Digital Audio – Quality Issues
Relatively recently, the concept of “multimedia” was included in our discourse, and now the computer is increasingly used as an entertainment center. Now the computer is forced to reproduce the sound that exists in it in the form of numbers.
Just as some connoisseurs of sound argue about the advantages of “tube” sound over “transistor” sound, there is an endless debate about which is better: digital or analog sound. Let’s try to figure it out.
For our ears, sound is air vibrations with a frequency of 20 Hz to 20 kHz, and the upper limit depends on age: in children it is 22-24 kHz, and in old age the perceived frequency decreases, up to 8 -12 kHz.
The frequencies of the indicated limits are perceived as vibrations, higher, they are not perceived by a person.
However, not all the detection bandwidth is used with the same intensity, so speech is clearly perceived in the range of 500 to 3500 Hz. But for listening to music, this is not enough. Ideally, the reproduced sound should not differ from the sound field of the microphone. That is, the recording and playback equipment must not introduce distortions within the limits of human perception.
The sound we hear from the speaker is electromechanically converted to an electrical signal during recording; then there is the amplification and processing of the analog electrical signal; analog to digital conversion; digital signal processing; frequency correction; recording procedure.
After the digitized sound is stored and transmitted. During playback, digital signal processing occurs first; follows the conversion from digital to analog; analog signal processing and amplification; electromechanical conversion to sound vibrations.
All of these procedures introduce their own distortions. The process of recording and sound processing takes place, as a rule, on studio equipment, which performs much better than home audio equipment. Therefore, although there are distortions, they are significantly less than the distortions introduced by home equipment at the playback stage. With amateur sound recording, errors appear in the recording stages.
The electromechanical conversion produced by the studio microphone produces a very weak signal that needs amplification.
Even in the ideal conditions of a professional recording studio, due to acoustic noise, the dynamic range of recorded music can be narrower than that provided by 16-bit audio.
When recording from multiple microphones, the signal is necessarily processed: channel volume levels are selected, noise is filtered, etc. Furthermore, the dynamic range of the signal is reduced, which leads to a significant increase in noise. But without this procedure, it would sound unsatisfactory when playing back the recording on a home computer.
The sound path has its own distortions, which can be divided into three groups:
1. Linear distortions are caused by the amplitude-frequency characteristic of the sound path and are a change in the ratio of the amplitudes and phases of various frequency components. Frequencies that were originally missing from the signal do not appear.
2. Non-linear distortion: a change in the shape of the original signal, which leads to the appearance of frequencies that are absent in the incoming signal, but depend on it.
3. Interference: the appearance of strange frequencies in the sound path that are not associated with the useful signal. Interference appears, for example, by electromagnetic interference, penetration into the sound path of the frequency of the supply voltage, etc.
However, all these distortions occur only in analog circuits (hence speculation about the frequency response of a digital output makes specialists smile). But don’t forget about the superficial defects of CDs, DVDs, and other optical storage media that store sound, leading to data loss.
The digitization of the signal is also associated with a lot of distortion, but first let’s look at the difference between analog and digital signals.
In an analog signal, the voltage changes smoothly over time, the signal is continuous. The digital signal is discrete, its value changes instantly. Furthermore, discretion is manifested in both frequency and amplitude region. Any change in signal value is sampled, and as a result, the values are rounded to the nearest whole number.
