AUDIO STREAMING

Analog to Digital

When a digital audio, video or image file is originally created, usually a wav file on audio, its format is PCM (Pulse-Code Modulation). Figure V is an overall illustration of the digital audio path from input to output, in which the digital signal processing is the block where codec encoding and decoding take place. PCM is the basic digital representation of a sampled analogue signal, in which the magnitude of the analogue signal is sampled regularly at uniform intervals in frequency (sampling frequency), with each sample being quantized and digitally represented with the highest precision possible (in audio usually 16 bit/sample). That quantization introduces quantization noise. Quantization noise is the difference between the analog signal and the digital representation, PCM, and arises as a result of the error in the quantization of the analogue signal, as it is illustrated in the figure.

With each increase in the bit/sample level, the digital representation of the analogue signal increases in fidelity, and the quantization noise becomes smaller. Since this method is the one that mostly keeps the integrity and recreates the original signal, it is used in devices and applications that doesn’t have computational or file size problems, such as Blu-Ray, DVD and Audio CDs. In applications that deal with the delay coming from the signal processing, i.e. that have requirements that doesn’t allow them to lose the time needed for processing or that can’t deal with the file size that results from a PCM encoding, compression has to be used and in some cases a selection of the original signal information has to be made so that the less perceived information can be discarded while affecting as few as possible the original signal perceived quality. The table  shows PCM encoding levels according to each range of frequencies and bits/sample.
 

	Frequency (Hz)	Sampling Rate (kHz)	bit/sample (PCM)	PCM bitrate (kbit/s)
Speech (telephone)	300-3400	8	8	64
Speech (wideband)	50-7000	16	8	128
Audio (medium band)	10-11000	24	16	384
Audio (wideband)	10-22000	48	16	768

Speech and Audio Sampling
Source: Audio and Video Communication: Theoretical Classes Presentations

Looking at the table, wideband audio has a PCM bitrate of 768 kbit/s with a single channel. The CD audio rate is higher, because it includes two channels and can be given by:

CD Audio
2 Channels (Stereo)
2 * 22 kHz Frequency channel
2 * 22 kHz * 2 Sampling rate
44 kHz * 2 * 16 = 1.41 Mbit/s

CDs data rate is about 1.5 Mbit/s and that’s the reason why no compression is needed. However, CDs are not the only technology available and internet and wireless applications are not compatible with such high data rates. In streaming, for example, this bitrate is too high to guarantee a good audio experience to all users, due to their different internet connections. Codecs are the tools that allow encoding the PCM format into a format with the compression that keeps the ratio: perceived quality per file size, higher as possible after decoding, while making viable the file transmission or live broadcasting.

As in any other system there are advantages and disadvantages that come along. From the advantages that come from converting from analog to digital one can emboss:

• easier to process, store and transmit;
• mobility;
• random access;
• fast forward and fast reverse;
• editing and mixing.

Among the disadvantages one should highlight:

• makes piracy easier;
• quality losses even in PCM, despite unperceptive.

To the end user there is no doubt digital formats are much more friendly and that’s why the digital succeeded over the analog.