Research on iOS audio technology-Audio format

**什么是音频格式**

It was a long time before I could figure out the problem. Audio format actually refers to the type of container, in the popular point is the type of sound files, such as "I Love You china. mp3", the audio format of this sound file is MP3.
Here is a little introduction to some audio coding things. Many of the first involved in this field (for example, I ha), it is easy to confuse the audio format and audio coding differences and connections, such as the audio format has MP3 format, audio encoding has MP3 encoding, this time most people do not understand.
Audio encoding essence is an algorithm, we have to get the original sound data, can not be directly put into the file, we need to use for different purposes for these data processing, such as compression to make its volume smaller, this time to use the audio encoding. Audio coding is an algorithm that people design for a variety of needs.
The data is ready and you need to save the data to a file for a long time. While you are saving sound data, you may want to store some other data, even scripts, for some purposes. As a result, music files became a hybrid. In order for the player to know what the sound file is mixed with, it needs to give it a specific audio format.
So the audio format and audio encoding are different.

**特点**

To play or process audio files in the computer, that is, to the sound file number, mode conversion, the process is also composed of sampling and quantification, the sound that the human ear can hear, the lowest frequency is from 20Hz to the highest frequency 20khz,20khz the ear is not audible, Therefore, the maximum bandwidth for the audio file format is 20KHZ, so the sampling rate needs to be between 40~50khz and more quantization bits per sample. The standard for audio digitization is the signal-to-noise ratio of 16-bit -96db per sample, which uses a linear pulse-coded modulation PCM with equal lengths for each quantization step. In the production of audio files, it is the adoption of this standard.

**分类**

Audio formats can generally be divided into two main categories:

First, OST (Non-compressed)

There are many non-compressed data formats, the most popular is the WAV format. WAV files are flexible in format and can store multiple types of audio data. It is a good choice to save the original recording data. The WAV format is based on the riff file format, and the riff format is similar to the AIFF and IFF formats.
BWF (broadcast sonic format) as a successor to WAV, is a standard audio format created by the European Broadcasting Federation. Metadata can be stored in the BWF file. BWF files are also based on the riff file format, and the extension is WAV. (only mentioned on wiki, but not found)

Second, compression

Compression classes can also be divided into two small categories:
1, non-destructive, such as ape, Flac,tak,tta, Wv,lpac, Au,alac
2, damage, such as Mp3,realaudio,ogg,vqf,wma,atrac, Musepack, AAC,AMR
The lossy file format is based on an acoustic psychology model that removes sounds that are difficult or impossible for humans to hear, such as a sound with a very low volume behind it.

