Tag: what is digital audio out pcm

Audio. Digital and Analog Audio Part 6


Free Download Mp4Gain
picture

Audio. Digital and Analog Audio Part 6

Digital Audio

ANALOG AUDIO PROCESSING

digital audio

Any processing of an analog audio signal is accompanied by a certain loss of its quality (frequency, phase, non-linear distortions occur), but it is necessary. The main types of processing are as follows:

amplification of the signal to the level required for transmission, recording or playback through the speaker: having sent the signal from the microphone to the speaker, we will not hear anything: it is necessary to pre-amplify it in terms of level and power, while providing the ability to adjust the volume.

frequency filtering: infrasound, which is harmful to health at certain frequencies, and ultrasounds are cut off from the useful sound range (20 Hz – 20 kHz). In many cases, the range is deliberately reduced (the voice phone channel has a band from 300 Hz to 3400 Hz, the frequency band of metered radio stations is significantly limited). For loudspeaker systems, which usually have 2-3 bands, separation is also necessary, which is usually carried out in the crossover filters already at the level of the amplified (powerful) signal.

frequency correction (equalization): tone control, compensation for uneven recoil due to acoustic properties of the room, compensation for losses in transmission lines, studio processing to achieve the desired “color” of sound, suppression of feedback parasitic acoustics (“whistle”), etc., etc.

Noise suppression: there are special dynamic noise reduction schemes that analyze the signal and reduce the bandwidth in proportion to the level and frequency of the RF components (“denoisers”, “dehissers”). In this case, the noise that is above the bandwidth of the signal is cut off and the remaining noise is more or less masked by the signal itself. Such schemes always lead to a very noticeable degradation of the signal, but in some cases their use is appropriate (for example, when working with a recorded speech or on intercom radio stations). For analog sound recording equipment, compressor / expander-based noise cancellers (“compander” eg Dolby B, dbx systems) are also used, the work of which is less perceptible to the ear.
Impact on dynamic range: In order to make the playback of music programs in ordinary home systems, including car radio, rich and expressive enough, the dynamic range is compressed, making the sound of quiet sounds more strong. Otherwise, in addition to the occasional bursts of fortissimo (in classical music), you will have to listen to the silence from the speakers, especially given the noisy environment. For this, devices called compressors are used. In some cases, on the contrary, it is required to expand the dynamic range, then expanders are used. And to exclude exceeding the maximum level, which will lead to clipping (limiting the signal from above, accompanied by very high non-linear distortions, perceived as wheezing), limiters are used in studies.

special effects for studios, EMP, etc.: available to sound engineers and musicians there is a large number of special equipment to give the sound the desired color or to obtain a specific effect. These are various distorters (the sound of an electric guitar becomes hoarse, grainy), wah-wah prefixes (amplitude modulation that causes a characteristic “croaking” effect), enhancers, and exciters (devices that affect the color of the sound, in In particular, it can give the sound a “tube” tint); flangers, choruses, etc.

sound mixing, echo / reverb: recording in studios is usually done in multi-channel form, then, using mixers, the phonogram is reduced to the required number of channels (usually 2 or 6). In this case, the sound engineer can “push forward” one or another solo instrument recorded on a separate track, changing the loudness ratio of different tracks. Sometimes multiple copies of a lower level are superimposed on the signal with a certain time shift, thus simulating natural reverb (echo). Currently, similar and other effects are mainly achieved using signal processors that process digital signals.


Free Download Mp4Gain
picture


Mp4Gain Main Window
picture


Mp4Gain Features
picture


Free Download Mp4Gain
picture

Audio. Digital and Analog Audio Part 5

Audio. Digital and Analog Audio Part 5

Digital Audio

Any amplification path is non-linear, so harmonic distortion always occurs – new frequency components spaced 3, 5, 7, etc. in frequency. of the tone that generates them (odd harmonics) or in 2, 4, 6, etc. times (even).

Digital Audio

 

The threshold of visibility of harmonic distortions varies widely: from a few tenths or even hundredths of a percentage to 3-7%, depending on the composition of the harmonics. Even the harmonics are less noticeable, since they are in line with the fundamental tone (the difference in frequency is twice corresponding to one octave).

