MPEG-1 Audio Layer II
From Wikipedia, the free encyclopedia
| MPEG-1 Audio Layer 2 | |
|---|---|
| File extension: | .mp2 |
| MIME type: | audio/mpeg |
| Type of format: | Audio |
MPEG-1 Audio Layer II (MP2, sometimes Musicam) is an audio codec defined by ISO/IEC 11172-3. An extension exists: MPEG-2 Layer II and is defined in ISO/IEC 13818-3. The file extension for files containing such audio data is usually .mp2. While it has largely been superseded by MP3 for PC and Internet applications, it remains a dominant standard for audio broadcasting as part of the DAB digital radio and DVB digital television standards. It is also used internally within the radio industry, for example in NPR's PRSS Content Depot programming distribution system.
For details and a short historic introduction to MP2, see MP3.
Contents |
[edit] History of Musicam development from MP2 to MP3
 |
This article does not cite any references or sources. (April 2007) Please help improve this article by adding citations to reliable sources. Unverifiable material may be challenged and removed. |
A lot of the early history of Musicam technological development is available in scientific communications (AES, IEEE), Internet documents and in German publications (see References). The Musicam technology (also called MPEG Audio Layer II by ISO) designed by CCETT, IRT, Philips was, according to H.G. Mussmann (ISO MPEG Audio chairman), the core of the MP3 standard. It was originally designed for the purpose of digital radio and TV broadcasting (DAB, DMB, DVB) but, once standardized by ISO MPEG was further promoted due to its Layer III (MP3) additional layer.It was solidified at the same time when Internet became widely used and widely available. Both MP2 (Musicam)and MP3 belong to the MPEG 1 Audio compatible family of standards. This compatibility reveals an interesting application in digital radio DMB receivers which allow for the real time transcoding of the broadcast Layer II (MP2) digital sound flow into MP3 stored music on flash memories.
Some important (mostly undocumented) events in the development of Musicam stand out.
- Musicam (as a psychoacoustical compression algorithm) was nearly perfectly developed especially with regards to glockenspiels (related to the xylophone) audio samples. It has been shown to be particularly efficient on high quality percussive sounds (impulses) thanks to the very efficient time-domain concealment characteristics of its polyphased filterbank.
- Musicam was chosen as the core of the MPEG Audio standard due to its especially high audio coding performances on highly critical audio material such as triangle, glockenspiel, castanet, symphonic orchestra, male and female voices. Subjective audio testing done by professional experts in the most critical conditions ever implemented have contributed to show its high level of transparency with respect to 16 bits and 18 bits audio sounds and to determine the appropriate level of compression for professional and consumer broadcast applications.
- In 1990, it was disclosed to the scientific community as a structured multi layer algorithm for the coding of high quality sounds with one of the simplest decoder in terms of hardware and software. The latter was easily implemented on integer DSP like Motorola 56000 or on more complex DSP families like Texas Instrument TMS 320x.
- Musicam is based on a perceptual digital codec which uses at the encoder a 32 subbands filterbank and a model of the human auditory system either monaural or binaural (the last one in joint stereo with stereo intensity). It has defined a compressed frame format based on 1152 sample intervals at various sampling rates with a sync word and a signalling header. This standardized frame format is consequently used also in the upper MP3 layer (Layer III).
- It took some 9 months and one extra layer (Layer III = MP3) of codec complexity to turn it into the well known MP3 format by introducing complementary signal processing tools such as additional MDCT transform, entropy coding and joint stereo mode (stereo intensity).
- AAC and to a lesser extent Ogg Vorbis and AC-3 audio codecs are still affected by the same fundamental problem in the codec model that the triangle, kabuki, glockenspiel and crysaglott revealed -- coding signals with complex impulses and high energy transients are always poorly reproduced.
This is the fundamental difference between Musicam (the MPEG1 Audio family) and the subsequent audio compression codecs. Musicam based codecs (MP2, MP3) have put first a focus on time domain critical audio sequences which are more typical in classical music and professional applications (studio) whereas subsequent codecs have ignored this demanding requirement to focus on less critical (frequency domain)light music audio materials.
- MP2 was proposed by the Advanced Television Research Consortium as a contender for the then US only ATSC standard's audio subsystem. This proposal existed in the drafts of the ATSC standard from versions (0.5 -- 0.9) but not in the final release version (1.0.
Musicam was subsequently replaced in the US with its standardized equivalent MPEG-2 Layer II for the DTV "Grand Alliance" shootout, but lost out to Dolby AC-3.
References:
Genesis of the MP3 Audio Coding Standard - Hans Georg Musmann: http://www.tnt.uni-hannover.de/~musmann/index.php
in IEEE Transactions on Consumer Electronics, Vol. 52, Nr. 3, pp. 1043-1049, August 2006
MUSICAM Source Coding - Yves-François Dehery, AES 10th International Conference: Kensington, London, England, (7-9 Sept 1991), pp 71-79.
