how does mp3 encoding work Archives

Mp3 (an audio encoding method) Part 3

Free Download Mp4Gain

Mp3 (an audio encoding method) Part 3

MP3 ENCODING

To generate bit-compliant (Layer 1.Layer 2.Layer 3) MPEGAudio files, ISO MPEG Audio committee members developed reference simulation software in C called ISO 11172-5.

MP3 ENCODING

It can demonstrate the first real-time DSP-based hardware decoding of compressed audio on some non-real-time operating systems. Various other MPEG audio was developed in real time for digital broadcasting (DAB radio and DVB TV) for consumer receivers and set-top boxes.
Later on July 7, 1994, Fraunhofer-Gesellschaft released the first MP3 encoder called l3enc.
The Fraunhofer development team selected the .mp3 extension on July 14, 1995 (previously the extension was .bit). Using Winplay3 (released September 9, 1995), the first real-time software MP3 player, many people were able to encode and play MP3 files on their own personal computers. Since hard drives at the time were relatively small (such as 500MB), this technology was essential for storing entertainment music on computers.
MP2, MP3 and Internet
In October 1993, MP2 (MPEG-1 Audio Layer 2) files appeared on the Internet and were often played by Xing MPEG Audio Player and later MAPlay developed by Tobias Bading for Unix. MAPplay was first released on February 22, 1994 and ported to the Microsoft Windows platform.
The only MP2 encoder products at first were Xing Encoder and CDDA2WAV, a CD ripper that converts audio tracks from CDs to WAV format.
Often considered the father of the online music revolution, the Internet Underground Music Archive (IUMA) was the first hi-fi music site on the Internet, with thousands of licensed MP2 recordings before MP3 and the web became popular. .
From the first half of 1995 to the end of the 1990s, MP3 began to flourish on the Internet. MP3’s popularity is largely due to the success of companies and software packages such as Winamp released by Nullsoft in 1997 and Napster released by Napster in 1999, and they are mutually reinforcing. These programs make it easy for normal users to play, create, share and collect MP3 files.
The debate about sharing MP3 files between peers has spread rapidly in recent years, mainly because compression makes file sharing possible, uncompressed files are too large to share. Since MP3 files are widely spread over the Internet, Napster has been sued by some of the major record labels to protect their copyright (see Copyright).
Commercial online music distribution services, such as the iTunes Music Store, often choose other proprietary or DRM-enabled music file formats to control and limit the use of digital music. Formats that support DRM are used to protect copyrighted material from copyright infringement, but most protection mechanisms can be broken in some way. Computer experts can use these methods to generate unlocked files that can be freely copied. One notable exception is Microsoft’s Windows Media Audio 10 format, which has yet to be cracked. If a compressed audio file is desired, the recorded audio stream must be compressed and the sound quality will be degraded.
streaming audio quality
Because MP3 is a lossy compression format, it offers a variety of options for different “bit rates,” that is, the number of encoded data bits needed to represent the audio per second. Typical speeds are between 128 kbps and 320 kbps (kbit/s). In contrast, the uncompressed audio bitrate on a CD is 1411.2 kbps (16 bits/sample × 44100 samples/sec × 2 channels).
MP3 files encoded with lower bit rates generally play at a lower quality. If you use too low a bitrate, “compression artifact” (sounds not present in the original recording) will appear during playback. A good example of compression noise is the sound of compressed cheering; due to its randomness and sharp changes, encoder errors are more pronounced and sound like echoes.

Free Download Mp4Gain

Mp4Gain Main Window

Mp4Gain Features

Free Download Mp4Gain

Mp3 (an audio encoding method) Part 2

mp3 3ncoding

MPEG-1 Audio Layer 2 encoding began as a digital audio broadcast (DAB) managed by Egon Meier-Engelen at the German Deutsche Forschungs- und Versuchsanstalt für Luft- und Raumfahrt (later known as Deutsches Zentrum für Luft- und Raumfahrt, German Space Center). )draft.

