The Synthesis of Speech Based On LPC
Team Members and Roles:
Name Number Role
Hu Qian 200531200215 Make PowerPoint slides
Qu Yongxin 200631200298 Write MATLAB codes and give the final presentation
Liu Chaojun 200631200154 Write the final report
Liang Juanlan 200631200148 Research and collect useful information and make
PowerPoint slides
Project Description:
The past decade has witnessed substantial progress towards the application of low-rate speech
coders to civilian and military communications as well as computer-related voice applications.
And the Linear Predictive Coding (LPC) method plays an important role in it.
LPC is particularly effective in manipulating speeches. LPC is one of the most powerful speech
analysis techniques, and also one of the most useful methods for encoding good quality speech at
a low bit rate. It provides extremely accurate estimates of speech parameters, and also relatively
efficient for computation.
However, with LPC to reach the same goal, there are different ways. In this project, we have
taken two methods to recreate speeches. One is easier and the other is harder. Through these
computations, we can gain more understanding of LPC and show how linear predictive coding
mathematically approximates the parameters of the vocal tract to reduce a speech signal to a state
which is noticeably synthetic but still understandable. We can tell the differences between them,
too.
Introduction:
Under normal circumstances, speech is sampled at 8000 samples/second with 8-bit to represent
each sample. This provides a rate of 64000 bits/second. LPC reduces this to 2400 bits/second by
breaking the speech into segments, and then sending the voiced/unvoiced information, the pitch
period and the coefficients of the filters that represent the vocal tract for each segment.
At this reduced rate, the speech has a distinctive synthetic sound. And there is a noticeable loss of
quality. However, the speech is still audible and it can be also easily understood. Since there is
information loss in linear predictive coding, it is a loss form of compression.
The application of linear predictive analysis to estimate speech parameters is often called linear
predictive coding (LPC).
LPC is one of the most powerful speech analysis techniques, and one of the most useful
methods for encoding good quality speech at a low bit rate. It provides extremely accurate
estimates of speech parameters, and is relatively efficient for computation.
The linear predictive method of speech analysis approximates the basic parameters of speech.
This method is based on the assumption that a speech sample can be approximated as a linear
combination of previous speech samples.
The following figure illustrates the block diagram of prediction:
Z
r
S[n + r] Error
Output
+
-
S[n]
