Analog and digital audio difference
Is analog better than digital audio?
What is best connection for audio?
Frequently Asked Questions
Analog and digital audio difference
Is analog better than digital audio?
What is best connection for audio?
Frequently Asked Questions
HD audio file conversion
Audio Basis - articles about audio
Many disputes, what is better analog or digital audio signal, can digital sound be good as analog, are continued despite digital devices successfully and rapidly replaced analog ones. There is an opinion, the main difference between digital and analog audio is analog closer to original. It looks like, the analog waveform is smoother than "stairs" of digital music. On the other hand, a digital signal is not "stairs" actually. Not all is quite so simple. Read to end the easy and detailed explanation by audio software developer Yuri Korzunov.
For quick entry to the topic, read frequently asked questions. After it, read the full article to better understand what is better analog or digital audio, and why.
Technically, digital systems and records (their mediums) are capable cause lesser degrading sound issues, than analog ones.
The analog degrading causes sound "colorization", which may sound "nice" subjectively.
On the other hand, digital systems have different quality due to various implementations. And important features of a digital system may be worse than similar features of an analog one.
Example: if a digital system has a frequency limitation of 16 kHz, it may sound worse than an analog system with 18 kHz limitation.
Resume: What's better analog or digital is a matter of personal taste and available system.
Read more about the abilities and issues of digital systems here...
Modern digital recordings have lesser-distortion abilities.
The analog recording has specific distortions that may "embellish" sounding for listener.
If you look for a more exact reproduction of acoustic sources before microphone, digital systems are more recommended.
Read more about sound quality...
Digital signal is a form of coding of analog signal.
Digital sound has these benefits comparing an analog one:
Read more...
Digital signals have zero losses in communication channels and on mediums, in most cases. It was achieved due to special coding and recovering ability with probability close to 100%.
Analog signals are distorted in the channels and in mediums due to noise, crosstalk and other interferences, mechanical issues.
Of course, if we compare an original analog sound with analog one, restored from digital record of this original, we get a difference.
But if we compare the original and its analog recording, the difference will be more.
Both digital and analog are music storage ways. Both lose part of information of an acoustic waves before microphone.
But, digital losses are lesser in modern audio systems.
Digital data (audio, video and others) may be stored and transferred without losses. And the losses may be detected and fixed via special math.
Read more...
If external DAC (stand-alone or inside AV-receiver), is better that built in digital-to-analog converter, digital output is preferable.
Read more...
In general, optical mediums, that contains digital audio, should sound better than analog recordings. However, it is not so simple.
Read more...
Digital signal is more reliable to distortions due to special coding and recovering abilities. See more...
Digital signal may allow checking data integrity and, sometimes, restoring of broken information. See more...
Digital audio output send audio signal in a digital form. It used on different devices like digital audio CD/SACD/DVD-players, AV-receivers, TV.
Digital audio output in TV is interface to send digital sound to AV-receiver or DAC. Such interfaces may have types: HDMI, SPDIF optical or coaxial.
If you have alternative between digital and analog output at a device (CD-player, as example), it means, that the device have built-in DAC (digital-to-analog converter). You can use its analog output.
Alternatively, you can use digital output. If external device (DAC, AV-receiver) have better DAC, it can give sound advantages.
But, some digital outputs may have sample rate/bit depth limitations by technical or other issues. And external DAC will not used fully.
Also, an external DAC may have resolution limitations on its own.
So before the decision, learn:
We can see, that digital signal is a virtual math model of an analog signal in the digital domain. I.e., despite, the digital signal consists of separate samples, and the first wish is to connect lines in the "stairs", a discrete digital signal is continuous waveform math point of view.
If we ideally upsample digital signal to the infinite sample rate (interpolate/restore virtual samples between real samples), the digital waveform will be uninterrupted and equal to analog one.
The connection line is the same to steel ruler, that fixed in the points of real samples.
The same way is restoring a digital signal to analog form. There we get a continuous signal for any sample rate.
It happens, because the analog filter of a DAC is ideal interpolator, that smoothly and correctly connect digital samples.
We can see it in the picture above.
In the article, we'll consider what is better sound quality: digital or analog?
