Sample Rate
Samples captured per second; the ceiling on the frequencies a system can represent
Related
Sample rate is the number of times per second an audio signal is sampled, measured in hertz. At 44.1 kHz the system takes 44,100 points each second; at 48 kHz it takes 48,000. Denser points allow higher frequencies to be captured, and make larger files.
The Nyquist theorem states that capturing a frequency f requires a sample rate above 2f. At 44.1 kHz the theoretical ceiling is 22.05 kHz, which covers the 20 kHz limit of human hearing. 48 kHz leaves more margin and is common in video and broadcast. Rates above 96 kHz help mainly in recording and processing, not in playback.
The usual mistake is to assume a higher rate always sounds better on playback. Frequencies above the hearing limit are inaudible; the value of high rates lies more in reducing aliasing in conversion and non-linear processing, and in leaving headroom for later work.
Related → Bit Depth
Sources
- 采样率、奈奎斯特频率与常用工程采样率，参 The New Grove Dictionary of Music and Musicians, "Sound recording" 与 Modern Recording Techniques（Huber & Runstein）；本文为原创表述
Used for fact-checking only; all prose is original.