mp3 encoding

This project is funded by the European Union as a EUREKA research project, and its name is commonly known as EU-147. The study period for EU-147 was from 1987 to 1994.
2. By 1991, two proposals had emerged: Musicam (called Layer 2) and ASPEC (Adaptive Spectrum Sensing Entropy Coding). The Musicam method proposed by Philips of the Netherlands, CCETT of France, and the Institut für Rundfunktechnik of Germany was chosen due to its simplicity, error robustness, and lower computational effort in high-quality compression. The Musicam format based on subband coding is a key factor in determining the MPEG audio compression format (sample rate, frame structure, header, sample points per frame). This technology and its design philosophy are fully integrated into the definition of ISO MPEG Audio Layer I, II and later Layer III (MP3) formats. The standard was developed by Leon van de Kerkhof (Layer I) and Gerhard Stoll (Layer II) under the auspices of Prof. Mussmann (University of Hannover).
3. A working group consisting of Leon Van de Kerkhof from the Netherlands, Gerhard Stoll from Germany, Yves-François Dehery from France and Karlheinz Brandenburg from Germany absorbed design ideas from Musicam and ASPEC and added their own design ideas to develop an MP3. MP3 can achieve MP2 sound quality from 192 kbit/s to 128 kbit/s.
4. All of these algorithms eventually became part of the first group of MPEG standards, MPEG-1, in 1992, resulting in the international standard ISO/IEC 11172-3 published in 1993. Further work on MPEG audio was eventually became part of the MPEG-2 standard, a second group of MPEG standards developed in 1994, officially known as ISO/IEC 13818-3, first published in 1995.
5. The compression efficiency of the encoder is generally defined by the bit rate, because the compression rate depends on the number of bits (: in: bit depth) and the sampling rate of the input signal. However, there are often products that use CD parameters (44.1 kHz, two channels, 16 bits per channel, or 2×16 bits) as the compression ratio reference, and the compression ratio using this reference is usually higher, which which also shows that the compression ratio is very important for lossy compression problems.
6. Karlheinz Brandenburg used Suzanne Vega’s song Tom’s Diner on CD to test MP3 compression algorithms. This song is used because the song’s smooth and simple melody makes it easier to hear glitches in the compressed format during playback. Some jokingly refer to Suzanne Vega as “the mother of MP3”. Some more serious and critical audio extracts (glockenspiel, triangle, accordion…) from the EBU V3/SQAM reference CD are used by professional audio engineers to assess the subjective perceived quality of the MPEG audio format.

Mp3 (an audio encoding method)

Mp3 encxoding

MP3 is an audio compression technology, its full name is Moving Picture Experts Group Audio Layer III, called MP3.

mp3 encoding

It is designed to drastically reduce the amount of audio data. Using MPEG Audio Layer 3 technology, music is compressed into a smaller capacity file with a compression ratio of 1:10 or even 1:12, and for most users, playback quality is not as good as the original uncompressed. audio Significant decrease. It was invented and standardized in 1991 by a group of engineers at the Fraunhofer-Gesellschaft research organization in Erlangen, Germany. Music stored in the form of MP3 is called MP3 music, and a machine that can play MP3 music is called an MP3 player.