In addition to harmonic distortions, intermodulation distortions occur, which are the differential products of the frequencies of the signal spectrum and its harmonics. For example, at the output of an amplifier, at the input of which two frequencies of 8 and 9 Hz are applied (with a sufficiently non-linear characteristic), a third (1 kHz) will appear, as well as several others: 2 kHz (as the difference of the second harmonics of the fundamental frequencies), etc. … Intermodulation distortion is especially annoying to the ear, as it generates many new sounds, including those that are dissonant to the main ones.

What an audiophile can hear, and not only hear, but also explain to a sound engineer, can be completely invisible to the average listener.

Noise and distortion are largely masked by the signal, but they themselves mask low-level signals that fade or lose clarity. Therefore, the higher the signal-to-noise ratio, the better. Actual sensitivity to noise and distortion will vary based on individual hearing characteristics and training. The level of noise and distortion that does not affect the transmission of speech can be completely unacceptable for music. What an audiophile can hear, and not only hear, but also explain to a sound engineer, can be completely invisible to the average listener.

ANALOGUE AUDIO TRANSFER
Traditionally, audio signals were transmitted over cables and over the air (radio).

Distinguish between unbalanced (classic cable) and balanced transmission line. Unbalanced has two wires: signal (direct) and return (ground). Such a line is very sensitive to external interference, so it is not suitable for transmitting a signal over long distances. Often implemented with a shielded cable, the shield is grounded.

cifrovoe-i-analogovoe-audio-4.jpg
FIG. 4. Unbalanced screened line

The balanced line assumes three wires: two signal wires, through which the same signal flows, but in antiphase, and ground. On the receiving side, the common mode noise (induced in both signal wires) is mutually subtracted and completely disappears, and the useful signal level is doubled.

FIG. 5. Balanced screened line

Unbalanced lines are often used inside devices and for short distances, mainly on user routes. In the professional sphere, balance prevails.

In the figures, the shield connection points are shown conditionally, as they must be selected “in place” each time to achieve the best results. Most of the time, the screen is connected only on the signal receiver side.

Audio. Digital and analog audio Part 3

Audio. Digital and analog audio Part 3

DIGITAL AUDIO

Modern autumn sound sources are diverse and digital media are becoming more and more common: CDs, DVDs, although vinyl records are also preserved. We continue to listen to radio, both terrestrial and via cable (radio hotspots). Sound accompanies television shows and movies, not to mention such a familiar phenomenon as telephony.

Digital Audio

 

A computer receives an increasing share in the world of audio, allowing it to conveniently archive, combine and process sound programs in the form of files. In the digital age, digitized speech and music are transmitted through digital channels, including the Internet, without serious losses in transportation. This is done with digital encoding and the loss is due solely to compression, which is used most often. However, in digital media, either it does not exist at all (CD, SACD), or lossless audio compression algorithms are used (DVD Audio, DVD Video). In other cases, the degree of compression is determined by the required level of quality of the soundtrack (MP3 files, digital telephony, digital television, some types of media).

cifrovoe-i-analogovoe-audio-1.jpg
FIG. 1. Conversion of acoustic sound vibrations into an electrical signal

The reverse conversion of electrical vibrations to acoustic vibrations is carried out using speakers built into radios and televisions, as well as separate acoustic systems, headphones.

Sound is called acoustic vibrations in the frequency range 16 Hz to 20,000 Hz.

Sound is called acoustic vibrations in the frequency range 16 Hz to 20,000 Hz. Below (infrasound) and above (ultrasound), the human ear does not hear, and within the sound range, the sensitivity of hearing is very uneven. , its maximum falls at a frequency of 4 kHz. To hear sounds of all frequencies at the same volume, you must play them at different levels. This technique, called loudness, is often implemented in home computers, although its result cannot be considered unequivocally positive.

cifrovoe-i-analogovoe-audio-2.jpg
FIG. 2. Equal volume curves
(Click on the image to zoom)

The physical properties of sound are generally not presented in linear values, but in relative logarithmic values, decibels (dB), as this is much clearer in numbers and more compact in graphics (otherwise one would have to operate with values ​​that they have many zeros before and after the decimal point, and the second would be easily lost in the context of the first). The ratio of two levels A and B in dB (say voltage or current) is defined as:

With u [dB] = 20 lg A / B. If we talk about powers, then C p [dB] = 10 lg A / B.

