MP3 (formally MPEG-1 Audio Layer III or MPEG-2 Audio Layer III ) is an audio encoding format for digital audio. Originally defined as the third audio format of the MPEG-1 standard, it is maintained and extended again - defines additional bit rate and support for more audio channels - as the third audio format of the next MPEG-2 standard. The third version, known as MPEG 2.5 - is expanded to support better bit rates - generally applied, but not a recognized standard.
MP3 (or mp3 ) as a file format generally designates a file containing the elementary stream of audio data and MPEG-1 encoding video, without any other complexity than the MP3 standard.
In the MP3 aspect relating to audio compression - the clearest standard aspect for end users (and best known) - MP3 uses lossy data compression to encode data using imprecise estimates and partial data being discarded. This enables a large file size reduction when compared to uncompressed audio. The combination of small size and acceptable fidelity led to an explosion in music distribution over the Internet in the mid to late 1990s, as the technology allowed when bandwidth and storage were still in premium condition. The MP3 format soon became associated with the controversy surrounding copyright infringement, music piracy, ripping files/sharing services MP3.com and Napster, among others. With the advent of portable media players, product categories also include smartphones, MP3 support remains near universal.
MP3 compression works by reducing (or predicting) the accuracy of certain sound components that are considered to be beyond the hearing ability of most humans. This method is usually referred to as perceptual coding, or psychoacoustic modeling. The remaining audio information is then recorded in an efficient manner of space. Compared to CD-quality digital audio, MP3 compression can generally reach 75 to 95% reduction in size. For example, MP3s encoded at a constant bitrate of 128 kbit/s will produce files about 9% of the original audio CD size.
Also designed as a streamable format, the transmission segment can disappear without affecting the ability to decode the next segment.
MP3 is designed by the Mobile Image Experts Group (MPEG) as part of the MPEG-1 standard, and then MPEG-2. The first sub-groups for audio were formed by several teams of engineers at CCETT, Matsushita, Philips, Sony, AT & amp; T-Bell Labs, Thomson-Brandt, and others. MPEG-1 Audio (MPEG-1 Part 3), which includes MPEG-1 Audio Layer I, II and III, was approved as a draft committee for ISO/IEC standards in 1991, completed in 1992, and published in 1993 as ISO/IEC 11172-3: 1993. An MPEG-2 Audio compatible backward (MPEG-2 Part 3) extension with sample and lower bit rate was published in 1995 as ISO/IEC 13818-3: 1995.
Video MP3
Histori
Development
The MP3 lossy audio data compression algorithm takes advantage of the perceptual limitations of human hearing called auditory hearing. In 1894, American physicist Alfred M. Mayer reported that a tone could be made inaudible by another tone with a lower frequency. In 1959, Richard Ehmer described a complete set of auditory curves on this phenomenon. Ernst Terhardt et al. create an algorithm that describes hearing masking with high accuracy. This work was added to various reports from authors dating back to Fletcher, and for work that initially determined critical ratios and critical bandwidth.
The psychoacoustic masking codex 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 in the United States. Krasner was the first to publish and produce speech hardware (can not be used as musical bit compression), but the publication of his results as a relatively unclear Lincoln Laboratory Technical Report did not immediately affect the mainstream of the development of psychoacoustic codecs. Manfred Schroeder has become a well-known and respected figure in the worldwide acoustic and electrical engineering community, but his paper is not overly concerned, as it illustrates the negative results because of certain speech properties and the linear predictive encoding (LPC) present in speech.
Both Krasner and Schroeder are built on the work done by Eberhard F. Zwicker in the field of tuning and masking critical frequency bands, which in turn are built on fundamental research in the area from Bell Labs of Harvey Fletcher and his colleagues. Various audio compression algorithms (mostly perceptions) are reported in IEEE Referee Journals in Selected Areas of Communication. The journal reported in February 1988 on a variety of established and working audio compression technologies, some of which use auditory hearing as part of their fundamental design, and some show real-time hardware implementations.
The Moving Picture Experts Group (MPEG) was founded in 1988 by the initiative of Hiroshi Yasuda (Nippon Telegraph and Telephone) and Leonardo Chiariglione. Yasuda leads the initiative in Japan, called Digital Audio and Picture Architecture (DAPA), while Chiariglione leads an initiative in Europe, called Coding of Moving Images for Storage (COMIS). The two eventually met in May 1988 to work on global standards.
