Basic MPEG knowledge

1. What is MPEG?

The full name of MPEG Moving Pictures Experts Group, a dynamic image Expert Group, is simply an expert group dedicated to studying audio and video specifications.

This expert group was established under the joint guidance of ISO/IEC (International standardsorganization/International Electrotechnical Commission, International Standardization Organization/International Electrotechnical Commission. This group is dedicated to studying a digital audio/video compression specification. Therefore, we finally find n Specifications suitable for different application environments.

Similar to the development process of things, it is also similar to studying so many norms. Different Versions of things are used for different applications in different periods. Also from simplicity to complexity. And the naming rules are all, according to Arabic numbers from small to big: MPEG-1, MPEG-2, MPEG-4, MPEG-7, the latest version, it seems that there is a MPEG-21.

2. Why is there no MPEG-3

It is estimated that someone wondered, why in the middle of the MPEG-3 how gone?

Is no MPEG-3, this is because at that time the designer did not plan well, resulting in a good design of the MPEG-2 performance is too good, can have the original intention to let mpeg3 dry live, so then simply don't design mpeg3, the original plan was canceled, there is no MPEG-3. For many people misunderstanding, think MP3 is MPEG-3, it is even more wrong. Next we will explain the origins of MP3 names.

3. What is MPEG-2.5?

Mpeg2.5, simply put, is not an official standard. MPEG 2.5 is an unofficial extended version for MPEG2 and supports lower sampling rates. For more information, find the following information online:

The MPEG sound standard provides three independent compression layers: layer I, layer II, and Layer III. You can select a layer based on your requirements and make your choice after weighing the complexity and sound quality.

A. layer I encoder is the simplest. The output data rate of the encoder is 384 kb/s. It is mainly used for digital compact cassette and DCC ).
B. The complexity of the encoder of layer II is moderate, and the output data rate of the encoder is 256 kb/s ~ 192 KB/s,
Its Applications include digital broadcast audio, DBA, digital music,
CD-I (compactdisc-interactive) and VCD (videocompact disc.
C. Layer III encoder is the most complex. The output data rate of the encoder is 64 KB/s, which is mainly used for Sound Transmission on ISDN.

MPEG-1 Layer III supports a sampling rate of 32, 44.1, 48 kHz, and a bit rate of 32 ~ 320 kbps.

The MPEG-2 Layer III supports a sampling rate of 16, 22.05, 24 kHz, and 8-bit rate ~ 160 kbps.

Fraunhofer extended this by dividing the low sampling rate supported by the original MPEG-2 by 2:
8, 11.025, and 12 kHz, the bit rate is the same as the MPEG-2, known as "MPEG 2.5 ".

4. Relationship between different frames in MPEG

For layer I and layer II, different frames are independent of each other. This means that you can intercept the audio file of MPEG at will, find the first correct frame header, decode it, and then put the remaining frames to play the video correctly. For Layer III, all frames are not necessarily independent, that is, they may be correlated. This uses byte reservoir. In the worst case, the first frame can be decoded only when nine frames of data are continuously cached.

5. MPEG-1 and MPEG-2 audio features

5.1 MPEG-1 audio (ISO/IEC 11172-3)

It describes three-layer Audio Encoding and has the following features:

A. One or two audio channels
B. Sampling Rate: 32 kHz, 44.1khz, or 48 khz
C. The bit rate ranges from 32 Kbps to 448 kbps.
D. Each layer has its own characteristics

5.2 MPEG-2 audio (ISO/IEC 13813-3)

It contains two extensions to the MPEG-1. Commonly known as MEGP-2/low sampling rate (LSF) and MPEG-2/multi-channel (multichannel ).

5.2.1 characteristics of MPEG-2/LSF:

A. One or two audio channels
B. sampling rate is only half of the MPEG-1
C. bitrate from 8kpbs to 256 kbps

5.2.2 features of MPEG-2/multi-channel:

A. Up to five full-range audio channels and one LFE (low frequency enhancement, also known as bass) Channel
B. The sampling rate is the same as the MPEG-1
C. For 5.1 channels, the maximum bit rate is 1 mpbs.

6. What are ISO/IEC 11172-3 and ISO/IEC 13818-3?

Because MPEG is only an organization under ISO/IEC, most of the MPEG audio specifications come from ISO/IEC.
Therefore, ISO/iec11172 and ISO/IEC 13818 are actually aliases of MPEG-1 and MPEG-2.

In addition, due to the MPEG-1 and MPEG-2, each has several parts, where the third part is about audio. So, ISO/IEC11172-3 and ISO/IEC13818-3, respectively, correspond to the MPEG-1 audio and MPEG-2 audio, that is, we often mention the MPEG audio files corresponding to the specification.

Refer:
[1] http://blog.163.com/againinput4@yeah/blog/static/122764271200981972911855