Hi-Fi in the digital age. Part 1
Hi-Fi ( English High Fidelity - high accuracy ) - a term meaning that the reproduced sound is very close to the original. Hi-Fi is also a sound engineering industry, where all equipment is designed to ensure maximum fidelity of sound reproduction.
There is no doubt that the need to listen to music is one of the essential human needs. Not the most necessary for survival - but one of the most important - after survival is ensured.
')
Before the invention of the phonograph - the only way to listen to music was live musicians.
The term Hi-Fi itself (High Fidelity - “high fidelity” reproduction) was introduced by the English engineer Hartley back in 1932 to advertise his permanent magnet loudspeakers. For the same years, the first experiments on the creation of stereo recordings occurred.
One of the first proposals of the stereo recording method with the help of two independent recorders on two sides of the sound groove was published in 1941 in the American magazine Electronics. However, the first stereo plates appeared only in 1957. But their appearance was preceded by another significant event - the invention of LP (Long Play) LP at 33.1 / 3 vol. per minute by Columbia Broadcasting. And it was the second audio revolution, since the LP could sound 50 minutes while listening to both sides, while the usual 78-rpm record was only 8 minutes. They had a frequency range of 30-13000 Hz with a signal-to-noise ratio of 35-40 dB, and in 1950 it increased by another 10 dB after the invention of a variable-pitch recording method.
Improving the quality of the sound of vinyl records was a powerful incentive to the rapid improvement of amplifiers and speakers.
The appearance of stereo recordings in 1957 accelerated the development of household appliances even more sharply, and the “Golden” 60s were marked by the heyday of stereo systems, which by the end of the decade had won an unconditional victory over mono.
The best models of stereo players could reproduce signals in the frequency band of 10-25000 Hz and even more. For such parameters, perhaps, it will not be a shame to a decent “vertaka” in our days.
Tape and tape will probably be omitted altogether, they were just as an episode which has already passed, while vinyl records are alive, and since 2005 they have experienced steady sales growth amid a decrease in interest in AudioCD (CDDA) and an increase in interest in digital lossless files. .
In the later years (the 80s and 90s) after the appearance of the CD and its wide distribution, component systems became widespread, including:
There are also hybrids FM tuner and integrated stereo amplifier - stereo receivers.
Note: The article deals only with hi-fi music equipment and stereo music. Multichannel theater systems are beyond the scope of this article (they are not designed for music at all) and (IMHO) they are not hi-fi technology.
Nowadays, the above kit still remains a common way to assemble a Hi-Fi kit at home, but more modern digital formats and technologies are on the way.
We will talk about them ...
0.1 Nowadays, any sound in digital form is recorded in a PCM stream (in pulse code modulation (PCM, pulse-code modulation)) of varying degrees of discretization, but if we are talking about music CDDA, this is a stream of 16 bit 44.1 KHz. And if this is digitizing from vinyl or studio master recording, then the ADCs available to those who did digitization: allow you to record 20 or 24 bits and from 96 to 192 KHz.
The transport in the CD player is quite important, because any interruption or wrong data in the PCM digital stream will cause a misinterpretation of this DAC stream. In expensive specialized CD players, transport is given a lot of attention, the accuracy of the mechanism and the laser, isolation from the power supply circuits, amplification circuits, a heavy chassis and leg spikes to reduce vibrations, etc.
Usually it is WAV (just pure PCM digital stream), FLAC, APE, ALAC, WAVPack (PCM lossless compressed stream).
In such a “transport”, the continuous unpacking of compressed formats into a PCM stream plays an important role, modern processors cope with this without problems.
But such a transport has some problems: for example, noise from fans, interference and pickups from electrical circuits, which in one way or another can affect the PCM flow.
DAC can be of different types:
Here the scheme is as follows: PCM stream -> DAC -> Timbre Block -> Pre-Amplifier (Pre-Amp) -> Line Out .
There are also direct signal options that bypass timbre and preamplifiers (Direct / Pure Audio modes) from a DAC to adigital SPDIF output (coaxial (RCA) or optical (Toslink) line output.
It is in the DAC that the first obvious loss of sound quality occurs, even if it is theoretically assumed that digital PCM stream goes perfectly clean, without loss or interruption.
The “figure” differs from the “analog” in two main features:
- A signal that is continuously changing in analog form is represented in digital form by a fixed number of discrete numerical values. These numerical values ​​represent the signal is not constant at all times, but only at certain points in time, moments of quantization.
- The quantization moments are determined by the D / A converters, which have a master oscillator to control the quantization frequency or the sampling frequency.