Genesis MP3 technology is fully explained in a paper by Professor Hans Musmann, who leads the ISO MPEG Audio group for several years. In December 1988, MPEG called for audio coding standards. In June 1989, 14 audio encoding algorithms were submitted. Due to certain similarities between these coding proposals, they are grouped into four development groups. The first group is MUSICAM, by Matsushita, CCETT, ITT and Philips. The second group is ASPEC, by AT & amp; T, France Telecom, Fraunhofer Gesellschaft, Deutsche and Thomson-Brandt. The third group is ATAC, by Fujitsu, JVC, NEC and Sony. And the fourth group is SB-ADPCM, by NTT and BTRL.
Direct MP3 predecessors are "Optimum Coding in Frequency Domain" (OCF), and Perceptual Transform Coding (PXFM). These two codecs, along with block-switching contributions from Thomson-Brandt, are merged into a codec called ASPEC, submitted to MPEG, and which win the quality competition, but it is mistakenly rejected because it is too complicated to implement. The first practical implementation of audio perceptual coder (OCF) in hardware (Krasner hardware is too complicated and slow for practical use), is the implementation of a psychoacoustic transformation coder based on the Motorola 56000 DSP chip.
Other predecessors of the MP3 format and technology can be found in perceptual codecs based on integer arithmetic 32 sub-band filterbank, driven by the psychoacoustic model. It was primarily designed for Digital Audio Broadcasting (digital radio) and digital TV, and its basic principles were expressed to the scientific community by CCETT (France) and IRT (Germany) in Atlanta during the IEEE-ICASSP conference in 1991, after working on MUSICAM along with Matsushita and Philips since 1989.
The codec was incorporated into the broadcast system using COFDM modulation shown on the air and in the field along with Canadian Radio Canada and CRC during the NAV (Las Vegas) event in 1991. The audio part of the broadcasting system was based on two encoder chips (one for subband transformation, one for the psychoacoustic model designed by the G. Stoll (German IRT) team, later known as the psychoacoustic model I) and real time decoder using a single Motorola 56001 DSP chip running arithmetic integer software designed by YF Tim Dehery (CCETT, France). The simplicity of the corresponding decoder along with the high audio quality of this codec uses for the first time 48 kHz sampling frequency, 20 bit/sample input format (the highest sampling standard available in 1991, compatible with AES/EBU professional digital studio input standard) is the main reason for later adopting the characteristics of MUSICAM as a basic feature for a sophisticated digital music compression codec.
During the development of MUSICAM encoding software, the Stoll and Dehery team made a thorough use of a set of high quality audio assessment materials selected by a group of audio professionals from the European Broadcasting Union and later used as a reference for musical compression assessment. codecs. Subband coding techniques are found to be efficient, not only for the perceptual coding of high quality sound materials but especially for encoding critical percussion sounds (drums, triangles,..) due to specific temporal masking effects of MUSICAM. sub-band filterbank (this advantage being a special feature of the short transformation coding technique).
As a doctoral student at the University of Erlangen-Nuremberg Germany, 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 derived from OCF and PXFM, representing the results of the Brandenburg collaboration - working as a postdoc in AT & amp; 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 a relatively small contribution from sub-band coders psikooustic MP2. In 1990, Brandenburg became an assistant professor at Erlangen-Nuremberg. While there, he continued to work on musical compression with scientists at the Fraunhofer Society (in 1993 he joined staff from the Fraunhofer Institute). The song "Tom's Diner" by Suzanne Vega is 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 affect Vega's subtlety.
Standardization
In 1991, there were two proposals available that were rated for the MPEG audio standard: MUSICAM ( M request pattern adjusted to U niversal S ubband I ntegrated C oding A nd M ultiplexing) and ASPEC ( A daptive S p < P erceptual E ntropy C oding). As proposed by the Dutch company Philips, the French research institute CCETT, and the German standards organization Institute for Broadcast Technology, MUSICAM's technique was chosen for its simplicity and robustness, and for its high computing efficiency. The MUSICAM format, based on sub-band coding, forms the basis for the MPEG Audio compression format, combining, for example, frame structure, header format, sample rate, etc.
While many MUSIC technologies and ideas are incorporated into the definitions of MPEG Audio Layer I and Layer II, only bank filters and data structures based on 1152 instances of framing (file format and byte-oriented flow) of MUSICAM remain in Layer III (MP3), as part of the bank hybrid filters that are not computationally efficient. Under the leadership of Professor Musmann of Hanover University, standard editing was delegated to the Dutchman Leon van de Kerkhof, to German Gerhard Stoll, to the Frenchman Yves-FranÃÆ'§ois Dehery, who worked on Layer I and Layer II. ASPEC is a joint proposal of AT & amp; T Bell Laboratories, Thomson Consumer Electronics, Fraunhofer Society and CNET. It provides the highest coding efficiency.