Motion Picture Expert Compression Standard Audio Layer 3 foreign name Moving Picture Expert Group Audio Layer III research organization Fraunhofer-Gesellschaft type audio coding advantage Drastically reduce the amount of audio data defect sound quality loss
content
1 Features
2 story
▪ origin
▪ go to the masses
3 audio quality
4 patent issues
transmission characteristics
MP3 converts the time-domain waveform signal to a frequency-domain signal by taking advantage of the human ear’s insensitivity to high-frequency sound signals and splits it into multiple frequency bands, using different compression rates. for different frequency bands and increasing the compression ratio for high frequencies (even ignoring the signal) Use a small compression ratio for low frequency signals to ensure that the signal is not distorted. In this way, it is equivalent to discarding the high-frequency sound that is basically inaudible to the human ear [1], keeping only the audible low-frequency part, thus compressing the sound with a compression ratio of 1:10 or even 1: 12. Because the full name of this compression method is called MPEG Audio Player3, people call it MP3 for short.
According to the MPEG specification, AAC (Advanced Audio Coding) in MPEG-4 will be the next generation of the MP3 format.
Compared to CD, FLAC and APE lossless compression formats, the sound quality of the highest parameter MP3 (320 Kbps) is not much different.
MP3 players are dying
When they first came out, MP3 players were at the forefront of the digital revolution. However, sales of iPods and other MP3 players in the UK fell sharply in 2012 as consumers turned to other digital products such as smartphones.
In 2012, sales of MP3 players in the UK market were £110m ($178m), just 29% of the £381m in 2011, according to market research firm Mintel. Mintel expects total MP3 player sales in the UK market to halve by 2017. In the worst case scenario, total MP3 player sales in the UK market will be just 25 million dollars five years later. [23]
1. MP3 is a data compression format;
2. Discards pulse code modulation (PCM) audio data that is not important to the human ear (similar to JPEG is a lossy image compression), resulting in a much smaller file size;
3. MP3 audio can be compressed according to different bit rates, providing a variety of trade-offs between data size and sound quality. The MP3 format uses a mixed conversion mechanism to convert audio domain signals. time in frequency domain signals;
4. 32 band polyphase integral filter (PQF);
Modified discrete cosine filter (MDCT) of 5, 36 or 12 taps; each subband size can be independently selected between 0…1 and 2…31;
6. MP3 not only has extensive client software support, but also has a lot of hardware support, such as portable media players (referring to MP3 players), DVD and CD players, outgoing calls

Audio Coding Part 4

Audio Coding

Transmission encoding format

Audio Coding

PCM encoding
PCM Pulse Code Modulation is short for Pulse Code Modulation. In the text above, we mentioned the general PCM workflow. We don’t need to care which calculation method is used in the final PCM encoding. We just need to know the advantages and disadvantages of the PCM encoded audio stream. The biggest advantage of PCM encoding is good sound quality and the biggest disadvantage is its large size. Our common audio CD uses PCM encoding, and the capacity of one CD can only hold 72 minutes of music information.
WAV format
This is an old audio file format, developed by Microsoft. WAV is a file format that complies with the RIFF (Resource Interchange File Format) specification. All WAVs have a file header that contains encoding parameters for the audio stream. WAV does not have strict rules for encoding audio streams. In addition to PCM, almost all encodings that support the ACM specification can encode WAV audio streams. Many friends do not have this concept. Let’s take AVI as an example, because AVI and WAV are very similar in file structure, but AVI has one more video stream. There are many types of AVIs we have come into contact with, so we often need to install some decoders to watch some AVIs. DivX, which we have come into contact with a lot, is a type of video encoding. AVI can use DivX encoding to compress video streams, of course we can also use other code compression. Similarly, WAV can also use a variety of audio codecs to compress its audio stream, but we commonly use WAV whose audio stream is processed by PCM encoding, but this does not mean that WAV can only use PCM codec, it is also you can use MP3 codec. in WAV Just like AVI, as long as the corresponding Decode is installed, you can enjoy these WAVs.
On the Windows platform, WAV based on PCM encoding is the best supported audio format. All audio programs can support it perfectly. Because it can meet higher sound quality requirements, WAV is also the preferred format for music creation and editing. Suitable for storing musical material. Therefore, WAV based on PCM encoding is used as an intermediate format, and is often used in the mutual conversion of other encodings, such as MP3 to WMA.
MP3 encoding
As the most popular audio compression format, MP3 is widely accepted by everyone. Various MP3-related software products emerge in a never-ending stream, and more hardware products start to support MP3 as well. Many VCD/DVD players that we can buy are compatible with MP3. , and there are more portable MP3 players, etc. Although several of the major music companies are extremely displeased with this open format, they cannot prevent the survival and spread of this compressed audio format. MP3 has been in development for 10 years and is short for MPEG (MPEG: Moving Picture Experts Group) Audio Layer-3, which is an encoding scheme derived from MPEG1. MP3 can achieve an incredible 12:1 compression ratio and still maintain basically audible sound quality. In the days when the hard drive was expensive, users quickly accepted MP3. With the popularity of the Internet, hundreds of millions of users accepted MP3. users At the beginning of the release of MP3 encoding technology, it was actually very imperfect. Due to a lack of research on sound and human hearing, almost all early mp3 encoders were crudely encoded and the sound quality was severely damaged. With the continuous introduction of new technologies, mp3 encoding technology has been improved over and over again, including two major technical improvements.