**常见格式**

Acoustic ——————————————————————————————
CD
The audio quality of the CD format is relatively high. So to speak the audio format, CD is naturally the leading pioneer. In the "Open file type" of most playback software, you can see. cda format, this is the CD track. The standard CD format, also known as 44.1K sampling frequency, 88k/second, 16-bit quantization bit, because the CD track can be said to be nearly lossless, so its voice is basically loyal to the original. CDs can be played on CD players, and can be replayed using various playback software on your computer. A CD audio file is a. cda file, which is just an index information, does not really contain sound information, so regardless of the length of the CD music, the "*.CDA file" seen on the computer is 44 bytes long.
WAVE
WAVE (. WAV) is a sound file format developed by Microsoft and IBM that complies with the Piffresource Interchange File Format Files specification, which holds the audio information resources of the Windows platform and is supported by the Windows platform and its applications. ". WAV "format supports MSADPCM, CCITT A Law and other compression algorithms, support a variety of audio bits, sampling frequency and channel, the standard format of WAV file and CD format, is also 44.1K sampling frequency, rate 88k/seconds, 16 bit quantization bit, see it, WAV format sound file quality and CD-similar, is currently popular on the PC voice file format, almost all audio editing software "know" wav format.
AIFF
The AIFF (Audio Interchange File format) format and the AU format, which are very similar to WAV, are supported in most audio editing software, among other common music formats. AIFF is an abbreviation for the Audio Interchange file format. is an audio file format developed by Apple, supported by the MAC platform and its applications, and LiveAudio also supports AIFF format in the Netscape browser. So everyone is not common. AIFF is the standard audio format above Apple Apple computers and is part of QuickTime technology. This format is characterized by the fact that the format itself is irrelevant to the meaning of the data and is therefore favored by Microsoft and is thus developed in WAV format. AIFF is a very good file format, but because it is on the Apple Computer format, so on the PC platform has not been very popular. However, since Apple computers are used in the multimedia publishing industry, almost all audio editing software and playback software support the AIFF format more or less. As long as the Apple computer is still there, AIFF will always have a place. Because of the containment nature of AIFF, it supports many compression techniques.
Lossless Compression ———————————————————————————— –
APE
Ape (Monkey's audio), is a common lossless audio compression encoding format, with an. ape extension and sometimes a. mac extension. When compressing CD audio, a typical monkey ' s audio file tends to be close to 600~700k bit/sec, while MP3 up to 320K bit/sec, typically users will only be assigned to 128~192k Bit/sec.
The ape file structure is defined by monkey ' s audio. Monkey ' s audio provides software for converting to other audio file formats. Monkey ' s audio is the software for compressing/decompressing ape files. It is named after a monkey pattern on its main interface. Monkey's audio is an important tool for compressing ape formats, and can also be decompressed for ape files.
Characteristics:
1, compression rate: Compression ratio is generally around 55%
2, Codec: encoding, decoding speed is slightly slow, low-profile computer will have a lag
3, error handling: No error handling function, if the file is damaged, the data after the damaged location may be lost
4, sound quality: In the sound quality, relative to WMA, MP3, AAC and other lossy data compression format has an absolute advantage
5. Sampling Rate:
6. Resolution:
7. Open Source: Monkey's audio is a free software for open sources, the license agreement is not free software but quasi free software (semi-free software) is marginalized, many based on gnu/ Linux Linux distributions or other operating systems that can only be based on free software cannot earn
8, Other: hardware support
FLAC
FLAC (free Lossless audio Codec), Chinese literal translation is a lossless audio compression encoding (note: "Free" here refers to freedom rather than free). FLAC is a well-known free audio compression coding, which is characterized by lossless compression of audio files. Unlike other lossy compression encodings such as MP3 and WMA (9.0 versions support lossless compression), it does not destroy any of the original audio information, so you can restore the sound quality of your music discs.
Characteristics:
1, compression rate: Compression rate is slightly less than ape
2, Codec: More advanced technology, less resources, decoding speed than ape faster
3. Error handling: Only fixed-point sampling is supported, and floating point sampling is not supported, ensuring that no approximate errors affect sound quality. When the data stream is corrupted, the data loss is limited to the damaged data frame, and typically only a short fragment is lost.