The working group consisting of van de Kerkhof, Stoll, Leonardo Chiariglione Italia (CSELT VP for Media), French Yves-FranÃÆ'§ois Dehery, Karlheinz Brandenburg Germany, and American James D. Johnston (USA) took the idea of ​​ASPEC, bank of Layer II, adding some of their own ideas like stereo coding together MUSICAM and creating MP3 format, designed to achieve the same quality at 128 Kbit/s as MP2 at 192 Kbit/s.
The Algorithm for MPEG-1 Audio Layer I, II and III was approved in 1991 and completed in 1992 as part of MPEG-1, the first standard suite by MPEG, resulting in an international standard ISO/IEC 11172-3 (aliased MPEG-1 Audio or MPEG-1 Part 3 ), published in 1993. Files or data streams conforming to this standard shall handle sample rates of 48k, 44100 and 32k and continue to be supported by the current MP3 player and decoder. So the first generation of MP3 is defined 14 * 3 = 42 interpretation of MP3 frame data structure and layout size.
Further work on MPEG audio was completed in 1994 as part of the second suite of MPEG standards, MPEG-2, better known formally as an international standard ISO/IEC 13818-3 (aka MPEG-2 Part 3 or backward compatible MPEG-2 Audio or MPEG-2 Audio BC ), originally published in 1995. MPEG-2 Part 3 (ISO/IEC 13818 -3) sets an additional 42 bit rate and sample rate for MPEG-1 Audio Layer I, II and III. The new sampling rates are exactly half that originally defined in MPEG-1 Audio. This decrease in sampling rate serves to cut the available frequency fidelity in half while also cutting the bitrate up to 50%. MPEG-2 Part 3 also enhances MPEG-1 audio by allowing encoding of audio programs with more than two channels, up to 5.1 multichannel. MPEG-2 coded MP3 produces half of MPEG-1's corresponding MPEG-1 bandwidth reproduction for piano and song.
The third generation of "MP3" style data streams extends the MPEG-2 idea and implementation but is named MPEG-2.5 audio because MPEG-3 already has a different meaning. This extension was developed in Fraunhofer IIS, the registered patent holder of MP3s by reducing the frame sync field in MP3 headers from 12 to 11 bits. As in the transition from MPEG-1 to MPEG-2, MPEG-2.5 adds an additional half of the additional sampling rate available using MPEG-2. Thus expanding the MP3 coverage to include human speech and other applications but requires only 25% of the bandwidth (frequency reproduction) possibly using the MPEG-1 sampling rate. Although not an ISO-certified standard, MPEG-2.5 is widely supported by digital digital players and cheap brand names as well as computer-based software MP3 encoders (LAME), decoders (FFmpeg) and players (MPC) adding 3 * 8 = 24 frame types Additional MP3. Each MP3 generation thus supports 3 precise sampling levels half 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 then given in the article. MPEG-2.5 is supported by LAME (since 2000), Media Player Classic (MPC), iTunes, and FFmpeg.
MPEG-2.5 is not developed by MPEG (see above) and has never been approved as an international standard. MPEG-2.5 is thus an unofficial or exclusive extension to the MP3 format. Nevertheless it is ubiquitous and especially advantageous for the application of low level human speech bits.
- ISO standard ISO/IEC 11172-3 (alias MPEG-1 Audio) menetapkan tiga format: MPEG-1 Audio Layer I, Layer II dan Layer III. ISO standard ISO/IEC 13818-3 (a.k.a. MPEG-2 Audio) mendefinisikan versi tambahan MPEG-1 Audio: MPEG-2 Audio Layer I, Layer II dan Layer III. MPEG-2 Audio (MPEG-2 Bagian 3) tidak boleh disamakan dengan MPEG-2 AAC (MPEG-2 Part 7Â – ISO/IEC 13818-7).
The efficiency of compression encoders is usually determined by bit rate, because the compression ratio depends on the bit depth and the sampling rate of the input signal. However, the compression ratio is often published. They can use the Compact Disc (CD) parameter as a reference (44.1 kHz, 2 channels at 16 bits per channel or 2ÃÆ'â € "16 bits), or sometimes the Digital Audio Ribbon (DAT) parameter SP (48 kHz, 2ÃÆ'â € "16 bits). This last reference compression ratio is higher, indicating a problem with the use of the term compression ratio for a lossy encoder.
Karlheinz Brandenburg used the recording of Suzanne Vega's song CD "Tom's Diner" to assess and refine the MP3 compression algorithm. The song was chosen for its almost monophonic nature and broad spectrum content, making it easier to hear imperfections in the compression format during playback. Some refer to Suzanne Vega as "MP3 Mother". This particular track has an interesting property because two channels are almost, but not entirely, the same, leading to a case where Binaural Level Binaural Depression causes spatial unmasking noise artifacts unless the encoder correctly recognizes the situation and implements similar corrections detailed in the psychoacoustic model AAC MPEG-2. Some of the more important audio quotes (glockenspiel, triangle, accordion, etc.) are 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 format. LAME is the most advanced MP3 encoder. LAME includes a VBR variable bit rate encoding that uses quality parameters rather than bit rate objectives. Then version 2008) supports n.nnn quality targets that automatically select the MPEG-2 or MPEG-2.5 sampling rate appropriate for human speech recording that requires only 5512 Hz bandwidth resolution.
