Digital audio works by taking snapshots of an analog sound wave at regular intervals, recording the amplitude at each moment, and storing those values as numbers. Sample rate is how frequently those snapshots are taken, measured in samples per second (or kilohertz, where 1 kHz = 1,000 samples per second). Play all those values back in sequence at the same rate and you reconstruct the original wave — which is the core idea behind pulse-code modulation (PCM).
The Nyquist theorem
There's a hard mathematical limit on what sample rate can capture. The Nyquist theorem states that to accurately reproduce a frequency, you need to sample at least twice as fast as that frequency. To capture a 10 kHz tone, you need a minimum of 20,000 samples per second. To capture the top of human hearing at 20 kHz, you need at least 40,000 samples per second.
This is exactly why the CD standard landed at 44,100 Hz (44.1 kHz): it covers the full human hearing range (up to ~22 kHz) with a small safety margin, and it fits a certain amount of audio onto the physical space of a disc. It wasn't arbitrary — it was the minimum sufficient sample rate for full-bandwidth audio, chosen deliberately.
If you try to represent a frequency higher than half the sample rate, the digital system can't do it accurately and produces a false, lower-frequency artifact instead. This is called aliasing — it sounds like a weird, metallic ringing that has no relationship to the original audio. Anti-aliasing filters in audio converters prevent this by rolling off frequencies above the Nyquist limit before sampling.
Common sample rates and when they appear
- 44.1 kHz — the CD standard, and the most common rate for music distribution. Every streaming platform, most downloaded files, and the vast majority of DJ libraries. Perfectly sufficient for a dance floor.
- 48 kHz — the standard for video and broadcast audio. If you're watching a film or a TV show, the audio is almost certainly at 48 kHz. Some DJ controllers and audio interfaces default to this rate.
- 96 kHz and above — high-resolution audio, used in professional recording and production. Captures frequencies well beyond human hearing, which is useful during processing and editing (where math on the audio can create new frequencies) but largely irrelevant for final playback through a PA.
Why DJs should care
For most DJ purposes, sample rate is a decision you make when building your library and setting up your software, not something you adjust on the fly. A few things worth knowing:
- 44.1 kHz is enough. A dance floor PA — even a very good one — is not going to reveal any difference between a 44.1 kHz file and a 96 kHz file. The room acoustics, the crowd, and the distance from the speakers dwarf any theoretical advantage of a higher sample rate.
- Sample rate mismatches cause problems. If your audio interface is set to 48 kHz and your software is outputting at 44.1 kHz, the system has to resample on the fly — which can introduce artifacts or, worse, cause audio to play back at slightly the wrong pitch or speed. Match your software and hardware settings.
- Lossy formats don't store raw samples. MP3 and AAC files use perceptual coding to compress audio — they don't store PCM samples at all in the same way. The sample rate of a lossy file describes the rate it's decoded to on playback, not a raw snapshot rate.
Sample rate and quantization noise are the two axes of digital audio quality. Sample rate controls the frequency ceiling. Bit depth (the precision of each amplitude measurement) controls the dynamic range and noise floor. Most music is 16-bit/44.1 kHz — and for playback through a PA, that's more than adequate.