4. Sound quality: Same ape
5. Sample rate: Supports any sampling rate from 1Hz to 655,350hz and can be fine-tuned by 1Hz
6, Bit rate: support any PCM bit resolution, from 4 to 32bit
7. Open Source: Support for most unix-like systems (including Linux,bsd,solaris and Mac OS X), Windows,beos and OS/2
8, Other: hardware support
TAK
TAK (Tom's lossless audio Kompressor) is a music codec with a lossless sound quality. But for now, there are few supported software, just like foobar2000 and Winamp and other well-known music playback software to play with the plug-in.
Characteristics:
1. Compression Ratio: High compression ratio approximate to ape
2, Codec: Close to FLAC encoding, decoding speed, support multi-threaded encoding (1.0.3 later version)
3, error handling: With error tolerance (single bit error will not affect more than 250ms), with error detection (each frame has a 24-bit CRC)
4. Sound quality: Same ape
5. Sample rate: Support up to 192khz
6, Bit rate: support up to 24bit
7, open Source: Not yet open sources (in the future to book in C + + Open source code). Although the original author has not yet disclosed the source code, the FFmpeg developer has implemented an open source Tak decoder through reverse engineering and has entered into the FFmpeg
8, Other: Support APEv2 tag, support streaming media
TTA
TTA (True audio) is a free and simple real-time lossless audio codec. The TTA is a lossless audio compression based on adaptive predictive filtering, which can have the same or better compression effect compared to other current formats.
Characteristics:
1. Compression Ratio: data can be compressed to 30%-70%
2, Codec: real-time encoding, decoding algorithm, fast operation, low system requirements
3. Error Handling:
4. Sound quality: Same ape
5. Sampling Rate:
6, Bit rate: 8bit, 16bit, 24bit Integer and 32bitIEEE floating-point WAV format audio file
7, open Source: Support for multi-platform free software and open sources of code
8, Other: hardware support, support ID3V1 and ID3v2 two kinds of label information
WV
WV (WavPack) is a free, open source lossless audio compression format developed by David Bryant with a file suffix of. WV.
WavPack introduces a unique "hybrid" mode that uses an additional file to also have the advantage of lossy compression. Unlike other methods that generate only one file, this pattern generates two files, one of which is a relatively small, high-quality lossy compressed file that can be used separately, and the other is a "remediation" file for lossless data recovery with lossy files. For some users, this means that they no longer have to consider the use of lossy or lossless compression.
Characteristics:
1, compression rate: For ordinary pop music, usually between 30%-70%, for classical music and other wide range of music, usually can get a higher proportion
2, Codec: Fast and efficient compression and decompression
3, error Handling: The robustness of error
4. Sound quality: Same ape
5. Sample rate: Supports very high sample rates
6, Bit rate: 8bit, 16bit, 24bit, 32bit Integer, and 32bit floating-point representationWAV format audio file
7. Open Source: Open sources, published in a similar way to BSD license
8, Other: hardware support, support streaming media, support ID3v1, APEV2 label
MPEG-4 ALS
MPEG-4 ALS (lpac,lossless predictive audio Compression), also known as Audio Lossless encoding, is a lossless audio data compression method.
It is an extension of the MPEG-4 audio standard, which was finalized in December 2005.
MPEG4 ALS is similar in arithmetic to FLAC, which is simply a quantized linear predictive coding predictor using Golomb coding or bounded Gilbert Moore coding to encode the remainder, possibly due to the lack of available encoders and decoders, to 20 This format was still not accepted by the public in 06.
Characteristics:
1. Compression Ratio:
2, Codec:
3. Error Handling:
4. Sound quality:
5. Sampling Rate:
6, Bit rate:
7, open-source nature:
8. Other:
AU
Audio file is a digital audio format introduced by Sun Corporation. The AU file was originally a digital sound file under the UNIX operating system. Because the Web servers on the Internet were primarily Unix-based, the files in AU format were also commonly used in sound file formats in today's Internet.
Characteristics:
1. Compression Ratio:
2, Codec:
3. Error Handling:
4. Sound quality:
5. Sampling Rate:
6, Bit rate:
7, open-source nature:
8. Other:
ALAC
The ALAC (Apple Lossless audio Codec) is an Apple lossless audio compression encoding format. Also because it is lossless compression, it sounds exactly the same as the original file, and will not be changed by decompression and compression. The main difference between the ALAC and the MP3 is that during the coding process, the MP3 cancels the audio data from a small portion of the high-frequency and low-frequency portions, while the ALAC records truthfully and does not delete any detail data from the audio.
It was published on April 28, 2004 in one of the iTunes4.5 and QuickTime6.5.1 parts.