Become public
The implementation of reference simulation software, written in C language and later known as ISO 11172-5 , was developed (in 1991-1996) by members of the MPEG Audio ISO committee to produce bit compliant. MPEG Audio Files (Layer 1, Layer 2, Layer 3). It was approved as a draft of the ISO/IEC technical reporting committee in March 1994 and printed as a CD document 11172-5 in April 1994. It was approved as a draft technical report (DTR/DIS) in November 1994, completed in 1996 and published as an international standard ISO/IEC TR 11172-5: 1998 in 1998. The reference software in C language was then published as a freely available ISO standard. Working non-real time on a number of operating systems, it is able to demonstrate the first real time hardware decoding (DSP based) from compressed audio. Some real-time implementations of the MPEG Audio encoder and decoder are available for digital broadcasting purposes (DAB radio, DVB television) to the consumer receiver and set the top box.
On July 7, 1994, Fraunhofer Society released the first MP3 encoder software called l3enc. The file extension .mp3 was selected by the Fraunhofer team on July 14, 1995 (previously, the file has been named .bit ). With the first real-time MP3 player software, WinPlay3 (released September 9, 1995), many people can encode and play MP3 files on their PC. Because of the relatively small hard drive at that time (~ 500-1000 MB), lossy compression is essential for storing non-instrument based music (see tracker and MIDI) for playback on the computer. As an expert Jonathan Sterne notes, "An Australian hacker takes over using a stolen credit card, hackers then reverse engineer the software, write a new user interface, and distribute it for free, calling it" thank you Fraunhofer "".
Internet distribution
In the second half of the 1990s, MP3 files began to spread on the Internet, often through underground pirated track networks. The first known experiment in Internet distribution was organized in the early 1990s by the Internet Underground Music Archive, better known as the IUMA acronym. After several experiments using uncompressed audio files, this archive began delivering on the original low-speed Internet around the world some MPEG Audio files that were compressed using MP2 (Layer II) format and later on the MP3 files used when the standard was fully completed. MP3 popularity began to increase rapidly with the advent of Nullsoft audio player, Winamp, released in 1997. In 1998, the first portable digital MPMan portable audio solid developed by SaeHan Information Systems headquartered in Seoul, South Korea, was released. and the Rio PMP300 was sold thereafter in 1998, despite efforts by the RIAA.
In November 1997, the mp3.com website offered thousands of MP3s created by independent artists for free. The small MP3 file size enables file sharing of peer-to-peer music that is widespread from CDs, which was previously almost impossible. The first peer-to-peer filesharing network, Napster, was launched in 1999. The ease of creating and sharing MP3s resulted in widespread copyright infringement. Big record companies argue that this free music distribution reduces sales, and calls it "music piracy". They reacted by pursuing lawsuits against Napster (which were eventually closed and then sold) and against individual users involved in file sharing.
Unauthorized MP3 file sharing continues on the next generation peer-to-peer network. Some official services, such as Beatport, Bleep, Juno Records, eMusic, Zune Marketplace, Walmart.com, Rhapsody, the recording industry approved Napster's reincarnation, and Amazon.com sold unlimited music in MP3 format.
Maps MP3
Design
File structure
MP3 files consist of MP3 frames, consisting of headers and data blocks. This frame sequence is called elementary flow. Because "reservoir byte", the frame is not an independent item and usually can not be extracted at the boundaries of arbitrary frames. The MP3 Data Block contains audio information (compressed) in terms of frequency and amplitude. The diagram shows that the MP3 Header consists of synchronization words, which are used to identify the beginning of a valid frame. This is followed by a bit indicating that this is an MPEG standard and two bits indicating that layer 3 is used; then MPEG-1 Audio Layer 3 or MP3. After this, the value will be different depending on the MP3 file. ISO/IEC 11172-3 defines the range of values ​​for each header section along with the header specification. Most MP3 files currently contain ID3 metadata, which precedes or follows the MP3 frame, as noted in the diagram. The data stream may contain an optional checksum.
Stereo together is only done on a frame-to-frame basis.
