8/12/2019 Sub Band Coding
1/16
SUB BAND CODING
S.I.Rosaline
II sem/ COS
8/12/2019 Sub Band Coding
2/16
Introduction
Compression schemes depend on datacharacteristics.
Vector Quantizationsource output has high
degree of clustering. Differential Encodingsource output sample
to sample difference is small.
Scalar Quantizationsource output is random
But, most source outputs exhibit acombination of characteristics.
4/6/2014 2Rosaline
8/12/2019 Sub Band Coding
3/16
Introduction(contd..)
Decomposing source blocks into different
frequency bandrequires TRANSFORM
CODING.
DisadvantageBlocking (generation of coding
artifacts at the block edges).
Overcome by LOT( Lapped Orthogonal
Transform).
4/6/2014 3Rosaline
8/12/2019 Sub Band Coding
4/16
Example
4/6/2014 4Rosaline
8/12/2019 Sub Band Coding
5/16
Consider a sequence {xn} as shown above.
Inference - fast sample to sample variation
- slow varying trend (dotted lines)
To extract trendavg. samples in a moving
windowsmooths out rapid variations, makes
evident the slow trend.
Avg {xn} , window size -2generate new
sequence {yn} .
4/6/2014 5Rosaline
8/12/2019 Sub Band Coding
6/16
Now {yn} easily encoded compared with {xn}.
However, we want to encode {xn} . Hence
done by, encoding {yn} by difference equation{zn}.
Decomposinganalysis
Recomposingsynthesisimplemented using
discrete time filters.
4/6/2014 6Rosaline
8/12/2019 Sub Band Coding
7/16
FILTERS
Filter - A system that isolates different
frequency components.
Low pass filter, high pass filter, Band pass filter
cut off frequency.
Magnitude Transfer function - ratio of
magnitude of input & output of the filter as a
function of frequency.
4/6/2014 7Rosaline
8/12/2019 Sub Band Coding
8/16
Ideal & Practical Magnitude Response
Low Pass Filter
4/6/2014 8Rosaline
8/12/2019 Sub Band Coding
9/16
Realistic filtersripple occurs in pass band &
stop band.
Filtersdigital in nature. Hence analog s/gundergoes sampling first.
SamplingNyquist theorem2f0 .
If Nyquist ruleviolatedaliasing occursoriginal signal reconstruction is impossible
anti aliasing filter used.
Input output relation of the filter is:
4/6/2014 9Rosaline
8/12/2019 Sub Band Coding
10/16
Filters used in Sub Band Coding
4/6/2014 10Rosaline
8/12/2019 Sub Band Coding
11/16
Filter in sub band codingQuadrature Mirror
Filter.
High pass impulsive response is
Filters are symmetric :
Filters with fewer taps are less efficient than
that with more taps. ExampleJohnston eight tap filter & Smith
Barnwell eight tap filter.
4/6/2014 11Rosaline
8/12/2019 Sub Band Coding
12/16
4/6/2014 12Rosaline
8/12/2019 Sub Band Coding
13/16
Basic Sub Band Coding Algorithm
4/6/2014 13Rosaline
8/12/2019 Sub Band Coding
14/16
Analysis
Source output passed thro analysis filter bank
Passband of the filters can be overlapping ornon overlapping.
Reducing the number of samples-
decimation/downsampling. Amount of decimation depends on the ratio of
the bandwidth of the filter output to the filter
input. After decimation, the output is encoded using
ADPCM, PCM, or vector quantization.
4/6/2014 14Rosaline
8/12/2019 Sub Band Coding
15/16
Non Overlapping & Overlapping filter
banks
4/6/2014 15Rosaline
8/12/2019 Sub Band Coding
16/16
Synthesis
Quantization and codingQuantized & codedcoefficients used to reconstruct the originalsignal. (Bit Allocation)
Encoded samplesDecodedUpsampledpassed thro a bank of reconstruction filters output added to get the final output.
Three major componentsanalysis &synthesis filters, Bit allocation, EncodingScheme.
4/6/2014 16Rosaline