Sound quality has 2 definitions:
Both audio types have the same source - analog musical or another signal.
However, the next stage is storing and reconstruction to played back an analog signal. And the stage causes different losses for various audio systems. We can compare an analog vs digital system at this stage only.
Definition #1 (distortion level) depends on the sensitivity, accuracy, and precision of a measurement tool.
Definition #2 (ear test) can be partially free from subjective biasing due to big numbers. Measurements are provided via blind tests. However, it is not a simple procedure.
Distortion-level sound quality and ear-test sound quality may be uncorrelated in some cases. It happens, because various patterns of the measured distortions may impact differently to ears and brain.
EXAMPLE:
Even, a higher level of distortions is not meant better-perceived quality in ear test.
There are tricks in mixing and post-production, that can improve the sound, despite altering the original sound.
The author would compare the original sound and sound, containing some "nice-sounding" distortions, with "tomato juice" and "tomato juice with pepper and salt".
A few salt and pepper may improve the juice taste. No seasoning is "dry sound". Too much seasoning is "dirty sound".
Different improvements may be there. They depend on personal taste.
May be natural notes of original juice, hidden by the seasoning, are preferable for some people.
Look an infographic of analog vs digital audio quality comparison, where essential factors are shown.
We can see, that before the original acoustical signal comes to medium, it pass through electro-mechanical (microphone) and electrical devices. To throw out factors, that common for all systems, below we'll consider original electrical signals directly before and after the mediums (tape, vinyl, digital unit) and its conversion devices (tape heads, vinyl recorder and player cartridge, including circuits, ADC, DAC).
Let's consider an analog audio system.
A vinyl recording is produced mechanical way (see simplified workflow):
In each of the stages, some mechanical deviations appear.
"Wow and flutter" effect is caused due to rotation speed deviation or magnetophone tape stretching.
The speed deviation alters the time of playback of the same tape or groove length. It causes frequency bias, that leads to total waveform distortions.
The speed deviation happens during recording and playback.
Mechanical processes are sensitive to environmental temperature. As an example, it can cause an altering of play in bearings. The play may cause variations of "wow and flutter" effect.
With wear and tear of the mechanical parts, audio parameter precision and accuracy degrade. It can cause increasing of distortions.
Also, when the tape edges are bound together, during the tape unwind, the tape is ripped. The oxide from the backing is pulled, that create pinholes. Read details in the article: "Saving the Bob Dylan Archive from Adhesion Syndrome"
When a tonearm playback LP disk grooves at different distances from its center, angle of a stylus tip relative the groove is changed. It alters playback characteristics. The issue may be partially solved via tangential tonearm, that moves to the LP center with constant angle. But any mechanical device has parameter deviations, that impact to sound.
The main issue of analog mediums is noise.
Vinyl disks are sensitive to dust. It causes temporary mechanical distortions of the disk surface. The disk cleaning is very desirable.
A tape has noise due to magnet layer physical processes. There are some noise reduction systems. But it can't extremely reduce noise, like digital systems.
Tapes and vinyl records also wear and tear with storage and/or using time. Vinyl disks may be warped, scratched. Tapes may be stretched, well-worn and degaussed.
Electronic elements (transistors, tubes, others) of analog musical devices wear with time. Values of the element's parameters may come out of allowable range and lead to higher non-linear distortions and other.
Electronic components are a source of analog noise.
Degrading of electrical, magnetic, mechanical parameters also touch cartridges of LP players and tape heads.
In general, analog mediums put more loudness limitations for recorded audio stuff, than digital ones.
EXAMPLE:
LP has a limitation on an album track width (time length) at the disk side. A mastering engineer can control the record grooves and the track width by volume altering. But, when several tracks are located on one side of the disk, a similar volume should be provided.
Digital recordings have lesser issues with loudness and its normalization. Because even popular 24-bit formats have a wide dynamic range for recording. 32- and 64-bit float point formats have extremely low noise level and overload tolerance for audio processing.
Output analog signal is limited by the maximal level. Close to the level, non-linearity is increased dramatically and output signal can't grow above the level despite the input signal is rising. So, the output waveform is distorted.
It's called "overload". And such output waveform distortion is called "clipping".