In practice, the digital signal encounters such problems: interruption of the signal, instability of the clock generator, a drop in the signal characteristics in the cable, parasitic pickups. All this leads to a change in the waveform and small shifts in the time domain, or jitter. Insufficiently accurate calculation of "jitter" can lead to its accumulation when a digital signal passes through a path.
In sound, these are distortions similar to detonation, a phenomenon caused by the uneven movement of a magnetic tape in an analog tape recorder due to the imperfection of the tape drive mechanism. However, the distortion introduced by digital jitter is much more noticeable than the distortion of sound caused by detonation. Apparently, this is due to the greater “softness” and “smoothness” of detonation distortion (it can be said, “analog jitter”), due to the elasticity of the magnetic tape and the inertia of the mechanical elements of the tape drive mechanisms.
And besides this, there are distortions and inaccuracies in the recovery of an analog signal from a discrete (for example, dynamic range or non-linear distortion), and the type of DAC and algorithms for recovering an analog signal from a discrete PCM stream play an important role. As well as the resolution of the DAC (from 16bit / 44.1Khz to 24bit / 192Khz) and the presence of oversampling steps.
Some digital processors use the so-called one-bit D / A converters (noise-converting DAC, Delta-Sigma DAC, DAC with oversampling). All these converters work on the same principle: to convert a binary code into an analog signal instead of a resistor matrix with different weights of bits, a single-bit DAC uses only two states, zero and one. A one-bit code is a series of pulses of constant amplitude varying in duration.
These DACs are the cheapest and also contain a digital filter, a dual-channel DAC, a dual current-to-voltage converter, a dual adjusted resistor for adjusting the current, and two analog output stages. That is a kind of cheap „all-in-one pie“. From it, the signal already goes to the linear output, or to the operational amplifiers (output to the headphones, for example).
Their problems: Processors with single-bit DACs do not have enough bass elasticity and dynamics. They may have a soft and transparent sound, but they often lack the rhythmic intensity and “sharpness” of multi-bit DACs.
The most expensive DAC, because it requires full digital filtering, but providing the maximum speed and accuracy of conversion.
Inside these DACs there is a resistor-type ladder circuit. Each "step" corresponds to one binary code category. Consequently, a 16-bit DAC will have 16 impedance steps, which provide 65,536 possible input codes. The accuracy of the resistors of these resistors is crucial. Any deviation from the values ​​of R / 2R leads to the fact that the weight of the corresponding bit becomes more or less than necessary.
This is a rather complicated device and very expensive, an example is the cost of such DACs from Burr-Brown.
DAC resolution is different from 8-10-12-16 bits \ 22-44.1-48 KHz to 20-24 bits \ 96-192 KHz.
Some differences in the accuracy of the restoration of an analog signal are best demonstrated by a picture of restored sinusoidal signals:
These are signals at the D / A output at 24 bits and 20 bits (-110 dB)
These are signals at the D / A output at 24 bits and 20 bits (-120 dB)
And another picture, the dependence of THD non-linear distortion coefficient and DR dynamic range on the number of bits in the DAC:
The greater the resolution (“bit” and “kilohertz”), the more accurate the sound recovered from PCM. But in practice, of course, not everything is so simple, any DAC also has an error in conversion, which, among others, includes two very important characteristics - nonlinearity and non-monotonicity of the conversion characteristic (CP). Here, as elsewhere, the rule applies: "the more expensive the better."
Here, in general, that's all, further amplification of the signal and speakers. This is a different conversation for other articles.
But I want to say finally, in a simple way:
DACs built-in sound cards for motherboards and laptops (AC97 and new HD Audio) from the manufacturers Analog Devices and Realtek, etc. - All of them are the simplest one-bit DACs, and are made on the principle of maximum cheapness.
All this has a negative effect on sound and is heard even in less-normal full-size multimedia speakers. And the best way to improve the sound from a PC or laptop is to buy an external DAC (for example, for Creative Professional E-MU musicians or Cambridge Audio DAC music lovers, etc.) or a separate sound card with good DACs and operational amplifiers.
A couple of links interested in:
1. CD-players DAC and Transport List
2. Bit rate, resolution and brief characteristic of DAC chips
3. Digital to Analog Converters
4. Basic concepts and terms
There is no doubt that the need to listen to music is one of the essential human needs. Not the most necessary for survival - but one of the most important - after survival is ensured.
')
A bit of history
Before the invention of the phonograph - the only way to listen to music was live musicians.
The term Hi-Fi itself (High Fidelity - “high fidelity” reproduction) was introduced by the English engineer Hartley back in 1932 to advertise his permanent magnet loudspeakers. For the same years, the first experiments on the creation of stereo recordings occurred.