In addition to the frequency range, which determines the human ear’s sensitivity to tone, there is also the concept of loudness range, which shows the ear’s sensitivity to loudness level and covers the range from the lowest audible sound to the ear (threshold sensitivity) to the strongest, beyond which is the pain threshold. The sensitivity threshold is taken as a sound pressure of 2 x 10-5Pa (Pascal), and the pain threshold is pressure, 10 million times higher. In other words, the audibility range, or the pressure ratio between the loudest and the lowest sound, is 140 dB, which is markedly higher than the capabilities of any audio equipment due to its own noise. Only high definition digital formats (SACD, DVD Audio) match the theoretical limit of dynamic range (the ratio of the loudest sound reproduced by the equipment to the noise level) 120 dB, CD provides 90 dB, vinyl record – approximately 60 dB.

cifrovoe-i-analogovoe-audio-3.jpg
FIG. 3. Hearing sensitivity range

Only high definition digital formats (SACD, DVD Audio) match the theoretical dynamic range limit

Noise is always present in the audio path. This is both the intrinsic noise of the amplifying elements and the external interference. Signal distortions are divided into linear (amplitude, phase) and non-linear or harmonic. In the case of linear distortion, the signal spectrum is not enriched with new components (harmonics), only the level or phase of the existing ones changes. Amplitude distortions that violate the original level relationships at different frequencies result in audible timbre distortions. For a long time it was believed that phase distortions were not critical to hearing, but today the opposite has been shown: both timbre and sound localization are highly dependent on the phase relationships of the signal’s frequency components. .

Audio. Digital and analog audio

Audio. Digital and analog audio

Digital Audio

Although we assimilate most of the external information with the help of our eyes, sound images are no less important to us and often even more.

Digital Audio

Try watching a movie with the sound turned off; in 2-3 minutes you will lose the thread of the plot and the interest in what is happening, no matter how big the screen and the high quality image. Therefore, a pianist played off-screen in silent movies. If you remove the picture and leave the sound, the movie can be “heard” like a fascinating radio show.

Hearing gives us information about what we do not see, since the sector of visual perception is limited, and the ear captures the sounds that come from everywhere, complementing the visual images.

Hearing gives us information about what we do not see, since the visual perception sector is limited, and the ear captures sounds from all directions, complementing visual images. At the same time, our hearing with great precision can locate an invisible sound source in direction, distance, speed of movement.

They learned to convert sound into electrical vibrations long before images. This was preceded by a mechanical recording of sound vibrations, whose history dates back to the 19th century.

Accelerated progress, including the ability to transmit sound at a distance, was made possible by electricity, with the advent of amplification, acoustic and electroacoustic technology and transducers – microphones, pickups, dynamic heads, and other emitters. Today, audio signals are transmitted not only over cables and over the air, but also over fiber optic communication lines, primarily in digital form.

Acoustic vibrations are converted into an electrical signal, usually by microphones. Any microphone contains a moving element whose vibrations generate a current or voltage in a certain way. The most common type of microphone is the dynamic one, which is a reverse speaker. The vibrations of the air set in motion a membrane that is rigidly connected to a moving coil in a magnetic field. A condenser microphone is, in fact, a condenser, one of whose plates vibrates in time with the sound, and with it the capacitance between the plates changes. Ribbon microphones use the same principle, only one of the plates is freely suspended. Similar to a condenser electret microphone, whose plates, in the process of oscillation, generate by themselves an electric charge proportional to the amplitude of the oscillations. Many models of microphones have a built-in amplifier (the level of the signal directly from the acoustic-electric transducer is very low). Unlike a microphone, the pickup of an electric musical instrument registers vibrations not from air, but from a solid body: a string or the soundboard of an instrument. The cartridge reads the disc slot using a stylus mechanically connected to moving coils in a magnetic field, or magnets if the coils are stationary. Or the vibrations of the needle are transmitted to the piezoelectric element which, under mechanical stress, generates an electrical charge. In magnetic recording, an audio signal is recorded on a magnetic tape and then read with a special head. Finally, in cinematography, optical recording was traditionally adopted: an opaque soundtrack was applied from the edge of the film,

In synthesizers, sound is born directly in the form of electrical vibrations, there is no primary transformation of acoustic waves into an electrical signal.

