What Is Sound?

Sound begins with movement. Pluck a guitar string, hit a drum, or speak, and something starts to vibrate. That vibration pushes and pulls the air around it. The result is a travelling pattern of pressure changes called a sound wave.

Your ears turn those pressure changes into signals your brain can understand. Recording equipment follows a different path. A microphone turns the pressure wave into a changing electrical signal. An audio interface or sound card then measures that signal and stores the measurements as numbers.

Those numbers form digital audio. Music software can edit them, compress them, analyse them, and use them as input for AI music systems.

Amplitude and frequency: size and speed

Two basic properties make a sound wave easier to understand:

1. Amplitude describes the strength or size of the signal. On a waveform, a larger vertical movement means a larger recorded amplitude.
2. Frequency describes how many times the wave repeats each second. It is measured in hertz, written as Hz. A 440 Hz tone repeats 440 times every second.

Frequency is closely linked to pitch. Faster repetition is usually heard as a higher pitch. Amplitude is linked to signal level, but a larger digital amplitude does not always mean that a sound will feel equally louder to a listener.

Sample rate: how often do we measure?

Think of filming a spinning wheel. One photograph tells you very little, but thousands of photographs can capture the movement. Digital audio works in a similar way. A recorder takes repeated measurements of the incoming signal.

A sample rate of 44.1 kHz means the signal is measured 44,100 times per second. A rate of 48 kHz means it is measured 48,000 times per second.

The sample rate also limits the highest frequency that can be represented. To reproduce a frequency, the sample rate must be at least twice as high. This is why a 44.1 kHz recording can represent frequencies up to about 22.05 kHz.

A higher sample rate creates more measurements and more data. Its main technical effect is to give the recording system a wider frequency range.

Bit depth: how precise is each measurement?

Sample rate tells us how often the signal is measured. Bit depth tells us how many possible numerical values are available for each measurement.

1. 8-bit audio provides 256 possible values.
2. 16-bit audio provides 65,536 possible values.
3. 24-bit audio provides 16,777,216 possible values.

More available values give the system a greater dynamic range and a lower theoretical noise floor.

One way to picture this is as a staircase. A low bit depth gives the staircase fewer, larger steps. A higher bit depth gives it many smaller steps, so the stored value can follow the original signal more closely.

What is an audio channel?

An audio channel is one separate stream of measurements.

A mono recording has one channel. A stereo recording normally has two, one for the left side and one for the right. The two stereo channels can contain slightly different signals, which lets producers place sounds across a wider listening space.

Channel count is separate from sample rate and bit depth. A recording can be mono or stereo at 44.1 kHz, 48 kHz, or another supported sample rate.