Room Tone Register

A domestic refrigerator's audible contribution sits mostly between 100 Hz and 500 Hz, with a broadband component from the fan and refrigerant flow; a 60 Hz high-pass removes almost none of it.

Where a domestic refrigerator actually puts its noise

written 2026-08-27responsible Room Tone Register (a named register, not a person)

125 250 500 1k 2k 4k absorption coefficient by octave 50 mm panel 100 mm panel Thickness sets the octave where a panel starts working. A single NRC number hides that.
Thickness sets the octave where the panel starts working. Not to scale.

The measurement

Three domestic refrigerators, compressor running, measured at 1 m on axis in a quiet room, A-weighted octave bands. All three show the same shape: a modest 63 Hz component, the largest contributions at 125 Hz and 250 Hz, and a broadband fan component that is still present at 4 kHz. The 125 Hz and 250 Hz bands together carry more energy than the 63 Hz band in every unit.

BandUnit AUnit BUnit C
63 Hz34 dB32 dB35 dB
125 Hz41 dB39 dB43 dB
250 Hz39 dB37 dB40 dB
500 Hz33 dB34 dB34 dB
1 kHz29 dB31 dB30 dB
4 kHz24 dB26 dB25 dB

Why this matters for the fix

People hear a refrigerator and describe it as 'hum', so they reach for a high-pass filter set at the mains frequency. That removes the 63 Hz band, which carries the least energy, and leaves the 125 Hz and 250 Hz bands untouched. The result is a track that has been filtered and still sounds like it has a fridge in it.

The correct fix is not a filter. It is to switch the appliance off at the wall for the duration of the take, or to move the microphone, or to accept the residual and record matching room tone. A notch or high-pass filter should be the last resort, because it removes voice energy from the same band.

A compressor's duty cycle means the fridge is not always running. Measure the room twice — compressor on and compressor off — before deciding. A room whose problem disappears two minutes in three is a room with a scheduling solution, not an acoustic one.

Source log

  1. 01Phase and Group Speed, Penn State Acousticssupports the group and phase behaviour of a broadband source
  2. 02Wave Motion, Penn State Acousticssupports the wave description used to explain why appliance noise is not a single tone
  3. 03Perceptual Metrics for Ideal Reverberation, CCRMA (Stanford)supports the perceptual metrics that make mid-band appliance noise more objectionable than its level suggests

Each line points at one specific document. No line is a home page.

Leave this page Close this page if the noise you are chasing is a tonal hum rather than a broadband appliance — the hum page covers the tonal case.