Encoding and decoding
The MPEG-1 standard does not include the exact specifications for the MP3 encoder, but provides examples of psychoacoustic models, loop rates, and the like in non-normative parts of the original standard. MPEG-2 doubles the number of supported sampling levels and MPEG-2.5 adds 3 more. When this is written, the suggested implementation is dated enough. The standard implementer should set up their own matching algorithm to remove portions of the information from the audio input. As a result, many different MP3 makers became available, each producing different quality files. Comparison is widely available, making it easy for potential encoding users to research the best options. Some encoders who are experts in coding with higher bit rates (such as LAME) are not always as good as lower bit rates. Over time, LAME evolved on the SourceForge site to become a de facto CBR MP3 encoder maker. Then the ABR mode is added. Work evolves on the actual variable bit rate using quality objectives between 0 and 10. Finally numbers (such as -V 9600) can produce low quality bit rate sounds of good quality only 41 kbit/dt using MPEG-2.5 extensions.
During encoding, 576 time domain samples were taken and converted into 576 frequency domain samples. If there were any temporary, 192 samples were taken instead of 576. This was done to limit the temporal spread of the quantization noise accompanying the transients. (See psychoacoustics.) Frequency resolution is limited by the size of a small long block window, which lowers the coding efficiency. The time resolution can be too low for very temporary signals and may cause percussion percussion.
Because of the structure of the filter bank tree, the problem of pre-echo becomes worse, because the combined impulse response of the two filter banks is not, and can not, provide an optimal solution in time/frequency resolution. Additionally, a combination of two filter bank outputs creates an alias problem that must be handled partially by the "aliasing compensation" stage; However, it creates excess energy to be encoded in the frequency domain, thereby lowering the coding efficiency.
Decoding, on the other hand, is carefully defined in the standard. Most decoders are "bitstream compliant", meaning that the decompression output they generate from a given MP3 file will be the same, in a certain degree of rounding tolerance, as outputs that are mathematically determined in high ISO/IEC standard documents (ISO/IEC 11172-3 ). Therefore, decoder comparisons are usually based on how efficient their computing is (ie, how much memory or CPU time they use in the decoding process). Over time, these concerns become less of a problem because the CPU speed is diverted from MHz to GHz. The overall encoder/decoder delay is not defined, meaning there is no official provision for gapless playback. However, some encoders like LAME can attach additional metadata that will allow players who can handle it to provide smooth playback.
Quality
When doing a lossy audio encoding, such as creating an MP3 data stream, there is a trade-off between the amount of data generated and the sound quality of the results. The person producing the MP3 selects the bit rate, which determines how many kilobits per second the desired audio. The higher the bit rate, the larger the MP3 data stream, and, generally, the closer it will be heard to the original recording. With too low bit rates, compression artifacts (ie, sounds that are not in the original recording) can be heard in reproduction. Some audio is difficult to compress due to randomness and sharp attacks. When this type of audio is compressed, artifacts such as rings or pre-echoes are usually heard. Examples of applause or triangular instruments with relatively low bit rates provide a good example of compression artifacts. Most subjective tests of perceptual codecs tend to avoid using this type of sound material, however, the artifacts generated by percussion sound are almost invisible because of the special temporal masking feature of the 32 sub-band filterbank of Layer II on which the format is based.
In addition to the bit rate of the encoded audio section, the MP3 encoded sound quality also depends on the quality of the encoder algorithm as well as the complexity of the encoded signal. Because the MP3 standard allows little freedom with encoding algorithms, different encoding makers have very different quality features, even with identical bit rates. For example, in a public listening test featuring two initial MP3 encoders set around 128 kbit/s, one scored 3,66 on a scale of 1-5, while the other scored only 2.22. Quality depends on the choice of encoder and encoding parameters.
This observation led to a revolution in audio encoding. Early bitrate is the primary and only consideration. When MP3 files are the simplest type: they use the same bit rate for all files: this process is known as Constant Bit Rate (CBR) encoding. Using a constant bit rate makes encoding simpler and less CPU-intensive. However, it is also possible to create a file where bit rates are changed across files. This is known as the Bit Rate Variable. The bit reservoir and VBR encoding are actually part of the original MPEG-1 standard. The concept behind them is that, in every part of the audio, some parts are easier to compress, such as silence or music that only contains a few tones, while others will be more difficult to compress. Thus, the overall quality of the files can be increased by using lower bit rates for less complex and higher parts for more complex parts. With some advanced MP3 encoders, it is possible to specify the quality provided, and the encoder will adjust the appropriate bit rate. Users who want a certain "quality setting" that is transparent in their ears can use this value when encoding all their music, and generally do not have to worry about doing a personal hearing test on every piece of music to determine the correct bit rate.