At the picture, we can see the smooth transient-clipping zone in point A. Such transient type leads to "soft" clipping of output waveform.
As rule, analog signal cause soft clipping - linearity is decreased smoothly.
Let's consider a digital audio system.
All its parts, that conduct or convert, from/to analog signal, cause phase, frequency, non-linear distortions.
Electronic elements of analog parts of the digital musical devices wear with time like the same items in analog systems.
Stability of the clock signal of ADC (analog-to-digital converter) and DAC (digital-to-analog converter) is very important. There are deviations:
The issue may be considered as similar to "wow and flutter". But, currently, digital-system clock stability is very high, comparing speed deviation of analog systems, in most cases.
Digital music systems may not be pure "digital". Because humans can receive analog acoustic waves only. But, in the author opinion, we have some trend to reduce the analog part share in audio systems.
D-class power amplifiers, based on sigma-delta modulation (like DSD), exists a long time. But they still not popular enough despite these amplifiers allow to get rid of digital-to-analog converters (DAC) and are fine compatible with the DSD.
If no additional processing, overload in digital format cause loud click due to positive value transformed to negative one momentarily.
To avoid digital overload, analog compressing and/or level limiting of an input signal of ADC (analog-to-digital converter) may be applied.
Special processing provides traditional clipping into the digital domain. The clipping may be:
Overload may be avoided via analog compression of input analog-to-digital converter's signal.
When overload might occur into the digital domain, float point formats may solve the issues. The formats provide 0 dB level for 1.0.
If overload happens (1.1, as an example) it will not be clipped. At the final stage, the overloaded peak level may be normalized back to 0 dB.
Digital versus analog audio systems
Ana |
Tape | Vinyl |
Digi |
Digi |
Digi |
Digi lossy |
|
---|---|---|---|---|---|---|---|
Noi |
De |
about -60 dB | about -70 dB |
de |
de |
depend on bit depth and ADC, -96 dB and bet |
de |
Clicks |
de |
no | yes |
de |
de |
almost im |
|
Me |
no | yes | no | ||||
Jitter (fast de |
FM/AM-mo |
no |
yes; as rule, de |
||||
Ex |
no, except phy |
in |
sig |
no, except phy |
in |
no, except impact to a me |
|
Tem |
im |
yes |
impact to radio wave pro |
almost no | don't impact to files | ||
Re |
no |
yes with spe |
|||||
Im |
yes | ||||||
Time de |
no, except cab |
yes | no, except cables | insignificant |
no, except phy |
||
Sen |
only wires | yes | only wires |
may be com |
depend on device safety, may be com |
What is most recommended: analog and digital audio outputs? Before answer the question let's consider the scheme.
We can't see a significant difference between paths from the digital mediums to analog devices at upper (analog output of the digital device) and lower (digital output) parts of the drawing.
Factors impact to sound quality:
Connection via digital vs analog outputs
Digital connection | Analog connection | |
---|---|---|
Interruptions, pauses, pops | long cable length, damaged cable | damaged cable |
Sound transparency | doesn't impact to sound quality | may impact to sound quality |
DAC | impact to sound quality |
Analog connector types
Type | May be unbalanced | May be balanced | Description |
---|---|---|---|
RCA | yes | no | As rule, for home applications. |
TRS (jack) | yes | yes | Implementations: jack, mini-jack, micro-jack. As rule, for music production and performance applications. |
XLR | yes | yes | As rule, for music production and performance applications. |
In general case, no difference for sound quality between these connectors. However, a balanced connection may improve the signal-to-noise ratio. It may be important in hard electromagnetic interference environment and for long cables. A necessity in balanced connection is defined experimentally by ears or measurements at an audio device output.
Digital connection types
Type |
To |
Sup |
Max. length, me |
Des |
---|---|---|---|---|
SPDIF |
about 9.2 (audio data without meta-in |
24 bit/192 kHz, ste |
10 (op |
|
USB | up to 3 200 (USB 3) |
De |
3 (de |
|
HDMI | up to 18 000 (HDMI 2.0) |
Sam |
15 (de |
|
Thun |
up to 40 000 (V3) |
Via Dis |
up to 2 (de |
Meta-information is data about transmitted information.