The history of MP3 from Fraunhofer institute: http://www.iis.fraunhofer.de/fhg/iis/EN/bf/amm/mp3history/index.jsp
[edit] MPEG Layer II characteristics
MPEG-1 Layer II is defined in ISO/IEC 11172-3
- Sampling rates: 32, 44.1 and 48 kHz
- Bitrates: 32, 48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 256, 320 and 384 kbit/s
An extension has been provided in MPEG-2 Layer II and is defined in ISO/IEC 13818-3
- Additional sampling rates: 16, 22.05 and 24 kHz
- Additional bitrates: 8, 16, 24, 32, 40, 48, 56, 64, 80, 96, 112, 128, 144 and 160 kbit/s
Format
The format is based on successive digital frames of 1152 sampling intervals with four possible formats:
- mono format
- stereo format
- joint stereo format (stereo irrelevance)
- dual channel (uncorrelated) format
[edit] How the MP2 Codec works
- MP2 is a sub-band audio encoder, which means that compression takes place in the time domain with a low-delay filter bank producing 32 frequency domain components. By comparison, MP3 is a transform audio encoder with hybrid filter bank, which means that compression takes place in the frequency domain after a hybrid (=double) transformation from the time domain.
- MP2 also referred to as MPEG Audio Layer II (or Musicam) is the core algorithm of the MP3 standards. All psychoacoustical characteristics and frame format structures of the MP3 codec are derived from the basic MP2 algorithm and format.
- The MP2 encoder may exploit or not interchannel redundancies depending on its encoding mode. In pure stereophonic mode, this makes MP2 less efficient than MP3 on low bitrates (lower than 192 kbit/s). For example, a 128 kbit/s MP3 encoded audio usually sounds, to the human ear, truer to the original source than the same audio encoded as 192 kbit/s MP2. However MP2 can reach similar encoding performances to MP3 stereophonic mode thanks to its Joint Stereo coding mode which removes stereo intensity irrelevance.
- MP2 performs better than MP3 on high bitrates (192 to 384 kbit/s) and is generally more error resilient than MP3, so MP2 is considered optimal, and is the de facto standard, for broadcast applications. Typically, private broadcasters worldwide compress their material at 256kb/s while their counterparts in public broadcasting (including the Canadian Broadcasting Corporation, Channel Africa, Deutsche Welle, Radio France Internationale, Radio Canada International, Radio Netherlands, the SABC, the Singapore Broadcasting Corporation, and VOA, to cite a few) use 384kb/s.
- Like MP3, MP2 is a perceptual codec, which means that it removes information that the human auditory system will not be able to perceive. To choose which information to remove, the audio signal is analyzed according to a psychoacoustic model, which takes into account the parameters of the human auditory system. Research into psychoacoustics has shown that if there is a strong signal on a certain frequency, then weaker signals at frequencies close to the strong signal's frequency cannot be perceived by the human auditory system. This is called frequency masking. Perceptual audio codecs take advantage of this frequency masking by ignoring information at frequencies that are deemed to be imperceptible, thus allowing more data to be allocated to the reproduction of perceptible frequencies.
- MP2 splits the input audio signal into 32 sub-bands, and if the audio in a sub-band is deemed to be imperceptible then that sub-band is not transmitted. MP3, on the other hand, transforms the input audio signal to the frequency domain in 576 frequency components. Therefore, MP3 has a higher frequency resolution than MP2, which allows the psychoacoustic model to be applied more accurately than for MP2. Because the psychoacoustic model can be applied more accurately, MP3 has greater scope to reduce the bit rate.
- The use of an additional entropy coding tool and this higher frequency accuracy justify why MP3 doesn't need as high a bit rate as MP2 to get an acceptable audio quality. Inversely MP2 shows a better behaviour than MP3, in the time domain, due to its lower frequency resolution which implies less codec time delay - easier time-editing resolution for cut and paste in professional studio - native ruggedness to the digital recording and digital transmission errors.
- Moreover, MP2 sub-band filter bank provides an inherent transient concealment feature due to the specific temporal masking effect of its mother filter. This unique characteristics of the MPEG 1 Audio family codecs implies a very good sound quality on audio signals with rapid energy changes such as percussive sounds both on the MP2 and the MP3 codecs which use the same basic sub-band filter bank.
Musepack is an MP2-based encoder with several improvements which make it perform better on lower bitrates.
[edit] Use in DVD-Video
All DVD-Video players in PAL countries contain stereo MP2 decoders, making MP2 a possible competitor to Dolby Digital in these markets. DVD-Video players in NTSC countries are not required to decode MP2 audio, though most do. While some DVD recorders record in MP2 and many consumer-authored DVDs use the format, as of 2006 mass produced discs with MP2 soundtracks were very rare.
[edit] Use in VCD and SVCD
MPEG layer 2 is the standard audio format used in the Video CD and Super Video CD formats (SVCD also supports the multichannel variant), as well as the related China Video Disc format.
[edit] Naming confusion
The term MP2 and filename extension .MP2 usually refer to data in the MPEG-1 Audio Layer II format, but can also refer to MPEG-2 BC Audio Layer II, a backwards compatible extension which adds support for multi-channel sound and lower sampling rates. The abbreviation MP2 is also sometimes erroneously used to refer to MPEG-2 video or MPEG-2 AAC audio formats.
[edit] See also
[edit] External links
Sites with further information
- MPEG Audio Resources and Software
- TooLAME - an MP2 encoder
- TwoLAME - a fork of the tooLAME code
- RFC 3003 - The document defining MIME type for MPEG-1 Audio Layer II
