Analog-Digital Processing
A digital signal is obtained from analog or is directly synthesized into digital (in electric musical instruments).
Converting from analog to digital involves two basic operations: sampling and quantizing. Discretization is the replacement of a continuous signal with a series of samples of its instantaneous values taken at regular intervals. According to the Kotelnikov-Chenon theorem, a discrete signal can be completely restored later, as long as the sampling frequency is at least twice the upper frequency of the signal spectrum. The samples are then quantized according to level: each of them is assigned a discrete value closer to the real one. The precision of quantization is determined by the bit width of the binary representation. The higher the bit depth, the more quantization levels (2N,
The audio CD format has a sampling frequency of 44.1 kHz and 16 bits. This gives 44 thousand samples per second, each of which can take one of 216 = 65536 levels (for each of the stereo channels).
In addition to the 44.1 kHz / 16-bit format, others are used in digital recording. Studio recording is generally done in 20-24 bit, then the data is converted to audio CD by recalculation. The extra bits are then discarded or (better) rounded, sometimes pseudo-random noise is added to reduce quantization noise (dither).
The most advanced custom audio formats are DVD Audio and Super Audio CD (SACD). DVD Audio adopts the MLP lossless data compression algorithm developed by Meridian. And SACD, unlike other formats, does not use pulse code modulation (PCM or PCM), but one-bit encoding of the DSD (Discrete Pulse Width Modulation) stream. SACDs come in single or double layer (hybrid) discs with a normal CD layer.
The most popular audio medium today is compact disc, despite certain limitations in sound quality seen by audiophiles. The reason for them is in the low sample rate: for an accurate reconstruction of signals near the upper limit of the audio range, a filter that is not physically workable is needed (its impulse response covers the negative time area). This is compensated to some extent by digital filtering with higher sampling and bit depth. The data on the disc is redundantly encoded (Reed-Solomon code) to ensure smooth playback in real time.
Broadband communication is required for digital audio transmission, especially for uncompressed high definition multichannel transmissions.
Figure: 1. Digitizing an analog signal and obtaining digital samples on CD Audio and SACD (right)
DIGITAL AUDIO TRANSMISSION
The communication lines for digital audio transmission can be cables, optical lines, and overhead radio.
For the transmission of PCM signals over wired lines, AES / EBU (balanced, coaxial), S / PDIF (unbalanced coaxial) interfaces have been developed, which provide transmission of various signals (clock frequency, digital word rate, channel data) over a cable. Inside the devices, these signals are transmitted separately, at the output of the transport mechanism they are encoded and at the input of a digital-to-analog converter (in two-block systems) they are separated again in a digital receiver.
Typically, a high-quality coaxial cable is used for digital audio transmission. There are also S / PDIF converters for fiber optic lines: AT&T ST and Toslink (the latter is standard in consumer equipment). And also, for the use of twisted pairs in Ethernet cable networks. The medium of distribution for compressed audio in the form of archived files is the Internet.
Like any digital signal, digitized audio is distributed and switched by special devices: distribution amplifiers, matrix switches, and conventional.
There is one factor that negatively affects digital signals, and often negates almost all of the advantages of digital audio over analog, including the ability to repeatedly copy, stream, and archive programs without any loss of quality: we are talking about jitter. Jitter is jitter, or the uncertainty of a transition from 0 to 1 and vice versa.