History of Digital Audio Part 2

History of Digital Audio Part 2

Digital Audio

Different formats use different methods of audio compression, but bit rate still plays a role as a measure of audio quality. The sample rate also plays an important role and the number of hertz shows how many parts per second the file is divided into. The lower limit of the sample rate for audio files is 44.1 kHz (44100 Hz), if it is lower, it is not sufficient.

digital audio

VBR vs CBR

Constant Bit Rate (CBR) and Variable Bit Rate (VBR) are two methods of obtaining Bit Rate. Constant bitrate means that you set a certain bitrate for the entire file, and with a variable bitrate, its value changes throughout the entire music file as needed.

CBR is like packing something in a larger box than necessary, and VBR packs in a box that matches the outline of its contents. People often use an overestimated bit rate of 320 kbps, when this is not necessary, often a VBR of 192 kbps is sufficient. By ear, you are unlikely to feel a difference.

DRM

DRM (Digital Rights Management) is the most terrible invention since the nuclear bomb and is best left untouched. Music stores primarily use DRM protection to protect it from illegal copying and use.

DRM files are not compatible with all players and you may forget to transfer files in MSC / UMS mode with them. DRM-protected music is usually in WMA or AAC formats. In short, the use of DRM only creates additional problems for people.

History of digital audio

History of digital audio

digital audio

By its nature, sound is an oscillatory movement of particles in an elastic medium that propagates in the form of waves. After it became clear that sound represents such vibrations, the idea came up of recording them by repeating the shape on solid material.

DIGITAL AUDIO

So, in 1877, Thomas Edison created a phonograph, a device for the mechanical recording and reproduction of sound. And in 1888, the German E. Berliner invented the gramophone – the era of gramophone records began, which became the first massive carriers of audio information.

Thomas Edison and his phonograph

FIG. Inventor Thomas Edison and His Record Making: The Phonograph

Having studied the laws of electromagnetism, man made successful experiments to convert sound waves into electromagnetic waves and preserve them. This is how magnetic tape appeared, which became widespread in the middle of the 20th century.

For digital technology to store, process, and reproduce sound, it is converted to digital format by an analog-to-digital converter (ADC), which converts an analog signal into a sequence of numbers. This is called Pulse Code Modulation (PCM).

It happens like this: the ADC measures the amplitude of an analog signal many times per second and outputs the results in the form of numbers. However, the measurement result does not exactly match a continuous electrical signal: it depends on the number of measurements and their precision.

The frequency at which the measurements are taken is called the sample rate, and the precision of the amplitude measurements indicates the number of bits used to indicate the result of the measurement. This parameter is called the bit depth. For example, if the sampling frequency is 44.1 kHz, this means that the signal is measured 44 100 times in one second.

For the analog signal to be accurately reconstructed from its samples, the sample rate must be twice the maximum audio frequency. That is, if the analog signal contains frequency components from 0 Hz to 20 Hz, then the frequency of its sampling must be at least 40 kHz.

Digital audio formats

Of course, for digitized sound to be stored, transmitted, and converted, there must be certain digital sound standards – audio formats. Today, there are many such formats, each of which uses its own sound processing algorithm. They also differ in the information carriers.

The most popular and widespread in the field of home use today are ordinary music CDs – CDs. There are also relatively new recording formats, Super Audio Compact Disk (SACD) and DVD-Audio (or simply DVD-A). In addition, formats that use digital data compression have become widespread.

The most popular among them is MPEG-1/2 / 2.5 Layer 3 (MP3). Microsoft also did not stay away from the sound industry, as it developed its own compression algorithm, WMA, which is also actively promoted in the market.

