The MP3 lossy audio data compression algorithm takes advantage of a perceptual limitation of human hearing called auditory masking. In 1894, the American physicist Alfred M. Mayer reported that a tone could be rendered inaudible by another tone of lower frequency. In 1959, Richard Ehmer described a complete set of auditory curves regarding this phenomenon. Ernst Terhardt et al. created an algorithm describing auditory masking with high accuracy. This work added to a variety of reports from authors dating back to Fletcher, and to the work that initially determined critical ratios and critical bandwidths.
The psychoacoustic masking codec was first proposed in 1979, apparently independently, by Manfred R. Schroeder, et al. from Bell Telephone Laboratories, Inc. in Murray Hill, New Jersey, and M. A. Krasner both in the United States. Krasner was the first to publish and to produce hardware for speech (not usable as music bit compression), but the publication of his results as a relatively obscure Lincoln Laboratory Technical Report, did not immediately influence the mainstream of psychoacoustic codec development. Manfred Schroeder was already a well-known and revered figure in the worldwide community of acoustical and electrical engineers, but his paper was not much noticed, since it described negative results due to the particular nature of speech and the linear predictive coding (LPC) gain present in speech.
Both Krasner and Schroeder built upon the work performed by Eberhard F. Zwicker in the areas of tuning and masking of critical frequency bands, that in turn built on the fundamental research in the area from Bell Labs of Harvey Fletcher and his collaborators. A wide variety of (mostly perceptual) audio compression algorithms were reported in IEEE's refereed Journal on Selected Areas in Communications. That journal reported in February 1988 on a wide range of established, working audio bit compression technologies, some of them using auditory masking as part of their fundamental design, and several showing real-time hardware implementations.
The Moving Picture Experts Group (MPEG) was established in 1988 by the initiative of Hiroshi Yasuda (Nippon Telegraph and Telephone) and Leonardo Chiariglione. Yasuda was leading an initiative in Japan, called the Digital Audio and Picture Architecture (DAPA), while Chiariglione was leading an initiative in Europe, called the Coding of Moving Images for Storage (COMIS). Both eventually met in May 1988 to work on a global standard.
The genesis of the MP3 technology is fully described in a paper from Professor Hans Musmann, who chaired the ISO MPEG Audio group for several years. In December 1988, MPEG called for an audio coding standard. In June 1989, 14 audio coding algorithms were submitted. Because of certain similarities between these coding proposals, they were clustered into four development groups. The first group was MUSICAM, by Matsushita, CCETT, ITT and Philips. The second group was ASPEC, by AT&T, France Telecom, Fraunhofer Gesellschaft, Deutsche and Thomson-Brandt. The third group was ATAC, by Fujitsu, JVC, NEC and Sony. And the fourth group was SB-ADPCM, by NTT and BTRL.
The immediate predecessors of MP3 were "Optimum Coding in the Frequency Domain" (OCF), and Perceptual Transform Coding (PXFM). These two codecs, along with block-switching contributions from Thomson-Brandt, were merged into a codec called ASPEC, which was submitted to MPEG, and which won the quality competition, but that was mistakenly rejected as too complex to implement. The first practical implementation of an audio perceptual coder (OCF) in hardware (Krasner's hardware was too cumbersome and slow for practical use), was an implementation of a psychoacoustic transform coder based on Motorola 56000 DSP chips.
Another predecessor of the MP3 format and technology is to be found in the perceptual codec MUSICAM based on an integer arithmetics 32 sub-bands filterbank, driven by a psychoacoustic model. It was primarily designed for Digital Audio Broadcasting (digital radio) and digital TV, and its basic principles disclosed to the scientific community by CCETT (France) and IRT (Germany) in Atlanta during an IEEE-ICASSP conference in 1991, after having worked on MUSICAM with Matsushita and Philips since 1989.
