LPC(Linear Predictive Coding)

PCM technique uses  64kbps for encoding the speech. But LPC needs just 13 kbps.
It ia an efficient speech analysis technique.
each frame is decided whether it is a buzz or hiss.
If it is a buzz, calculate the frequency and intensity and do the encoding.
For his there is no frequency. Only the intesity is calculated.

What is buzz and hiss?
Some consonants are produced with turbulent airflow, resulting in a hissy sound. If the sound source is periodic, it is buzz, it is chaotic it is hiss. Speech is a combination of buzz and hiss sounds.


In LPC-10e algorithm,  number 0 is treated as HISS and number greater than one  represents frquency of the buzz.
LPC-10e uses just 2.4kbps for speech.

What is residue?  Aim of LPC is to compress the speech. But certain portion of the speech could not be compressed. it  will not sit in the LPC model. This is residue. residue will take as many bits as the original speech signal.
Somehow this residue is to be compressed for efficiency. For this purpose  CODEBOOK is designed. A table is formed using the typical residual signals. The closest matching code is sent .
In the receive end, the residue corresponding to the CODE is used to EXCITE the FORMANT FILTER. Hence this technique is called as  Code Excited Linear Prediction (CELP).
CELP can encode at the rate of 4.8 kbps

LPC principle:
LPC starts with the assumption that the speech signal is produced by a buzzer at the end of a tube. The buzz produced by our throat+mouth is characterized by its intensity  and frequency. This throat+mouth forms a tube which is characterized by its resonances, which are called FORMANTSLPC analyzes the speech signal by estimating the FORMANTS, removing their effects from the speech signal, and estimating the intensity and frequency of the remaining buzz. The process of removing the formants is called inverse filtering, and the remaining signal is called the residue.
The numbers which describe the formants and the residue can be stored or transmitted somewhere else. LPC synthesizes the speech signal by reversing the process: use the residue to create a source signal, use the formants to create a filter (which represents the tube), and run the source through the filter, resulting in speech.