Below we can see throughputs for different audio resolutions. It should be lesser than the maximal throughput of a digital connection with an accounting of the meta-information.
Throughputs of audio resolutions
Bit depth, bit | Sample rate, kHz | Channels | Bitrate (uncompressed), Mbit/s | |
---|---|---|---|---|
1 | 24 | 96 | 2 | 4.608 |
2 | 24 | 192 | 5.1 | 27.648 |
3 | 32 | 384 | 2 | 24.576 |
4 | 32 | 768 | 2 | 49.152 |
5 | 1 | DSD64(2.8 MHz) | 2 | 5.645 |
6 | 1 | DSD64(2.8 MHz) | 5.1 | 16.934 |
7 | 1 | DSD1024(45.2 MHz) | 2 | 90.317 |
Read more about audio interfaces
Most popular audio editing are analog mixing and gain control. The mixing and gain control are implemented in analog circuits and components (transistor, tubes, operational amplifier, and others).
Digital audio is a form of analog audio coding. From a math point of view, no difference between analog and digital gain control. Look at the formula of gain control for both digital and analog audio:
output(t) = input(t) * gain,
where t is time.
At first glance, digital values are discrete and cause precision losses.
But, a digital signal is not "stairs". We should consider it as a continuous function with some bit depth, that causes a certain noise level.
Using float point math to gain control allows to apply gain and mixing transparently for audio stuff.
It is necessary to remember, if we multiply 16-bit integer values, the result should be placed to a 32-bit integer value and truncated back to 16 bit. After all processings, dither is recommended.
Result of multiplication of two 32-bit float values is 32-bit float value.
Analog gain control may be implemented in different ways. As an example, a switchable resistor kit can provide dB-scaled gain control.
1...2 dB-step implementation is significantly easier in float point formats, comparing 16/24-bit integer variables.
Using electronic components in gain control circuits cause non-linear distortions, noise, temperature instability.
But, for level adjusting of:
analog gain control may be used.
Digital vs analog mixing and gain control
Analog mixer | Digital mixer | |
---|---|---|
Noise level | about -60...-130 dB or worse (depend on electronic components, construction, signal sources) |
about -180 ... -200 dB and better (for float point formats) or worse for 16/24 bit |
Non-linear distortions | yes | no |
Overload | yes | may be avoided (depend on the implementation and sources) |
Resume:
Except for artistic purposes, analog mixing is not recommended for high fidelity music production.
Digital gain control, when it applied properly, is more precise when analog one and cause lesser issues.
In, some cases, we can apply analog gain control only.
Analog vs digital technologies in music and audio production, recording
Analog | Digital | |
---|---|---|
Noise level | higher | lower |
Parameter stability of full production system | lower | higher |
Mechanical distortions | yes | no |
Distortions in recording | higher | lower |
Distortions in editing | higher | lower |
Pre-ringing of filters | no | yes; no for minimum-phase filters |
Post-ringing of filters | yes | yes, approximately double post-ringing energy for minimum-phase filters, comparing linear filters with pre- and post-ringing |
Temperature, wear and tear, interference stability | lower | higher |
"Artistic" distortions and "analog" sound (vinyl, tape, tubes, analog compression, synthesizers,...) | yes | software emulation |
Project saving | high cost, sophisticated electro-mechanical drives of sliders with same positioning error | yes |
Precise/exact project saving | no | yes |
Lossless backup | no, time degrading by different reasons | yes |
Lossless copy (including factory) | no | yes |
Copying protection | no | optionally |
Distribution | only physical delivery | network, memory devices, physical delivery |
In the general application, a speakers are pure analog device. There are drivers with passive electrical components.
But some speakers contains amplifier(s) inside. Teoretically, it allow:
The amplifier may be digital.
The crossover may be implemented in the digital domain, that give more potential abilities, than crossover on the active components. But, DSD-input signal demands additional tricks for editing (EQ, gain contlrol). If the editing is done properly, it is almost transparent for the audio signal.
Optical connection is digital one. If you have better external DAC, that is better than built in mobile phone, DAP, etc., optical connection to the DAC may give sound-quality advantages.
Read more...
December 25, 2022 updated | since July 12, 2019