Sunteți pe pagina 1din 22

Audio and Video Compression

Chapter 4

Dr. Saeed Mahmud Ullah


Associate Professor
Electrical and Electronic Engineering
University of Dhaka
Audio Signal
• Audio signal is basically a continuously varying streams of data.
• Audio digitization process known as Pulse Code Modulation (PCM) involves sampling the
analogue audio signal
- minimum rate which is twice maximum frequency
- OR as determined by the BW of the communications channel - bandlimited signal
 Speech signal
 Maximum audio frequency- 10 kHz
 Sampling rate – 20 kHz
 Number of bits/sample- 12
 Bit rate- 240 kbps

 Maximum music frequency- 20 kHz


 Sampling rate – 40 kHz
 Number of bits/sample- 16
 Bit rate- 1.28 kbps (stereophonic) 2
Audio Signal compression

 Two ways to reduce bit rate


 Audio signal is sampled at lower bit rate/
fewer bits per sample
 Utilizing compression algorithms

3
Lecture 19 ICT4205: MC

Audio Compression
Differential Pulse code Modulation (DPCM)
• It is derivative of standard PCM and exploits the fact that, for most audio signals, the range
of the differences in amplitude between successive samples of the audio waveform is less
than the range of the actual sample amplitudes.
• Hence fewer number of bits are required than for a comparable PCM signal with the same
sampling rate.
• Essentially, previous digitized sample of a analog signal put in a register (R) – a temporary
storage facility.
• The difference signal (DPCM) is computed by subtracting the current contents of the
register (R0) from the new digitized sample output by the ADC (PCM).
• Value in register is updated: Computed difference signal is added to the previously
computed signal held in register
• Typical savings with DPCM, are limited to 1 bit: for a standard PCM voice signal reduces
bit rate requirement from 64 kbps to 56 kbps 4
Lecture 19 ICT4205: MC

Audio Compression

Sample N

Ro = Current content of register R and R1 = new/updated contents

5
6
ITU Recommendation G.721
8-order predictor is used

For higher difference value: 6


bit is used

For lower difference value: 5 bit


is used

7
ITU Recommendation G.722

8
9
1. Sensitivity of ear
2. Frequency Masking
3. Temporal Masking

10
11
Video Compression
20
21
22

S-ar putea să vă placă și