One of the first proposals of the stereo recording method with the help of two independent recorders on two sides of the sound groove was published in 1941 in the American magazine Electronics. However, the first stereo plates appeared only in 1957. But their appearance was preceded by another significant event - the invention of LP (Long Play) LP at 33.1 / 3 vol. per minute by Columbia Broadcasting. And it was the second audio revolution, since the LP could sound 50 minutes while listening to both sides, while the usual 78-rpm record was only 8 minutes. They had a frequency range of 30-13000 Hz with a signal-to-noise ratio of 35-40 dB, and in 1950 it increased by another 10 dB after the invention of a variable-pitch recording method.
Improving the quality of the sound of vinyl records was a powerful incentive to the rapid improvement of amplifiers and speakers.
The appearance of stereo recordings in 1957 accelerated the development of household appliances even more sharply, and the “Golden” 60s were marked by the heyday of stereo systems, which by the end of the decade had won an unconditional victory over mono.
The best models of stereo players could reproduce signals in the frequency band of 10-25000 Hz and even more. For such parameters, perhaps, it will not be a shame to a decent “vertaka” in our days.
Tape and tape will probably be omitted altogether, they were just as an episode which has already passed, while vinyl records are alive, and since 2005 they have experienced steady sales growth amid a decrease in interest in AudioCD (CDDA) and an increase in interest in digital lossless files. .
In the later years (the 80s and 90s) after the appearance of the CD and its wide distribution, component systems became widespread, including:
- Cd player
- FM tuner
- integrated stereo amplifier (or bundle of preamp and power amplifier)
- shelf or floor speakers
There are also hybrids FM tuner and integrated stereo amplifier - stereo receivers.
Note: The article deals only with hi-fi music equipment and stereo music. Multichannel theater systems are beyond the scope of this article (they are not designed for music at all) and (IMHO) they are not hi-fi technology.
Nowadays, the above kit still remains a common way to assemble a Hi-Fi kit at home, but more modern digital formats and technologies are on the way.
We will talk about them ...
Digital sound sources
0. Digital sound source:
0.1 Nowadays, any sound in digital form is recorded in a PCM stream (in pulse code modulation (PCM, pulse-code modulation)) of varying degrees of discretization, but if we are talking about music CDDA, this is a stream of 16 bit 44.1 KHz. And if this is digitizing from vinyl or studio master recording, then the ADCs available to those who did digitization: allow you to record 20 or 24 bits and from 96 to 192 KHz.
1. Transport (PCM digital stream transport):
1.1 In the case of playing from a CD player:
The transport in the CD player is quite important, because any interruption or wrong data in the PCM digital stream will cause a misinterpretation of this DAC stream. In expensive specialized CD players, transport is given a lot of attention, the accuracy of the mechanism and the laser, isolation from the power supply circuits, amplification circuits, a heavy chassis and leg spikes to reduce vibrations, etc.
1.2 In the case of playing a sound file from a computer or multimedia player:
Usually it is WAV (just pure PCM digital stream), FLAC, APE, ALAC, WAVPack (PCM lossless compressed stream).
In such a “transport”, the continuous unpacking of compressed formats into a PCM stream plays an important role, modern processors cope with this without problems.
But such a transport has some problems: for example, noise from fans, interference and pickups from electrical circuits, which in one way or another can affect the PCM flow.
2. DAC (DAC, Digital-to-Analog Converter)
DAC can be of different types:
- as a separate device: it receives a digital stream to the SPDIF or USB / FireWire input and outputs an analog signal to the line output.
- part of the cd player.
- part of the integrated amplifier or receiver.
Here the scheme is as follows: PCM stream -> DAC -> Timbre Block -> Pre-Amplifier (Pre-Amp) -> Line Out .
There are also direct signal options that bypass timbre and preamplifiers (Direct / Pure Audio modes) from a DAC to a
It is in the DAC that the first obvious loss of sound quality occurs, even if it is theoretically assumed that digital PCM stream goes perfectly clean, without loss or interruption.
For example, one of the problems DAC is "jitter":
The “figure” differs from the “analog” in two main features:
- A signal that is continuously changing in analog form is represented in digital form by a fixed number of discrete numerical values. These numerical values ​​represent the signal is not constant at all times, but only at certain points in time, moments of quantization.
- The quantization moments are determined by the D / A converters, which have a master oscillator to control the quantization frequency or the sampling frequency.