New audio file formats appear every year, but no player on the market supports the playback of all formats.

In fact, the term MP3 player is only correct for players that support the MP3 format. Let’s see what’s what in audio formats.

Before looking at the various audio file formats (codecs), let’s take a look at a few terms.

Bitrate

Bit rate is the space required for 1 second of music. With a bit rate of 128 kbps (kilobits per second) = 16 kbps (kilobytes per second), approximately 5 megabytes are needed for 5 minutes of music.

The higher the bit rate, the higher the quality of the music. But this as long as the bit rate of the original format is higher than the bit rate of the encoded format. By compressing a CD to MP3 at 320 kbps, you get better sound quality than 128 kbps, but converting from 128 kbps to 320 kbps will not improve the quality and may even degrade it.

Often times a 128kbps bit rate masquerades as CD quality, but this is not actually the case. If you have enough high-quality equipment, you will hear it immediately. Manufacturers like to give an estimate of the number of songs that go into a player at a very low bit rate, and many consumers are unaware that audio files vary in size. Therefore, you should not rely on the numbers in the advertisements, in fact, much less the songs in your collection can fit in the player.

Compression

Uncompressed audio takes up a lot of space. To reduce the size of audio files in formats such as MP3, programs cut off the part of the frequency range that the human ear cannot hear.

What is bitrate?

What is bitrate?

Bitrate

Bitrate

Bitrate

Bit rate: the number of bits of information used to store or transfer one second of data transmission: video and / or audio recordings, including compressed ones.

Bit rate is expressed in bits per second (bit / s, bps), as well as derived values: kilo (kbps, kbps), mega (Mbps, Mbps), etc.

For streaming video and audio formats (such as MPEG and MP3) that use lossy compression]], the bit rate expresses the degree of compression of the stream. Most of the time, the video and audio bit rate is measured in megabits per second.

Increasing the bitrate provides a significant increase in video recording quality, which is especially noticeable when shooting dynamic scenes and small details.

Encoding modes
There are three compression modes for data transmission:

CBR (constant bit rate): with constant bit rate;
VBR (variable bit rate): with variable bit rate;
ABR (Average Bit Rate): with an average bit rate.

Constant bit rate
Constant Bit Rate, CBR – A variant of streaming data encoding, in which the required bit rate is initially set, which does not change throughout the file.

Its main advantage is the ability to predict the size of the final file fairly accurately.

However, the constant bitrate option is not very suitable for video or audio content, the dynamics of which change over time, as it does not provide an optimal size / quality ratio.

Variable bit rate
With a variable bit rate, the VBR codec selects the value of the bit rate based on the parameters (the level of the desired quality), and during the encoded segment, the bit rate may change.

This method provides the best quality / size ratio for the output file, but its exact size turns out to be very unpredictable. Depending on the nature of the sound (or image, in the case of video encoding), the size of the resulting file may differ several times.

Average bit rate
Average bit rate, ABR is a hybrid of constant and variable bit rates: the value in Mbps is set by the user and the program varies it within certain limits. However, unlike VBR, the codec is careful to use the maximum and minimum possible values, without risking going beyond the average specified by the user. This method allows the most flexible setting of the processing speed and with much higher precision (compared to VBR) in predicting the output file size.

What is digital audio?

What is digital audio?

Digital Audio

Digital sound is nothing more than a combination of numbers.

DIGITAL AUDIO

With a certain algorithm, sound, such as air pressure, is converted into data streams and encoded for further processing and playback. Depending on the algorithm used, the music file has one format or another, one or another extension.

Remember that along with digital sound, there is analog sound, which is represented by a continuous electrical signal that reflects the change in the sound wave. The analog to digital sound conversion is a setting of the numerical value of the amplitude at a given time with a given density of values. Consequently, the more values ​​that are recorded, the more reliable and accurate the image of the digitized sound fragment is recreated. With such digitization, very voluminous data matrices emerge that, depending on the format used, differ in the sound quality / volume ratio of the final file.

Perhaps the main advantage of digital audio over analog is the ability to store and copy data indefinitely without losing the original quality (whereas when copying from one analog medium to another, a decrease in recording quality is quite noticeable).

