Digital audio is the digitization of the analog audio signal. According to Nyquist's sampling theorem, the sampling frequency must be at least twice the highest frequency of the signal to be sampled. Assuming a highest audio frequency of 20 kHz, the sampling signal must be at least above 40 kHz for good audio quality.
Different optical storage media use different sampling rates, which is also related to storage capacity. For example, audio signals on a CD-A are sampled at a rate of 44.1 kHz, while DVDs use both 44.1 kHz and 96 kHz. MPEG-1 audio with PCM uses sampling rates of 32 kHz, 44.1 kHz, and 48 kHz, and MPEG-2 uses sampling rates of 16 kHz, 24 kHz, 32 kHz, 44.1 kHz, 48 kHz, and 96 kHz.
In addition to the sampling rate, quantization plays a significant role in audio quality. For example, if an audio signal is quantized at 4 bits per sample, this results in 16 levels and a maximum level deviation of 6.25%. With 8 bits, the maximum level deviation is still 0.39% and with 16 bits only 0.0015%.
At the same time, the file size and memory requirements increase proportionally with the sampling rate and sampling depth. For example, one minute of uncompressed stereo digital audio on a CD-DA digitized at a sampling rate of 44.1 kHz and 16 bits on two channels requires 10.584 megabytes( MB) of storage. On a DVD, on the other hand, the storage requirement would be 34.56 MB per minute at a sampling rate of 96 kHz, quantization of 24 bits on two channels. With reduced bandwidth and a lower sampling rate, for example, a CD-ROM/XA in Class C with a sampling rate of 18.8 kHz and 16 bits for speech quality in monophony comes to a volume of 2.27 MB per minute.