4.4. Audio Compression

Simple Audio Compression Methods
Psychoacoustics
MPEG Audio Compression

Reference: Chapter 6 of Steinmetz and Nahrstedt

Reference: Davis Pan, "A Tutorial on MPEG/Audio Compression", IEEE Multimedia, Vol. 2, No. 2, pp. 60-74, 1995.

  

4.4.1 Simple Audio Compression Methods


The following are some of the Lossy methods:

  

4.4.2 Psychoacoustics


Human hearing and voice

Critical Bands

Sensitivity of human hearing in relation to frequency

Frequency Masking

Question: Do receptors interfere with each other?

Temporal masking

  

4.4.3 MPEG Audio Compression


Some facts

Steps in algorithm:

  1. Use convolution filters to divide the audio signal (e.g., 48 kHz sound) into 32 frequency subbands --> subband filtering.

  2. Determine amount of masking for each band caused by nearby band using the psychoacoustic model shown above.

  3. If the power in a band is below the masking threshold, don't encode it.

  4. Otherwise, determine number of bits needed to represent the coefficient such that noise introduced by quantization is below the masking effect (Recall that one fewer bit of quantization introduces about 6 dB of noise).

  5. Format bitstream

Example:

MPEG Layers

Effectiveness of MPEG audio

Layer

Target

Bit-rate

Ratio

Quality at

64 kb/s

Quality at

128 kb/s

Theoretical

Min. Delay

Layer 1

192 kb/s

4:1

---

---

19 ms

Layer 2

128 kb/s

6:1

2.1 to 2.6

4+

35 ms

Layer 3

64 kb/s

12:1

3.6 to 3.8

4+

59 ms

Further Exploration

MPEG Resources on the Web.
  
Top | Chap 4 | CMPT 365 Home Page | CS