5 min read
Why 48 kHz is the video standard, and what happens when you ignore it
48 kHz is not an arbitrary number. It divides cleanly into frame rates, and that is why every video spec asks for it.
Every broadcast spec, every cinema deliverable and every camera manual lands on the same figure: 48 kHz. Music, meanwhile, standardised on 44.1 kHz. The split is historical and mathematical, not a quality judgement.
Frame rates are the reason
At 48 kHz there are exactly 2000 samples per frame at 24 fps, 1920 at 25 fps and 1600 at 30 fps. Whole numbers make editing, conforming and frame accurate cutting predictable. At 44.1 kHz you get 1837.5 samples per frame at 24 fps, which is not a whole number of anything.
Where 44.1 kHz came from
44.1 kHz was chosen for the compact disc because it fit the video tape based recorders used to master digital audio in the late seventies. It stuck for music and never made sense for picture.
What goes wrong with a mismatch
- A 44.1 kHz file interpreted as 48 kHz plays roughly 9 percent fast and rises in pitch.
- Real time conversion inside an NLE costs CPU and can drift on long takes.
- Repeated conversions between rates accumulate small artefacts.
What about 96 kHz
96 kHz is useful when material will be heavily pitched down, as in sound design for creatures or vehicles. For dialogue and delivery it doubles file size for no audible gain, and it still conforms cleanly to 48 kHz because the ratio is exactly two.
Practical rule
Shoot 48 kHz, record 48 kHz, edit 48 kHz, deliver 48 kHz. Convert anything that arrives at another rate once, at ingest, with a proper resampler.