Digitization of audio is the process of changing an analog continuous sound wave into a discrete digital signal, which can be stored and processed by a computer or a digital device. The process is done through a series of three main steps: first step sampling, then quantization, and finally encoding. During the first part, sampling takes snapshots of the state of the analog waveform’s amplitude at certain usual times. The frequency of these snapshots is called the sampling rate; a higher rate captures more detail of the original wave.
Next, quantization measures the amplitude of each sample and assigns it a numerical value from a predefined range. The precision of this measurement depends on the bit depth, which determines the number of possible values available. A higher bit depth allows for a more accurate representation of the original amplitude. Finally, encoding converts these numerical values into binary code (sequences of 0s and 1s) that digital devices can understand and store.
Although digitization gives many advantages, such as storing, copying, and manipulating easily, it also has tradeoffs. The change process does not work well, and some information from the original analog wave gets inevitably lost. This can be the cause of a very slight distortion called quantization noise that can be observed at lower bit depths. Some audiophiles also say that digital audio can be less “warm” or “natural” compared to analog recordings because of the discrete nature of the digital signal. It is very important to choose appropriate sample rates and bit depths to minimize these downsides and achieve a very high-fidelity digital representation of the audio.