Characteristics:
1, Compression ratio: compression to the original capacity of 40%-60%, higher than MP3
2, Codec: encoding, decoding speed quickly
3. Error Handling:
4. Sound quality: Same ape
5. Sampling Rate:
6, Bit rate: non-compressed audio format (WAV, AIFF)
7, open Source: The ALAC encoder was published on October 26, 2011 in Apache license for the agreement to publish the source code
8, Other: currently portable digital multimedia player only ipod can play
lossy compression ———————————————————————————— –
MPEG
MPEG is the English abbreviation for the dynamic Image Expert group. This expert group was founded in 1988 and is dedicated to creating video and audio compression standards for CDs. MPEG audio files refer to the sound portion of the MPEG standard, which is the MPEG audio layer. The current music format on the internet is most common with MP3. Although it is a lossy compression, its greatest advantage is that it has a high compression ratio with minimal sound distortion. MPEG contains formats including: MPEG-1, MPEG-2, Mpeg-layer3, MPEG-4
MP3
The MP3 (MPEG3) format was born in Germany in the 80 's, so-called MP3 refers to the audio portion of the MPEG standard, which is the MPEG audio layer. According to the different compression quality and coding process is divided into 3 layers, respectively, corresponding to the ". Mp1 ",". MP2 ",". mp3 "These 3 kinds of sound files. MPEG Audio file compression is a lossy compression, while basically keeping the low audio portion of the distortion, but at the expense of the sound file 12KHz to 16KHz High Audio this part of the quality in exchange for file size. Because of its small file size and good sound quality, there is no other audio format to rival it at the beginning of its existence, thusThe development of the. mp3 format provides good conditions.
Characteristics:
1, compression rate: compression to the original capacity of the 8.5%-10%
2, Codec: encoding, decoding speed quickly
3. Error Handling:
4. Sound quality: The higher the sampling rate, the better the sound quality.
5. Sampling rate: up to 48kHz
6, Bit rate: variable encoding algorithm will be the interval value
7, open Source: Lame the perfect realization of the VBR (variable encoding rate) algorithm, and it is completely free software, and the development team composed of enthusiasts have been constantly developing and perfecting. On the basis of VBR, lame more developed ABR algorithm. The average bit rate of ABR (averagebitrate) is an interpolation parameter of VBR.
8, Other: Support ID3 tag, support streaming media,
RA
RA (RealAudio) has many versions, RealAudio 1, RealAudio 2, RealAudio 3 have recently reached RealAudio 11, these formats are different, depending on the media player is also different, the same player some RM files can be played out, Some do not, this time can see is not RealAudio encoded version of the problem.
RealAudio is mainly applicable to online music appreciation on the Internet. Most users are still using a modem with a lower rate, so a typical playback is not the best sound. Some download sites will prompt you to choose the best real file based on your modem rate. There are several main file formats for real: RA (RealAudio), RM (Realmedia,realaudio G2), RMX (RealAudio Secured), and much more. These formats are characterized by varying the quality of the sound depending on the bandwidth of the network, making it easier for the most affluent listeners to get better sound when most people hear a smooth sound.
1. Compression Ratio:
2, Codec:
3. Error Handling:
4. Sound quality:
5. Sampling Rate:
6, Bit rate:
7, open-source nature:
8. Other:
OGG
OGG (Ogg Vorbis) is a new audio compression format, similar to the existing music formats such as MP3. But a little different is that it is completely free, open and without patent restrictions. The Oggvorbis file has the extension *. OGG. The design format of this file is very advanced. This file format can be continuously improved in size and sound quality without affecting the old encoder or player. Vorbis uses lossy compression, but reduces losses by using more advanced acoustic models.
The latest version is the Libogg 1.2.0 released on March 26, 2010. [2] Another version of LIBOGG2 can also be found in the Xiph.org Foundation's SVN package library.
1, compression rate: can be continuously improved
2, Codec: can be continuously improved
3. Error Handling:
4. Sound quality: The same bit rate coded ogg is better than MP3 sounds.
5. Sampling Rate:
6, Bit rate:
7. Open Source: Free software released under the new BSD license
8. Other:
VQF
Yamaha development, its core is to reduce data flow but to maintain the sound quality method to achieve a higher compression ratio, VQF audio compression rate than the standard MPEG audio compression rate is nearly one-fold, can reach about 18:1 or even higher. It can be said that the technology is also very advanced, but due to poor publicity, this format is difficult to play.the. VQF can be played with the Yamaha player. At the same time Yamaha also offersConvert. wav files to the software for the *.VQF file. This document lacks features plus a lack of publicity.
