A recording sampled at 16 kHz can represent sound frequencies up to 8 kHz, half the sampling rate.

Checked against its source on October 17, 2026

What it measures: Sampling at 16 kHz is the rate the speech recognition literature treats as the wideband standard for dictation, while narrowband telephone audio is sampled at 8 kHz.
How it is calculated: 16 kHz / 2 = 8 kHz, the Nyquist limit: the highest frequency a sampled signal can represent
What it does not show: This is the sampling theorem applied to the 16 kHz figure. It says nothing about how much speech energy lies above 4 kHz.

Calculated from