94
Harmonic
Phase
0-45
Voiced/Unvoiced
Bits
12
Total
160
Advanced MBE (AMBE)



MBE coding rate at 2400 bps
AMBE coding rate at 1200/2400 bps
Four new features
1.
2.
3.
4.
Enhanced V/UV decision
Initial pitch detection
Refined pitch determination
Dual rate coding
Page 25 of 34
Enhanced V/UV decision

divide the whole speech frequency band
into 4 subbands and 2 subbands for 2.4 kbps
and 1.2 kbps respectively.
 That is to say only 4 bits and 2 bits are used
to encode U/V decisions for 2.4 kb/s and
1.2 kb/s vocoder respectively.
Page 26 of 34
Initial pitch detection

MBE takes 2 steps to detect the refined initial pitch period
– Spectrum matching technique to find the initial pitch period
– Using DTW-based (Discrete Time Wrapping) technique to
smooth the estimation


Computational complexity is very high
In MBE, a modified three-level center clipped autocorrelation method is used to detect the initial pitch period,
and also use a simple smoothing method to correct the
pitch errors.
Page 27 of 34
Redefined pitch determination




To find the best pitch the basic method is to compute the error between
the original speech spectrum and the shaped voiced speech spectrum
by first supposing a pitch period
The supposed pitch of which the spectrum error is minimum is chosen
as the last pitch
To reduce the computational complexity, AMBE uses a 256- point
FFT to get the speech spectrum, and 5-point window spectrum is used
to form the voiced harmonic spectrum.
To get the refined pitch, AMBE perform seven times of spectrum
matching process. In every time. AMBE first set a supposed pitch, then
shape a harmonic spectrum over the overall frequency band according
to the supposed pitch and window spectrum, and an error can be
calculated by subtracting the shaped spectrum from speech spectrum.
After the seven times of matching process, the refined pitch can easily
be determined
Page 28 of 34
Dual rate coding
Page 29 of 34
Parameter
2400 bps
1200 bps
Pitch
quantization
8
6
V/UV decision
4
2
Amplitude
quantization
41
19
total
53
27
Improved MBE (IMBE)

A 2400 bps coder based on MBE
 Substantially better than U.S government
standard LPC-10e
 The parameters of the MBE speech model :
– the fundamental frequency
– voiced/unvoiced information
– the spectral envelope.
Page 30 of 34
IMBE algorithm

estimate the excitation and system parameters
which minimize the distance between the original
and synthetic speech spectra (analysis by
synthesis)
 Once these parameters are estimated,
voiced/unvoiced decisions are made by
comparing the spectral error over a series of
harmonics to a prescribed threshold
Page 31 of 34
IMBE block diagram

Page 32 of 34
IMBE algorithm
block diagram
IMBE Coding





IMBE offered in 2.4, 4.8 and 8.0 kbps
Analysis and synthesis routines are the same
except the bit allocation
The fundamental frequency needs accuracy of
about l Hz. and requires about 9 bits per frame.
The V/UV decisions are encoded with one bit
per decision.
The remaining bits are allocated to error
control and the spectral envelope information.
Page 33 of 33