After the SOUNDVQ compressed audio file in the playback effect audition, almost no one can hear it and the original audio file differences. Playback VQF requires a computer to be configured for only Pentium 75 or higher, of course, if you use a Pentium 100 or above machine, VQF can run better. In fact, the CPU requirement for playback VQF is only about 5~10% higher than Mp3. VQF-TWINVQ technology, although developed by NTT and Yamaha, is free of charge for their application software.
1, compression ratio: about 5%, compression ratio is greater than MP3 and RA
2, Codec:
3. Error Handling:
4. Sound quality: Close to CD quality (16-bit 44.1kHz stereo)
5. Sampling Rate:
6, Bit rate:
7, open Source: NTT and Yamaha did not disclose the source code of VQF
8. Other:
WMA
Microsoft Development of WMA (Windows Media Audio), like the VQF format developed by the Japanese Yamaha company, is designed to achieve a higher compression rate than MP3 by reducing data traffic but maintaining sound quality. Some pure audio ASF files that encode all of their content using the Windows Media Audio encoding format also use WMA as the extension.
Microsoft has dramatically improved its engine in WMA 9, and in fact 64Kbps of WMA music can achieve the same quality as the 128Kbps MP3 music, about 1/3 less than the MP3 volume.
Another advantage of WMA is that content providers can add anti-copy protection through DRM (Digital rights Management) scenarios such as Windows Media rights Manager 7. This built-in copyright protection technology can limit the playing time and the number of plays and even play the machine and so on, this is a pirated mess of the music company is a boon.
Characteristics:
1, the compression rate: about 5% (only in the case of MP3 below the 192KBPS bit rate)
2, Codec:
3. Error Handling:
4, sound quality: Better than MP3 format, more than the RA format
5. Sampling Rate:
6, Bit rate:
7, open-source nature:
8, Other: DRM copyright protection, support streaming media, ID3 label, WMA9 version start to support lossless compression (Windows Media Audio 9 Lossless)
ATRAC
ATRAC (Adaptive Transform acoustic Coding), adaptive auditory conversion coding for the Sony Company in 1992 developed the audio lossy data compression technology, is also a generic term for related technical terms. In addition to Sony itself, other MiniDisc manufacturers, such as sharp and Panasonic, have developed their own atrac codecs.
The Sony Research and Development ATRAC first edition (to avoid confusion, called ATRAC1) has continued to develop the associated lossy compression technology ATRAC2, ATRAC3, ATRAC3plus, and lossless Atrac advanced Lossless. In fact, these five kinds of compression techniques are different from each other, except for the similar names. In addition, the numbers at the end of the ATRAC2 and ATRAC3 names are often misinterpreted as ATRAC version numbers, which are in fact part of the coding technology name.
ATRAC1, usually written as Atrac. To reduce the computational processing burden, the ATRAC1 encoding uses two times QMF (quadrature Mirror Filters) to divide the input audio into three sub-bands , the first separation of the high frequency (11.025~22.05khz), the second separation of the remaining mid-low frequency (0~5.5125khz, 5.5125~11.025khz). The sub-band is then MDCT (Modified discrete cosine Transform, variable address discrete cosine transform) to cut the block, and according to the sensitivity of the ear to the audio and adjust the allocation of the data block, is also called self-adaptation. Compression, Atrac according to auditory psychology, ignoring the hearing of the human ear, as well as the large volume shielding small sound, in order to achieve the purpose of data compression. ATRAC1 does not have the details of how the stipulate is distributed, allowing fine tuning to improve sound quality later.
1. Compression Ratio:
2, Codec:
3. Error Handling:
4. Sound quality:
5. Sampling Rate:
6, Bit rate:
7, open-source nature:
8. Other:
AAL
The AAL (ATRAC Advanced Lossless), published in the a&vfesta2005 in September 2005, is the only lossless compression specification in the ATRAC family, referred to as the AAL. The format can contain no distortion compression, destructive compression two parts. The destruction compression part may use the ATRAC3, the ATRAC3plus and so on the format, but the non-distortion part is the original audio to carry on the reversible lossless compression. In addition to transmitting the entire AAL file to the Walkman, it is possible to remove only the smaller atrac3/atrac3plus part. AAL compression rate is about 30~80%, because the AAL also contains destructive compression of the audio, so the format used to destroy the compression of the AAL will also affect the amount of compression. The AAL was first supported by the SonicStage version 3.3 released on November 1, 2005. Complete AAL data can be transmitted to a music player that fully supports the AAL, or it can only transmit atrac3/atrac3plus parts. On the hardware side, the NW-S700F and nw-s600 released by Sony to October 2006 fully support the AAL.