Audio Coding Part 3

Audio Coding

flow characteristics

With the development of the Internet, people have put forward requirements to listen to music online, so it is also required that the audio file can be played while reading, without the need to read the entire file and then play it, so that listening is can achieve without downloading. It can also be done while encoding and playing. It is this feature that you can live broadcast online and set up your own digital radio station has become a reality.
Transmission classification code
According to different coding methods, audio coding techniques are divided into three types: waveform coding, parametric coding, and hybrid coding. Generally speaking, waveform coding has high voice quality, but the coding rate is also high; parametric coding has a very low coding rate and the quality of the resulting synthesized speech is not high; hybrid coding uses parametric coding technology and waveform coding technology, coding rate and sound quality among them.
1. Waveform coding
Waveform coding refers to directly transforming the time-domain signal into a digital code without using any parameters of the generated audio signal, so that the reconstructed speech waveform is as consistent as possible with the waveform. waveform of the original speech signal. The basic principle of waveform coding is to sample the analog speech signal at a certain rate on the time axis and then quantize the amplitude samples hierarchically and represent them with codes.
The waveform coding method is simple, easy to implement, strong in adaptability, and good in voice quality. However, because the compression method is simple, it also has some problems: the compression ratio is relatively low, resulting in a higher encoding rate. Generally speaking, the complexity of waveform coding is relatively low and the coding rate is relatively high. Generally, the audio quality is quite high when the encoding rate is higher than 16 kbit/s. When the coding rate is less than 16 kbit/s, the sound quality is drastically reduced.
The simplest waveform coding method is PCM (Pulse Code Modulation), which just samples and quantizes the speech signal. The advantages are that the coding method is simple, the delay time is short, the sound quality is high, and the reconstructed speech signal is almost indistinguishable from the original speech signal. The disadvantage is that the coding rate is relatively high (64 kbit/s) and it is more sensitive to errors in the transmission channel.
2. Parameter coding
Parametric coding consists of extracting the parameters of the generated speech from the speech waveform signal and using these parameters to reconstruct the speech through the speech generation model, so that the reconstructed speech signal can maintain the semantics of the original speech signal as much as possible. . That is, the parameter encoding is based on the digital model generated by the voice signal, and then the model parameters are obtained from the digital model, and then the digital model is restored according to these parameters, and then the talks.
The coding rate of parametric coding is low, which can reach 2.4 kbit/s. The generated speech signal is restored using the established digital model. Therefore, the waveform of the reconstructed speech signal may be quite different from the waveform of the original speech signal. The distortion will be larger AND due to the limitations of the speech generation model, increasing the data rate does not improve the quality of the synthesized speech. However, although the sound quality of the parameter encoding is relatively low, the confidentiality is very good, and it has been used in the military. A typical parameter coding method is LPC (Linear Predictive Coding).
3. Mixed coding
Hybrid encryption refers to the simultaneous use of two or more encryption methods for encryption. This coding method overcomes the weakness of waveform coding and parametric coding, and combines the high quality of waveform coding and the low coding rate of parametric coding, and can achieve better results.

Audio Coding Part 2

Audio Coding

Reasons to use audio compression technology.