Perceived quality can be affected by ambient noise, listeners attention, and audience training and in many cases by audio equipment listeners (such as sound cards, speakers and headphones). Furthermore, adequate quality can be achieved with lower quality settings for lectures and human speech applications and reduces encryption time and complexity. A test given to new students by Stanford University Music Professor Jonathan Berger shows that students' preference for MP3-quality music has increased every year. Berger says the students seem to prefer the 'hiss' sound that MP3 carries to music.
An in-depth study of MP3 audio quality, voice artist and project composer Ryan Maguire "The Ghost in the MP3" isolates the sounds lost during MP3 compression. In 2015, he released the song "moDernisT" (anagram of "Tom's Diner"), composed exclusively of sounds removed during the MP3 compression of the song "Tom's Diner", a song originally used in MP3 standard formulation. A detailed description of the techniques used to isolate the sounds removed during MP3 compression, along with conceptual motivation for the project, has been published in 2014 Proceedings of the International Computer Music Conference.
Bit rate
Bitrate is the product of the sample rate and the number of bits per sample used to encode the music. Audio CD is 44100 samples per second. The number of bits per sample also depends on the number of audio channels. CDs are stereo and 16 bits per channel. So multiplying 44100 with 32 gives 1411200 - bitrate from uncompressed digital audio CD. MP3 is designed to encode this 1411 kbit/s data at 320 kbit/s or less. Because the less complex parts are detected by the MP3 algorithm the lower bitrate can be used. When using MPEG-2 instead of MPEG-1, MP3s only support lower sampling rates (16000, 22050 or 24000 samples per second) and offer bitrate options as low as 8 kbit/s but not higher than 160 kbit/s. By lowering the sampling rate, the MPEG-2 III layer eliminates all frequencies above half the new sampling rate that may already exist in the audio source.
As shown in these two tables, the selected 14 bit rates are allowed in the MPEG-1 Audio Layer III standard: 32, 40, 48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 256 and 320 kbit/s, along with the 3 highest sampling frequencies available from 32, 44.1 and 48 kHz. MPEG-2 Audio Layer III also allows for 14 bit rates that are somewhat different (and mostly lower) from 8, 16, 24, 32, 40, 48, 56, 64, 80, 96, 112, 128, 144, 160 kbit/sec with a sampling frequency of 16, 22.05 and 24 kHz which is exactly half of the MPEG-1 MPEG-2.5 frame. Audio Layer III is limited to only 8 bits of level 8, 16, 24, 32, 40, 48, 56 and 64Ã, kbit/s with 3 sampling frequencies lower than 8, 11,025, and 12 kHz.
The MPEG-1 frame contains the most detail in 320 kbit/s mode with simple stillness and tone still requires 32 kbit/s. The MPEG-2 frame can capture up to 12 kHz required sound reproduction up to 160 kbit/s. MP3 files created with MPEG-2 do not have 20kHz bandwidth due to the Nyquist-Shannon sampling theorem. The frequency of reproduction is always very less than half the sampling frequency, and the imperfect filter requires a greater margin for error (noise level versus filter sharpness), so the 8 kHz sampling rate limits the maximum frequency to 4 kHz, while sampling the 48 kHz level limits the MP3 to maximum 24 kHz voice reproduction. MPEG-2 uses half and MPEG-2.5 is only a quarter of the sample rate of MPEG-1.
For the general field of human voice reproduction, the 5512 Hz bandwidth is sufficient to produce excellent results (for sound) using a 11025 sampling rate and VBR encoding of 44100 (standard) wave files. This is easily achieved using LAME version 3.99.5 and the command line " lame -V 9.6 lecture.WAV "English speakers average 41-42 kbit/s with a setting of -V 9.6 but this may vary with the number of recorded silences or the delivery rate (wpm). Resampling to 12000 (6K bandwidth) selected by LAME parameter -V 9.4 Similarly -V 9.2 selects 16000 sample rate and 8K generated lowpass filtering. For more info see Nyquist - Shannon. Older versions of LAME and FFmpeg only support integer arguments for variable bit rate quality selection parameters. The n.nnn (-V) quality parameters are documented in lame.sourceforge.net but are only supported in LAME with the new VBR style variable-speed bit rate selector - not the average bit rate (ABR).
The 44.1 kHz sample rate is commonly used for music reproduction, as it is also used for audio CDs, the main source used to create MP3 files. A large number of bit rate is used on the Internet. 128 kbit/s bit rate is usually used, at 11: 1 compression ratio, offers sufficient audio quality in a relatively small space. Since the availability of Internet bandwidth and hard drive size has increased, the higher bit rate of up to 320 kbit/s is widespread. The uncompressed audio stored on the audio CD has a bit rate of 1,411.2 kbit/s, (16 bits/samples ÃÆ'â € "44100 samples/second ÃÆ'â €" 2 channels/1000 bits/kilobit), so the bitrates are 128, 160 and 192Ã, kbit/s represents a compression ratio of about 11: 1, 9: 1 and 7: 1.