This codec incorporated into a broadcasting system using COFDM modulation was demonstrated on air and on the field together with Radio Canada and CRC Canada during the NAB show (Las Vegas) in 1991. The implementation of the audio part of this broadcasting system was based on a two chips encoder (one for the subband transform, one for the psychoacoustic model designed by the team of
G. Stoll (IRT Germany), later known as psychoacoustic model I) and a real time decoder using one Motorola 56001 DSP chip running an integer arithmetics software designed by Y.F. Dehery's team (CCETT, France). The simplicity of the corresponding decoder together with the high audio quality of this codec using for the first time a 48 kHz sampling frequency, a 20 bits/sample input format (the highest available sampling standard in 1991, compatible with the AES/EBU professional digital input studio standard) were the main reasons to later adopt the characteristics of MUSICAM as the basic features for an advanced digital music compression codec.
During the development of the MUSICAM encoding software, Stoll and Dehery's team made a thorough use of a set of high quality audio assessment material selected by a group of audio professionals from the European Broadcasting Union and later used as a reference for the assessment of music compression codecs . The subband coding technique was found to be efficient, not only for the perceptual coding of the high quality sound materials but especially for the encoding of critical percussive sound materials (drums, triangle, ..) due to the specific temporal masking effect of the MUSICAM sub-band filterbank (this advantage being a specific feature of short transform coding techniques).
As a doctoral student at Germany's University of Erlangen-Nuremberg, Karlheinz Brandenburg began working on digital music compression in the early 1980s, focusing on how people perceive music. He completed his doctoral work in 1989. MP3 is directly descended from OCF and PXFM, representing the outcome of the collaboration of Brandenburg—working as a postdoc at AT&T-Bell Labs with James D. Johnston ("JJ") of AT&T-Bell Labs—with the Fraunhofer Institute for Integrated Circuits, Erlangen (where he worked with Bernhard Grill and four other researchers – "The Original Six"), with relatively minor contributions from the MP2 branch of psychoacoustic sub-band coders. In 1990, Brandenburg became an assistant professor at Erlangen-Nuremberg. While there, he continued to work on music compression with scientists at the Fraunhofer Society (in 1993 he joined the staff of the Fraunhofer Institute). The song "Tom's Diner" by Suzanne Vega was the first song used by Karlheinz Brandenburg to develop the MP3. Brandenburg adopted the song for testing purposes, listening to it again and again each time refining the scheme, making sure it did not adversely affect the subtlety of Vega's voice.
In 1991, there were two available proposals that were assessed for an MPEG audio standard: MUSICAM (Masking pattern adapted Universal Subband Integrated Coding And Multiplexing) and
ASPEC (Adaptive Spectral Perceptual Entropy Coding). As proposed by the Dutch corporation Philips, the French research institute CCETT, and the German standards organization Institute for Broadcast Technology, the MUSICAM technique was chosen due to its simplicity and error robustness, as well as for its high level of computational efficiency. The MUSICAM format, based on sub-band coding, became the basis for the MPEG Audio compression format, incorporating, for example, its frame structure, header format, sample rates, etc.
While much of MUSICAM technology and ideas were incorporated into the definition of MPEG Audio Layer I and Layer II, the filter bank alone and the data structure based on 1152 samples framing (file format and byte oriented stream) of MUSICAM remained in the Layer III (MP3) format, as part of the computationally inefficient hybrid filter bank. Under the chairmanship of Professor Musmann of the University of Hanover, the editing of the standard was delegated to Dutchman
Leon van de Kerkhof, to German
Gerhard Stoll, to Frenchman
Yves-François Dehery, who worked on Layer I and Layer II. ASPEC was the joint proposal of AT&T Bell Laboratories, Thomson Consumer Electronics, Fraunhofer Society and CNET. It provided the highest coding efficiency.
A working group consisting of van de Kerkhof, Stoll, Italian Leonardo Chiariglione (CSELT VP for Media), Frenchman
Yves-François Dehery, German Karlheinz Brandenburg, and American James D. Johnston (United States) took ideas from ASPEC, integrated the filter bank from Layer II, added some of their own ideas such as the joint stereo coding of MUSICAM and created the MP3 format, which was designed to achieve the same quality at 128 kbit/s as MP2 at 192 kbit/s.
The algorithms for MPEG-1 Audio Layer I, II and III were approved in 1991 and finalized in 1992 as part of MPEG-1, the first standard suite by MPEG, which resulted in the international standard ISO/IEC 11172-3 (a.k.a. MPEG-1 Audio or MPEG-1 Part 3), published in 1993. Files or data streams conforming to this standard must handle sample rates of 48k, 44100 and 32k and continue to be supported by current MP3 players and decoders. Thus the first generation of MP3 defined 14*3=42 interpretations of MP3 frame data structures and size layouts.