audio coding

It is very easy to calculate the bit rate of a PCM audio stream, the value of the sample rate × the value of the sample size × the number of bps of the channel. A WAV file with a sample rate of 44.1 KHz, a sample size of 16 bits, and two-channel PCM encoding has a data rate of 44.1 K×16×2 = 1411.2 Kbps. We usually say that 128K MP3, the corresponding WAV parameter, is this 1411.2 Kbps, this parameter is also called data bandwidth, it is a concept with the bandwidth in ADSL. Divide the bit rate by 8 to get the data rate for this WAV, which is 176.4 KB/s. This means storing a 1-second sample rate of 44.1 KHz, a 16-bit sample size, and a two-channel PCM-encoded audio signal, which requires 176.4 KB of space, which is approximately 10.34 M in 1 minute, which is unacceptable. For most users, especially friends who like to listen to music on the computer, to reduce disk usage, there are only 2 ways to downsample or compress. Lowering the index is not advisable, so experts have developed various compression schemes. Due to different uses and target markets, the sound quality and compression ratio achieved by various audio compression encodings are different, and we will mention them one by one in the following articles. One thing is for sure, they are all compressed.
Frequency vs. Sampling Rate
The sample rate represents the number of times the original signal is sampled per second. The sample rate of most of the audio files that we see regularly is 44.1 KHz. What does this mean? Suppose we have 2 segments of sine wave signals, 20 Hz and 20 KHz respectively, each lasting one second, to correspond to the lowest and highest frequencies we can hear, and we sample these two signals at 40 KHz respectively. , we can get what kind of result? The result is: the 20 Hz signal is sampled 40K/20=2000 times per vibration, while the 20K signal is only sampled 2 times per vibration. Obviously, under the same sample rate, the low-frequency information is much more detailed than the high-frequency information. This is also the reason why some audiophiles accuse CDs of digital sound not being real enough, and 44.1KHz CD sampling cannot guarantee that high-frequency signals are recorded well. To better record high-frequency signals, a higher sample rate seems to be required, so some folks use a 48KHz sample rate when capturing audio tracks from CDs, which is undesirable! Actually, this is not good for sound quality. For the ripping software, keeping the same sample rate as the 44.1 KHz provided by the CD is one of the guarantees for the best sound quality, rather than improving it. A higher sample rate is only useful for analog signals, if the signal being sampled is digital, do not try to increase the sample rate.

Audio Coding

Sampling rate and sample size
Sound is actually a type of energy wave, so it also has the characteristics of frequency and amplitude, with frequency corresponding to the time axis and amplitude corresponding to the level axis.

Advanced Audio Coding

The wave is infinitely smooth and the chain can be considered to be made up of innumerable points. Since the storage space is relatively limited, in the process of digital encoding, the points of the chain must be sampled. The sampling process consists of extracting the frequency value of a certain point. Obviously, the more points that are extracted in one second, the richer the frequency information that can be obtained. To restore the waveform, there must be two sampling points in one vibration. The highest frequency that can be felt is 20kHz, so to meet the hearing requirements of the human ear, at least 40k samples per second are required, expressed in 40kHz, and these 40kHz are the sampling frequency. Our common CD has a sample rate of 44.1 kHz. It is not enough to have only frequency information, we must also obtain and quantify the energy value of this frequency to represent the strength of the signal. The number of quantization levels is an integer power of 2, and the sample size of our common CD bit is 16 bits, that is, 2 to the power of 16. Sample size is more difficult to understand than frequency. sampling, because it makes it seem abstract. For example, suppose a wave is sampled 8 times and the energy values corresponding to the sample points are A1-A8, but only use a sample size of 2 bits, as a result we can only keep the values of 4 points in A1 -A8 and discard the other 4. If we use the sample size of 3bit, all the information of 8 points is recorded. The higher the sample rate and sample size values, the closer the recorded waveform is to the original signal.
lossy and lossless
According to the sample rate and sample size, it can be known that compared to the natural signal, the audio encoding can only be infinitely close at most, at least the current technology can only do this. Compared to the natural signal, any digital audio encoding scheme is lossy because it cannot be fully restored. In computer applications, PCM encoding can achieve the highest level of fidelity, which is widely used for material preservation and music appreciation. It is used on CDs, DVDs, and our common WAV files. Therefore, PCM has become lossless encoding by convention, because PCM represents the best level of fidelity in digital audio, it does not mean that PCM can guarantee the absolute fidelity of the signal, and PCM can only be infinitely close in the greater extent. We usually include MP3 in the category of lossy audio encoding, which is relatively PCM encoding. The purpose of emphasizing the relativity of lossy and lossless encoding is to tell everyone that it’s hard to achieve true lossless, just like expressing pi with numbers, no matter how high the precision is, it’s infinitely close, not really equal to pi value.
Reasons to use audio compression technology
It is very easy to calculate the bit rate of a PCM audio stream, the value of the sample rate × the value of the sample size × the number of bps of the channel. A WAV file with a sample rate of 44.1 KHz, a sample size of 16 bits, and two-channel PCM encoding has a data rate of 44.1 K×16×2 = 1411.2 Kbps. We usually say that 128K MP3, the corresponding WAV parameter, is this 1411.2 Kbps, this parameter is also called data bandwidth, it is a concept with the bandwidth in ADSL. Divide the bit rate by 8 to get the data rate for this WAV, which is 176.4 KB/s. This means storing a 1-second sample rate of 44.1 KHz, a 16-bit sample size, and a two-channel PCM-encoded audio signal, which requires 176.4 KB of space, which is approximately 10.34 M in 1 minute, which is unacceptable. For most users, especially friends who like to listen to music on the computer, to reduce disk usage, there are only 2 ways to downsample or compress.