1. Compression Ratio:
2, Codec:
3. Error Handling:
4. Sound quality:
5. Sampling Rate:
6, Bit rate:
7, open-source nature:
8. Other:
Musepack
Musepack (earlier known as Mpegplus, mpeg+, or mp+) is a lossy compressed audio format based on the MP2 algorithm. It is encoded in a way that focuses on auditory penetration and is particularly good at 160kbit/s or above. Originally presented and developed by Andree Buschmann, Musepack was later taken over by Frank Klemm and is now maintained by the Klemm development team (Musepack Musepack Development) with the help of Frank Team,mdt.
Characteristics:
1. Compression Ratio:
2, Codec: Compared to MP3, AAC more efficient Huffman coding
3. Error Handling:
4. Sound quality:
5. Sampling Rate:
6, Bit rate: 3kbit/s to 1300kbit/s of the pure variable Code rate code
7, open Source: On the Microsoft Windows, Linux and Mac OS X and other platforms, on the official website of Musepack, in addition to the Musepack encoder and decoder, there are several media players dedicated third-party plug-ins, are issued by LGPL or BSD license
8, Other: Noise replacement technology, APEV2 label
AAC
AAC (Advanced audio Coding), appeared in 1997, based on MPEG-2 's audio coding technology. Co-developed by Fraunhofer IIS, Dolby Labs, T, Sony and other companies to replace the MP3 format. After the advent of the MPEG-4 standard in 2000, AAC re-integrates its features, adding SBR technology and PS technology to differentiate it from the traditional MPEG-2 AAC, also known as MPEG-4 AAC.
But until 2006, there was not much to store music in this format, and MP3 players that could play that format were few and far between. In addition, many music playback software on the computer supports AAC (provided the AAC decoder is installed), such as Apple itunes. But in the Mobile phone field, AAC's support has been very common, Nokia, Sony Ericsson, Motorola and other brands in the high-end products in the support of AAC (initially mainly LC-AAC, with the development of mobile phone performance, HE-AAC support has been extensive).
Characteristics:
1, compression ratio: about 5%
2, Codec:
3. Error Handling:
4. Sound quality: Better than almost all traditional coding methods in the same specifications
5. Sampling rate: up to 96kHz
6, Bit rate: 8bit, 16bit, 24bit, 32bit
7, open-source nature:
8. Other:
AMR
AMR Full Name Adaptive multi-rate, Adaptive multi-rate encoding, mainly used for mobile device audio (mobile phone call), compression ratio is larger, but relative to other compression format quality is poor, because more for the voice, call, the effect is very good.
Category 1. AMR: Also known as AMR-NB, compared to the following WB, the Voice bandwidth range: 300-3400hz,8khz sampling
Category 2. Amr-wb:amr Wideband, Voice bandwidth range: 50-7000hz 16KHz sampling
The AMR-WB sampling frequency is 16kHz, which is a wideband speech coding standard adopted by the ISO-T and 3GPP, also known as the G722.2 standard. The AMR-WB provides a voice bandwidth range of 50~7000hz, and the user can subjectively feel that the voice is more natural, comfortable and easy to distinguish than before. Compared with this, the GSM EFR (enhenced full rates, enhanced total rate encoding) sampling frequency is 8kHz and the voice bandwidth is 200~3400hz. The advantage of AMR-WB for narrowband GSM (full-speed channel 16k,gmsk) is that it can be encoded from 6.6kb/s, 8.85kb/s, and 12.65kb/s three, and when the network is busy, the encoder can automatically adjust the encoding mode to enhance QoS. In this application, the AMR-WB immunity is better than AMR-NB. The AMR-WB is applied to edge and 3G to fully demonstrate its advantages. Sufficient transmission bandwidth guarantee AMR-WB can be used from 6.6kb/s to 23.85kb/s a total of nine kinds of encoding, voice quality beyond the PSTN fixed telephone.

AMR is a patented product

The basic audio format has been introduced here. I always thought MIDI could not be classified as an audio format, so I gave him a separate category
MIDI
MIDI (musical Instrument digital Interface) format is used by people who often play music, and MIDI allows digital synthesizers and other devices to exchange data. The mid file format is inherited by MIDI. The mid file is not a recorded sound, but rather a set of instructions that record the sound and then tell the sound card how to reproduce the music. Such a MIDI file only uses about 5~10KB for every 1 minutes of music. Mid files are mainly used for original instrument works, amateur performances of pop songs, game tracks, and electronic greeting cards. The effect of the . Mid file replay depends entirely on the sound card's grade. the biggest use of the. mid format is in the field of computer composition. The . mid file can be written using the composer software, or it can be made into a. mid file by inputting the music of the external sequencer into the computer through the MIDI port of the sound card .

Research on iOS audio technology-Audio format