The most widespread and popular digital audio format today is MP3 (MPEG Layer 3). It was developed, after a series of intermediate formats and investigations, started in 1987, by the Fraunhofer Institute in Germany.

The developers of the format were faced with the task of simplifying and reducing the cost of shipping long musical fragments. As you know, one minute of a stereo signal from a CD (16 bit, 44.1 kHz sample rate) takes up about ten megabytes of memory. At the same time, unlike text or graphic files, the audio signal cannot be compressed without loss of quality. Thus, modem transmission of an uncompressed composition from an audio CD lasting 3 minutes at a data transfer rate of, say, 24 kbps will take several hours. Scientists at the Fraunhofer Institute managed to achieve multiple file size compression: on average, one minute of a compressed audio signal in MP3 format takes about 1 megabyte. The principle of compression is based on the removal of “unnecessary” sounds from the music file, to which the human ear is immune, or which duplicate each other.

The main factor that determines the relationship between file size and sound quality within a given format is the bit rate. Bit rate is an indicator of how much information a second of sound encodes. The higher it is, the less distortion and the closer the encoded composition is to the original. The most common on the Internet are compositions with 128 and 192 Kbps bitrates. The maximum bitrate supported by programs and devices that work with MP3 is 320 Kbps. In practice, only an expert or a professional who works with sound can notice the differences between an MP3 file with a 320 bit rate.

To optimize the size of MP3 music files while maintaining decent quality, a variable bit rate (abbreviation VBR – variable bit rate) is used. In this case, the encoding program divides the file into fragments of different spectral saturation and encodes them with a suitable bit rate. Most modern MP3 players support variable bit rate playback. A significant advantage of MP3 files is that they can contain the name of the artist, the name of the track and the album, the year of its release, etc. The set of this data is called ID3 tags. Most modern gamers can read and display them on the screen.

In 2001, Swedish Coding Technologies and Thomson Multimedia developed the MP3 Pro codec. It is MP3-based and as a result is fully MP3 backward compatible and only partially forward compatible. It uses SBR (Spectral Band Replication) technology, so the codec provides good quality at low bit rates. However, the encoding quality at medium to high bit rates is inferior to that of most other codecs. For this reason, this format is mainly used for broadcasts on the Internet and demonstrations of fragments of new musical compositions.

Another type of MP3 was the development of MP3 Surround, recently introduced by the creators of MP3: the Fraunhofer Institute. This format repeats all the characteristics of multi-channel sound, while still being compatible with standard stereo MP3: information describing the spatial characteristics of the sound is recorded on an additional track. By playing files of this format on special equipment capable of reading this track, you can obtain surround sound that conforms to the Surround 5.1 standard.

The beginning of the digital age

The beginning of the digital age

digital audio

binary code

digital audio

Although digital audio is the standard of music these days …

It has not always been this way.

Music originally existed only in the form of sound waves.

Then, with the development of technology, ways were discovered to convert it to other formats, such as:

Musical notation
electrical signals in cables
radio waves in the atmosphere
request on vinyl record
But more recently, in the age of computers, digital audio has become the main recording format, making it easy to copy and transfer songs.

The device that made this possible is called … digital converter.

Also, on how it works …

2. Digital converters
In recording studios, digital converters exist in 2 versions:

as a standalone device in top studios or …
as part of an audio interface in home studios.
To make binary code out of sound, they take tens of thousands of images (samples) per second to build a rough image of an analog wave.

This image is not entirely accurate, because in the moments between samples, the converter has to guess what is happening.

digital wave

As seen in the graphic above:

the red line shows an analog signal and …
black line shows conversion …
The results are not ideal, but sufficient to produce excellent sound quality.

And the difference depends mainly on …

3. Sampling rate
Take a look at this image:

sampling rate circuit

As can be seen …

By capturing more images per second, higher sampling rates:

Collect more real information,
Use less guesswork,
Creates a cleaner display from an analog signal
And in the end, you get the best sound quality.

