Sample Rate and Bit Depth: What 44.1 kHz Means

Sample Rate and Bit Depth: What 44.1 kHz Means

44,100 samples per second is a strange number, until you learn it was engineered around 1970s video recorders: three samples per TV line, 245 lines per field, 60 fields per second. The CD inherited it, and we have typed it into audio settings ever since.

Sample rate: how often the waveform is measured

An analog waveform is continuous; digital audio stores snapshots. The sample rate is how many snapshots per second. The Nyquist theorem says any frequency below half the sample rate can be captured faithfully, so 44.1 kHz covers up to 22.05 kHz, above most adults’ hearing (15–20 kHz).

Use Sample rate Bit depth Stereo data rate 1 minute
CD, music files 44.1 kHz 16-bit 1,411 kbps ~10.6 MB
Video, broadcast 48 kHz 16 or 24-bit 1,536 / 2,304 kbps ~11.5 / 17.3 MB
Studio recording 96 kHz 24-bit 4,608 kbps ~34.6 MB

48 kHz is more than music needs; it exists because it divides evenly into video frame rates, keeping audio and picture in sync without fractional samples — why cameras and editors default to it.

Bit depth: how precisely each sample is stored

Bit depth sets how many amplitude values each sample can take, and therefore the dynamic range: every extra bit adds about 6 dB. 16-bit gives a theoretical 96 dB; 24-bit gives 144 dB, more than any real room or microphone can use. The practical value of 24-bit is headroom while recording and mixing — set levels conservatively, raise them later, and never hear quantization noise.

When you reduce 24-bit to 16-bit, apply dither: a tiny amount of shaped noise that randomizes the rounding error. Without it, quiet fades can turn gritty.

Where the numbers stop mattering

192 kHz files sell well. On playback they reproduce nothing you can hear; their honest use is recording and heavy processing, where extra samples reduce artifacts from pitch shifting and filtering. For delivery, 16-bit at 44.1/48 kHz is plenty.

What does matter is resampling. The audio converter decodes in a browser audio context configured for 44.1 kHz and writes WAV as 16-bit PCM, so a 48 kHz video file comes out resampled to 44.1 kHz — harmless for a music copy, unnecessary for a video workflow. Channel count is preserved, so a mono recording stays mono.

Record high, deliver sensibly

Capture at 48 kHz/24-bit if the audio might meet video; export 16-bit/44.1 kHz for music files and 16-bit/48 kHz for video. And do not upsample after the fact — going from 44.1 to 96 kHz adds file size, not information.