MP3 COMPRESSION

To achieve such a dramatic reduction in the number of bits required to transmit an MP audio signal, use different techniques. These techniques include those based on perceptual coding and others such as byte reservation, stereo assembly or Huffman codes. Percentage coding consists of removing all the information that goes into the audio signal that the human ear is not capable of detecting. We will now describe them:

PERCEPTUAL CODING

Minimum hearing threshold The ear’s minimum hearing threshold is the power below which a tone at a given frequency is not capable of being detected by the ear. This threshold is non-linear. As we see in the figure, which represents the Fletcher and Mundson law, the frequencies in which we hear best are those between 2 and 5 Khz. Therefore frequencies outside that band are not totally essential since they will hardly be perceived. Therefore it is possible to remove the content of the audio signal outside these frequencies.

As we can see in the drawing, the range in which a lower power is needed for the tone to be heard is between 2 and 4 Khz.

The masking effect This effect consists in that, when an audio signal has a tone at a given frequency, it produces a masking effect at the frequencies close to it, so that if at these nearby frequencies the signal does not exceed a certain power threshold cannot be heard and therefore it is not necessary to encode them. The form that this power threshold will take according to the position of the tone or the masking tones is what is called the psychoacoustic model, which as the name itself indicates is a perception model that tries to emulate the perception of the human ear.

In this graph we can see how if we put a tone at 1 Khz of 60 dB (masking tone) and then we put another tone at, for example 1.1 Khz and we vary the frequency of this, it is not possible to detect the presence of this second tone until its power exceeds the threshold presented in the figure.

In this case we see various masking tones and the resulting new hearing thresholds. In MP3, what is done is to divide the spectrum to be transmitted (that is, between 2 and 5 Khz) into frequency subbands, so that the power of the subband is evaluated and the masking threshold is created in the nearby subbands. Nearby subbands that exceed that power threshold are coded and those that do not exceed it are not coded.

Furthermore, the masking is not only in appearance but also in time as we can see in the figure.

The byte reserve: Often, some passages of a musical piece cannot be encoded at the same rate without altering the quality of the music. MP · then uses a small byte reservation that acts as a buffer using the capacity of passages that can be encoded at a lower rate in the given stream.
The stereo assembly In the case of a stereo signal, the MP3 format can use a few more tools to further compress the data.
Intensity stereo (IS) The human ear is not able to locate with complete certainty the spatial origin of sounds for very high or very low frequencies. This technique takes advantage of this, recording some frequencies as a monophonic signal, so that a minimum of spatial content is subtracted from the sound.
Mid / Side (M / S) Stereo When the left and right channels are similar then a middle channel (L + R) and a side channel (LR) are created, which are encoded instead of encoding the left channel on one side and the right for another. In this way it is possible to reduce the transmitted data using fewer bits for the lateral channel. Then during playback the MP3 decoder will reconstruct the left and right channels.

Huffman Coding: This coding technique is used at the end of the whole process. It works by creating variable-length codes, so that the symbols that appear in the bitstream most likely have shorter codes. The translation between symbols and codes is done using a table. Each code has a unique prefix so that the codes can be decoded correctly despite their variable length. This type of coding allows on average to reduce by 20% the amount of data to be transmitted. It is an ideal complement to perceptual coding since, during great polyphonies, perceptual coding is very efficient since many sounds are masked, but nevertheless little information is identical and Huffman’s algorithm becomes inefficient. During pure sounds there are few masking effects, but Huffman encoding is very efficient since digitized sound contains many repeating bytes.