Further work on MPEG audio was finalized in 1994 as part of the second suite of MPEG standards, MPEG-2, more formally known as international standard ISO/IEC 13818-3 (a.k.a. MPEG-2 Part 3 or backwards compatible MPEG-2 Audio or MPEG-2 Audio BC), originally published in 1995. MPEG-2 Part 3 (ISO/IEC 13818-3) defined 42 additional bit rates and sample rates for MPEG-1 Audio Layer I, II and III. The new sampling rates are exactly half that of those originally defined in MPEG-1 Audio. This reduction in sampling rate serves to cut the available frequency fidelity in half while likewise cutting the bitrate by 50%.
MPEG-2 Part 3 also enhanced MPEG-1's audio by allowing the coding of audio programs with more than two channels, up to 5.1 multichannel. An MP3 coded with MPEG-2 results in half of the bandwidth reproduction of MPEG-1 appropriate for piano and singing.
A third generation of "MP3" style data streams (files) extended the MPEG-2 ideas and implementation but was named MPEG-2.5 audio, since MPEG-3 already had a different meaning. This extension was developed at Fraunhofer IIS, the registered patent holders of MP3 by reducing the frame sync field in the MP3 header from 12 to 11 bits. As in the transition from MPEG-1 to MPEG-2, MPEG-2.5 adds additional sampling rates exactly half of those available using MPEG-2. It thus widens the scope of MP3 to include human speech and other applications yet requires only 25% of the bandwidth (frequency reproduction) possible using MPEG-1 sampling rates. While not an ISO recognized standard, MPEG-2.5 is widely supported by both inexpensive Chinese and brand name digital audio players as well as computer software based MP3 encoders (LAME), decoders (FFmpeg) and players (MPC) adding 3*8=24 additional MP3 frame types. Each generation of MP3 thus supports 3 sampling rates exactly half that of the previous generation for a total of 9 varieties of MP3 format files. The sample rate comparison table between MPEG-1, 2 and 2.5 is given later in the article. MPEG-2.5 is supported by LAME (since 2000), Media Player Classic (MPC), iTunes, and FFmpeg.
MPEG-2.5 was not developed by MPEG (see above) and was never approved as an international standard. MPEG-2.5 is thus an unofficial or proprietary extension to the MP3 format. It is nonetheless ubiquitous and especially advantageous for low-bit rate human speech applications.
MPEG Audio Layer III versions
||First edition public release date
||Latest edition public release date
|MPEG-1 Audio Layer III
||ISO/IEC 11172-3 (MPEG-1 Part 3)
|MPEG-2 Audio Layer III
||ISO/IEC 13818-3 (MPEG-2 Part 3)
|MPEG-2.5 Audio Layer III
- The ISO standard ISO/IEC 11172-3 (a.k.a. MPEG-1 Audio) defined three formats: the MPEG-1 Audio Layer I, Layer II and Layer III. The ISO standard ISO/IEC 13818-3 (a.k.a. MPEG-2 Audio) defined extended version of the MPEG-1 Audio: MPEG-2 Audio Layer I, Layer II and Layer III. MPEG-2 Audio (MPEG-2 Part 3) should not be confused with MPEG-2 AAC (MPEG-2 Part 7 – ISO/IEC 13818-7).
Compression efficiency of encoders is typically defined by the bit rate, because compression ratio depends on the bit depth and sampling rate of the input signal. Nevertheless, compression ratios are often published. They may use the Compact Disc (CD) parameters as references (44.1 kHz, 2 channels at 16 bits per channel or 2×16 bit), or sometimes the Digital Audio Tape (DAT) SP parameters (48 kHz, 2×16 bit). Compression ratios with this latter reference are higher, which demonstrates the problem with use of the term compression ratio for lossy encoders.