In practice, the digital signal encounters such problems: interruption of the signal, instability of the clock generator, a drop in the signal characteristics in the cable, parasitic pickups. All this leads to a change in the waveform and small shifts in the time domain, or jitter. Insufficiently accurate calculation of "jitter" can lead to its accumulation when a digital signal passes through a path.
In sound, these are distortions similar to detonation, a phenomenon caused by the uneven movement of a magnetic tape in an analog tape recorder due to the imperfection of the tape drive mechanism. However, the distortion introduced by digital jitter is much more noticeable than the distortion of sound caused by detonation. Apparently, this is due to the greater “softness” and “smoothness” of detonation distortion (it can be said, “analog jitter”), due to the elasticity of the magnetic tape and the inertia of the mechanical elements of the tape drive mechanisms.
And besides this, there are distortions and inaccuracies in the recovery of an analog signal from a discrete (for example, dynamic range or non-linear distortion), and the type of DAC and algorithms for recovering an analog signal from a discrete PCM stream play an important role. As well as the resolution of the DAC (from 16bit / 44.1Khz to 24bit / 192Khz) and the presence of oversampling steps.
Consider some types of DAC
2.1 One-bit DAC. Delta-Sigma (DS) conversion. (Cirrus Logic “Crystal”, Realtek, Analog Devices, Philips Bitstream, MASH Matsushita, REM JVC)
Some digital processors use the so-called one-bit D / A converters (noise-converting DAC, Delta-Sigma DAC, DAC with oversampling). All these converters work on the same principle: to convert a binary code into an analog signal instead of a resistor matrix with different weights of bits, a single-bit DAC uses only two states, zero and one. A one-bit code is a series of pulses of constant amplitude varying in duration.
These DACs are the cheapest and also contain a digital filter, a dual-channel DAC, a dual current-to-voltage converter, a dual adjusted resistor for adjusting the current, and two analog output stages. That is a kind of cheap „all-in-one pie“. From it, the signal already goes to the linear output, or to the operational amplifiers (output to the headphones, for example).
Their problems: Processors with single-bit DACs do not have enough bass elasticity and dynamics. They may have a soft and transparent sound, but they often lack the rhythmic intensity and “sharpness” of multi-bit DACs.
2.2 D / A - Modified Resistive Matrix or R-2R (Burr-Brown)
The most expensive DAC, because it requires full digital filtering, but providing the maximum speed and accuracy of conversion.
Inside these DACs there is a resistor-type ladder circuit. Each "step" corresponds to one binary code category. Consequently, a 16-bit DAC will have 16 impedance steps, which provide 65,536 possible input codes. The accuracy of the resistors of these resistors is crucial. Any deviation from the values ​​of R / 2R leads to the fact that the weight of the corresponding bit becomes more or less than necessary.
This is a rather complicated device and very expensive, an example is the cost of such DACs from Burr-Brown.
2.3 Now a little about the resolution of the conversion, this allows us to speak about the accuracy and delicacy of handling the signal with which the DAC restores the analog signal
DAC resolution is different from 8-10-12-16 bits \ 22-44.1-48 KHz to 20-24 bits \ 96-192 KHz.
Some differences in the accuracy of the restoration of an analog signal are best demonstrated by a picture of restored sinusoidal signals:
These are signals at the D / A output at 24 bits and 20 bits (-110 dB)
These are signals at the D / A output at 24 bits and 20 bits (-120 dB)
And another picture, the dependence of THD non-linear distortion coefficient and DR dynamic range on the number of bits in the DAC:
The greater the resolution (“bit” and “kilohertz”), the more accurate the sound recovered from PCM. But in practice, of course, not everything is so simple, any DAC also has an error in conversion, which, among others, includes two very important characteristics - nonlinearity and non-monotonicity of the conversion characteristic (CP). Here, as elsewhere, the rule applies: "the more expensive the better."
Here, in general, that's all, further amplification of the signal and speakers. This is a different conversation for other articles.
But I want to say finally, in a simple way:
DACs built-in sound cards for motherboards and laptops (AC97 and new HD Audio) from the manufacturers Analog Devices and Realtek, etc. - All of them are the simplest one-bit DACs, and are made on the principle of maximum cheapness.
All this has a negative effect on sound and is heard even in less-normal full-size multimedia speakers. And the best way to improve the sound from a PC or laptop is to buy an external DAC (for example, for Creative Professional E-MU musicians or Cambridge Audio DAC music lovers, etc.) or a separate sound card with good DACs and operational amplifiers.
A couple of links interested in:
1. CD-players DAC and Transport List
2. Bit rate, resolution and brief characteristic of DAC chips
3. Digital to Analog Converters
4. Basic concepts and terms
Source: https://habr.com/ru/post/109901/
All Articles