Non-standard bit rates up to 640 kbit/s can be achieved with LAME encoder and freeformat options, although some MP3 players can play those files. According to ISO standards, decoders are only required to be able to decode up to 320 kbit/s. The earlier MPEG Layer III encoder uses what is now called the Constant Bit Rate (CBR). This software can only use a uniform bitrate on all frames in an MP3 file. Then a more sophisticated MP3 encoder can use the reservoir bit to target the average bit rate by selecting the encoding rate for each frame based on the sound complexity of the recording.
More sophisticated MP3 encoder manufacturers can produce variable bitrate audio. MPEG audio can use bitrate switching on a per-frame basis, but only a layer III decoder should support it. VBR is used when the goal is to achieve a fixed level of quality. The final file size of the VBR encoding is less predictable than the constant bitrate. The average bitrate is a type of VBR that is implemented as a compromise between the two: allow bitrate varies for more consistent quality, but is controlled to stay near the average value chosen by the user, for a predictable file size. Although the MP3 decoder must support VBR to meet the standards, historically some decoders have bugs with the VBR decode, especially before VBR encoding becomes widespread. The most evolved MP3 LAME maker supports the creation of VBR, ABR, and even the ancient CBR MP3 format.
Audio Layer III can also use "reservoir bits", a partial full frame capability to store parts of the next audio frame data, allowing temporary changes in effective bitrates, even in a constant bitrate stream. Internal handling of the reservoir bits increases the encoding delay. There is no band 21 scale factor (sfb21) for frequencies above about 16 kHz, forcing the encoder to choose between a less accurate representation in band 21 or less efficient storage in all bands under 21 bands, the latter producing bitrate wasted in VBR encoding.
Additional data
Additional data fields can be used to store user-specified data. Additional data is optional and the number of available bits is not explicitly given. Additional data is located after the Huffman bit code and the range to which the main_data_begin next frame points. mp3PRO uses additional data to encode their bits to improve audio quality.
Metadata
"Tags" in an audio file are part of a file containing metadata such as title, artist, album, track number, or other information about the file content. The MP3 standard does not define the tag format for MP3 files, nor is there a standard container format that will support metadata and remove the need for tags. However, some de facto standards for tag formats exist. In 2010, the most widespread was ID3v1 and ID3v2, and APEv2 was recently introduced. These tags are usually embedded in the beginning or end of an MP3 file, separate from the actual MP3 frame data. MP3 decoders either extract information from tags, or just treat them as non-MP3 junk files that can be deleted.
Play & amp; editing software often contains tag editing functions, but there is also a dedicated tag editor app for that purpose. In addition to metadata relating to audio content, tags can also be used for DRM. ReplayGain is a standard for measuring and storing the loudness of MP3 files (audio normalization) in its metadata tags, allowing players compatible with ReplayGain to automatically adjust the overall playback volume for each file. MP3Gain can be used to convert files reversibly based on ReplayGain measurements so that customized playback can be achieved on players without ReplayGain capabilities.
License, ownership and legislation
The patent-free MP3 encoding and encoding technology in the EU, all patents have expired by 2012. In the United States, the technology became patent-free substantially on April 16, 2017 (see below). The majority of MP3 patents expire in the US between 2007 and 2015. In the past, many organizations have claimed ownership of patents related to MP3 decoding or encoding. This claim causes a number of legal threats and actions from various sources. As a result, uncertainty about which patents should be licensed to create MP3 products without infringing patents in countries that allow software patents is a common feature in the early stages of technology adoption.
Complete initial MPEG-1 standards (spare parts 1, 2 and 3) are available to the public on 6 December 1991 as ISO CD 11172. In most countries, patents can not be filed after the previous work is published, and the patent expires 20 years after the date initial submissions, which can be up to 12 months later for submissions in other countries. As a result, the patents required to implement MP3 expires in most countries by December 2012, 21 years after the publication of ISO CD 11172.
The exception is the United States, where the applicable patent but filed before June 8, 1995 ends after 17 years from the date of issue or 20 years from the priority date. Long patent prosecution may result in a patent issuance more slowly than usual (see patent submarine). Various MP3-related patents expire on the start date of 2007-2017 in the United States. A patent for anything disclosed in ISO CD 11172 is filed a year or more after its publication is questioned. If only a known MP3 patent filed by December 1992 is considered, then the MP3 decoding has been patent-free in the US since September 22, 2015, when U.S. Patent 5,812,672 , which had PCT archiving in October 1992, expired. If the longest patent mentioned in the reference is taken as a measure, then MP3 technology becomes patent-free in the United States on April 16, 2017, when U.S. Patent 6,009,399 , held and maintained by Technicolor, expired. As a result, many free and open source software projects, such as the Fedora operating system, have decided to start sending MP3 support by default, and users no longer need to install unofficial packages managed by third party software repositories for MP3 playback or encoding.