Karlheinz Brandenburg used a CD recording of Suzanne Vega's song "Tom's Diner" to assess and refine the MP3 compression algorithm. This song was chosen because of its nearly monophonic nature and wide spectral content, making it easier to hear imperfections in the compression format during playbacks. Some refer to Suzanne Vega as "The mother of MP3". This particular track has an interesting property in that the two channels are almost, but not completely, the same, leading to a case where Binaural Masking Level Depression causes spatial unmasking of noise artifacts unless the encoder properly recognizes the situation and applies corrections similar to those detailed in the MPEG-2 AAC psychoacoustic model. Some more critical audio excerpts (glockenspiel, triangle, accordion, etc.) were taken from the EBU V3/SQAM reference compact disc and have been used by professional sound engineers to assess the subjective quality of the MPEG Audio formats. LAME is the most advanced MP3 encoder. LAME includes a VBR variable bit rate encoding which uses a quality parameter rather than a bit rate goal. Later versions 2008+) support an n.nnn quality goal which automatically selects MPEG-2 or MPEG-2.5 sampling rates as appropriate for human speech recordings which need only 5512 Hz bandwidth resolution.
A reference simulation software implementation, written in the C language and later known as ISO 11172-5, was developed (in 1991–1996) by the members of the ISO MPEG Audio committee in order to produce bit compliant MPEG Audio files (Layer 1, Layer 2, Layer 3). It was approved as a committee draft of ISO/IEC technical report in March 1994 and printed as document CD 11172-5 in April 1994. It was approved as a draft technical report (DTR/DIS) in November 1994, finalized in 1996 and published as international standard ISO/IEC TR 11172-5:1998 in 1998. The reference software in C language was later published as a freely available ISO standard. Working in non-real time on a number of operating systems, it was able to demonstrate the first real time hardware decoding (DSP based) of compressed audio. Some other real time implementation of MPEG Audio encoders and decoders were available for the purpose of digital broadcasting (radio DAB, television DVB) towards consumer receivers and set top boxes.
On 7 July 1994, the Fraunhofer Society released the first software MP3 encoder called l3enc. The filename extension .mp3 was chosen by the Fraunhofer team on 14 July 1995 (previously, the files had been named .bit). With the first real-time software MP3 player WinPlay3 (released 9 September 1995) many people were able to encode and play back MP3 files on their PCs. Because of the relatively small hard drives back in that time (~ 500–1000 MB) lossy compression was essential to store non-instrument based (see tracker and MIDI) music for playback on computer. As sound scholar Jonathan Sterne notes, "An Australian hacker acquired l3enc using a stolen credit card. The hacker then reverse-engineered the software, wrote a new user interface, and redistributed it for free, naming it "thank you Fraunhofer"".
In the second half of the 1990s, MP3 files began to spread on the Internet, often via underground pirated song networks. The first known experiment in Internet distribution was organized in the early 1990s by the Internet Underground Music Archive, better known by the acronym IUMA. After some experiments using uncompressed audio files, this archive started to deliver on the native worldwide low speed Internet some compressed MPEG Audio files using the MP2 (Layer II) format and later on used MP3 files when the standard was fully completed. The popularity of MP3s began to rise rapidly with the advent of Nullsoft's audio player Winamp, released in 1997. In 1998, the first portable solid state digital audio player MPMan, developed by SaeHan Information Systems which is headquartered in Seoul, South Korea, was released and the Rio PMP300 was sold afterwards in 1998, despite legal suppression efforts by the RIAA.
In November 1997, the website mp3.com was offering thousands of MP3s created by independent artists for free. The small size of MP3 files enabled widespread peer-to-peer file sharing of music ripped from CDs, which would have previously been nearly impossible. The first large peer-to-peer filesharing network, Napster, was launched in 1999. The ease of creating and sharing MP3s resulted in widespread copyright infringement. Major record companies argued that this free sharing of music reduced sales, and called it "music piracy". They reacted by pursuing lawsuits against Napster (which was eventually shut down and later sold) and against individual users who engaged in file sharing.
Unauthorized MP3 file sharing continues on next-generation peer-to-peer networks. Some authorized services, such as Beatport, Bleep, Juno Records, eMusic, Zune Marketplace, Walmart.com, Rhapsody, the recording industry approved re-incarnation of Napster, and Amazon.com sell unrestricted music in the MP3 format.