Now let’s talk about specific numbers:

Standard sample rates in professional audio:

44.1 kHz (CD)
48 kHz
88.2 kHz
96 kHz
192 kHz
44.1 kHz is the minimum sample rate due to a mathematical principle known as …

Kotelnikov’s theorem (Nyquist-Shannon)
To accurately record digital audio, converters must capture the full spectrum of human hearing between 20 Hz and 20 kHz.

According to Kotelnikov’s theorem …

Capturing a specific frequency requires at least 2 samples per cycle … to measure both the high and low points of a wave.

This means that a sample rate of 40 kHz or more is required to record frequencies up to 20 kHz. Therefore, the sampling frequency of CDs is slightly higher, 44.1 kHz.

Kotelnikov’s theorem

Cons of a high sample rate
Although the higher the sample rate, the higher the sound quality … but this just doesn’t happen.

The cons are:

Requires a lot of computing power
Less clues
Large audio files
So this is a constant search for a compromise. Professional studios find it easier to deal with high sample rates because they have the best equipment.

However, for most home studios, the standard 48 kHz sample rate is appropriate.

How does encoding work in digital audio? Part 5

How does encoding work in digital audio? Part 5

encoding digital audio

DSD offers significant advantages over PCM:

encoding digital audio

more precisely draw a wave;
increased immunity to noise;
an easier way to change and transmit a digital stream;
In theory, it is possible to reduce cost by simplifying DAC circuits, but due to backward compatibility, manufacturers are unlikely to do so.
Originally, SACDs used the DSD x64 format with a sample rate of 2822.4 kHz. The 44.1 kHz audio CD sample rate was taken as the basis, increased 64 times, hence the name x64. The following DSDs are currently in use:

x64 = 2822.4 kHz;
x128 = 5644.8 kHz;
x256 = 11,289.6 kHz;
x512 = 22,579.2 kHz;
declared DSD x1024.

DXD
There is a certain intermediate format between PCM and DSD called DXD – Digital eXtreme Definition. This is, in fact, high definition PCM: 352.8 kHz or 384 kHz with 24 or 32 bit quantization. It is used in studies for the processing and subsequent mixing of materials.

But this approach is flawed: firstly, it does not allow to use all the advantages of DSD, and secondly, the file size is larger than in DSD. At the moment, flagship DACs on the I2S input accept a PCM data stream with a sample rate of up to 768 kHz and a bit depth of up to 32 bits. It’s scary to even consider how much hard drive space an album will take up at this resolution.

DSD has practically separated from SACD. Now, the DSD format can often be found packaged in files with the DSF and DFF extensions. Many turntables have been released with the ability to record in DSF and DFF, lovers of good sound are increasingly digitizing vinyl records in the DSD format. But in recording studios, nobody wants to invest in unpopular formats, so they continue to rivet the sound with a minimum wage: 44.1 × 16.

DSD switching and data transmission
To transfer a digital transmission to DSD, a three-pin connection scheme is used:

DSD Clock Pin (DCLK) – sync;
Data input pin DSD Lch (DSDL) – left channel data;
Data input pin DSD Rch (DSDR): Right channel data.

Unlike I2S, DSD data transmission is extremely simplified. DCLK sets the clock rate of the bit sync, and the left and right channel data is transmitted sequentially through the DSDL and DSDR pins, respectively. Here there are no adjustments, recording and playback in DSD is done little by little. This approach provides the closest approximation to the analog signal, and due to the high frequency, the quantization noise is reduced and the reproduction precision is increased by an order of magnitude.

PDO
DoP is often used to carry DSD data streams, so it’s worth mentioning. DoP is an open standard for transferring DSD data over PCM frames (DSD over PCM). The standard was created to transmit a stream through controllers and devices that do not support direct DSD streaming (not native DSD).

The principle of operation is as follows: in a 24-bit PCM frame, the upper 8 bits are padded with ones; this means that DSD data is currently being transmitted. The remaining 16 bits are sequentially filled with DSD data bits.

For x64 DSD transmission with a single bit rate of 2822.4 kHz, a PCM sample rate of 176.4 kHz (176.4 x 16 = 2822.4 kHz) is required. For DSD x128 transmission at 5644.8 kHz, a PCM sampling rate of 352.8 kHz is already required.