Technicolor (formerly Thomson Consumer Electronics) claims to control Layer 3 license licenses in many countries, including the United States, Japan, Canada, and EU countries. Technicolor has been actively enforcing this patent. The MP3 license revenue from the Technicolor administration generated about EUR100 million for the Fraunhofer Society in 2005. In September 1998, the Fraunhofer Institute sent a letter to several MP3 software developers stating that a license was required to "distribute and/or sell decoders and/or encoders ". The letter claims that unlicensed products "infringe Fraunhofer and Thomson patents.In order to create, sell or distribute products using the [MPEG Layer-3] standard and thus our patent, you need to obtain a license under this patent from us." This has led to a situation where the MP3 LAME encoder project can not offer authorized binary users who can run on their computers. The project position is that as source code, LAME is just a description of how MP3 encoder can be implemented. Unofficially, compiled binaries are available from other sources.
Sisvel S.p.A. and its subsidiaries in the United States of Audio MPEG, Inc. previously sued Thomson for patent infringement on MP3 technology, but the dispute was settled in November 2005 with Sisvel giving Thomson licenses for their patents. Motorola followed soon after, and signed a contract with Sisvel to license MP3-related patents in December 2005. Except for three patents, the US patent managed by Sisvel has expired in 2015. The three exceptions are: US. Patent 5,878,080 , expires in February 2017; US. Patent 5,850,456 , expires in February 2017; and AS. Patent 5,960,037 , ending April 9, 2017.
In September 2006, German officials confiscated MP3 players from SanDisk booth at the IFA show in Berlin after the Italian patent company won an order on behalf of Sisvel against SanDisk in a licensing rights dispute. The order was later annulled by a Berlin judge, but the reversal was in turn blocked on the same day by another judge of the same court, "bringing the West Wild Patent to Germany" in the words of a commentator. In February 2007, Texas MP3 Technologies sued Apple, Samsung Electronics and Sandisk in an eastern Texas federal court, claiming a patent infringement of a portable MP3 player that the Texas MP3 says has been commissioned. Apple, Samsung, and Sandisk completed their claims against them in January 2009.
Alcatel-Lucent has confirmed several MP3 encodings and compression patents, allegedly inherited from AT & amp; T-Bell Labs, in its own litigation. In November 2006, before the company's merger, Alcatel sued Microsoft for allegedly violating seven patents. On February 23, 2007, the San Diego jury gave Alcatel-Lucent $ 1.52 billion in damages for a breach of two of them. The court subsequently overturned the verdict, however, found that one patent was not infringed and the other was not owned by Alcatel-Lucent; it's shared by AT & amp; T and Fraunhofer, who had licensed it to Microsoft, the judge decided. The defense ruling was upheld in appeal in 2008. See Alcatel-Lucent v. Microsoft for more information.
Alternate technology
Other lossy formats exist. Among these, MP3PRO, AAC, and MP2 are all members of the same technology family as MP3 and rely on almost identical psychoacoustic models. The Fraunhofer Society has many basic patents underlying this format as well, with others being held by Alcatel-Lucent, and Thomson Consumer Electronics. There are also open compression formats such as Opus and Vorbis available for free and without known patent restrictions. Some newer audio compression formats, such as AAC, WMA Pro and Vorbis, are free of some limitations attached to MP3 formats that can not be solved by MP3 encoder.
In addition to the lossy compression methods, lossless formats are a significant alternative to MP3s because they provide unchanged audio content, albeit with increased file size compared to lossy compression. Lossless formats include FLAC (Free Lossless Audio Codec), Apple Lossless and many others.
See also
- Comparison of audio encoding format
- MP3 Blog
- MP3 player
- Surround MP3
- MP3HD
- MPEG-4 Part 14
- Podcasts
- Portable media player
References
Further reading
- Geert Lovink (2014). "Reflection on MP3 Format: Interview with Jonathan Sterne". Computational Culture (4). ISSN 2047-2390.
External links
- MP3 in Curlie (based on DMOZ)
- MP3-history.com, MP3 Story: How MP3 was created, by Fraunhofer IIS
- News Archive MP3, Over 1000 articles from 1999-2011 focus on MP3 and digital audio
- MPEG.chiariglione.org, MPEG Official Website
- HydrogenAudio Wiki, MP3
- RFC 3119, More RTP Payload Tolerance for MP3 Audio
- RFC 3003, Audio/mpeg Media Type
Source of the article : Wikipedia
0 Comments