Post on 19-Jul-2015
transcript
Speech /Audio Coding Standard LPC-10
By
Sonawane Swapnil R
511006
Sub.: Speech
DEP-E&TC
VIIT Pune
1
What is Speech Coding?
• “Speech coding" = finding a representation of speech which can be transmitted efficiently through a digital channel.
• It is usually lossy coding, meaning that the waveform can not be completely reproduced by the decoder, instead, only the information which is useful to a human listener is retained.
2
Coding Algorithm
• ADPCM: Adaptive Differential Pulse Code Modulation
• LPC-10: LPC Vocoder with 10 coefficients
• CELP: Code Excited LPC
• RPE-LTP: Regular Pulse Excited LPC with Long Term Prediction
• VSELP: Vector Sum Excited LPC
• IMBE: Improved Multi-Band Excitation
3
LPC – 10/ FS-1015 :-
• BRIEF HISTORY :-
– Is a secure telephony speech encoding standard developed bythe United States Department of Defense and later by NATO. Thestandard was finished 1984.
– Algorithm uses linear predictive coding vocoder.
– The vocoder enables understandable speech, but the quality isvery unnatural and synthetic.
4
PROPERTIES :-
• 10 LP(linear predictor) coefficients are used.
• Bandwidth: 2.4kbps
• Samples/frame : 180 samples
• Bits/frame: 54 bits
• Frame Size: 22.5ms = 44.44 frames/sec
5
ANALYSIS :-
6
Conti..
• Analysis process extract from the speech signal the parametersrequired to model it.
• First parameter :- Type of speech signal (voiced or unvoiced).
• The result is a voicing indicator,
• When voiced, its period has to be estimated in order to reflect itsheight. This period, called as pitch .
7
1 Voiced segment
0 Unvoiced segment
Conti..
• The result of this analysis is a set of ten reflection coefficients(hence the name LPC 10) which sufficiently & faithfully describe thecross-sectional variations in the vocal tract.
• Finally, for each frame, the level of the speech signal is evaluated inorder to control the gain of the synthesizer on the synthesis side.
8
SYNTHESIS:-
9
Conti..
• The algorithms employed to synthesize the speech signal reflect theassumed speech production model.
• They include, in succession:
– a noise generator, used for unvoiced sounds;
– a periodic signal generator, to which the pitch is provided, for voicedsounds;
– a switch allowing selection of either generator according to the type ofspeech signal to be produced in the current frame;
10
Conti..
– a filter of order 10, which filters the excitation selected; it is at thislevel that the distinction between the different vowels and thedifferent consonants is made;
– a gain control system, which gives the synthetic signal the rightvolume;
– optionally, a “post-filtering” system, designed to mask certainimperfections in the synthesizer and to make the synthesized signalmore pleasant to the human ear.
11
VOCODER AT 2,400 BIT/S:-
12
LP Coefficients Pitch Voicing Energy
0 41 48 53
- The remaining 1 bit is for synchronization
SPEECH CODER COMPARISON:-
13
APPLICATIONS
• Digital telephony
• Satellite bradcasting
• Radio communications with secure voice transmissions
14
THANKU
15