Skip to content
Sign in

/insights · Engineering

How to measure reverb (RT60) with your phone, and when to trust it

Measure RT60 per octave band from a single hand clap: where to put the phone, how the Schroeder decay curve is read (T20/T30), why a phone mic is fine for time measurements, and when you need a sweep through a speaker instead.

August 26, 2026 7 min read roomlabrt60reverbroom acousticshome studioacoustic treatmentiso 3382

Clap once in your room and listen. That tail you hear is reverberation, and the number that describes it is RT60: the time it takes sound to decay by 60 dB. A bare bedroom sits around 0.6 to 1.0 s. A treated vocal corner gets under 0.3 s. A bathroom can pass 1.5 s.

You can measure it with the phone in your pocket, and for most decisions (which room, which corner, did those panels do anything) the phone is genuinely enough. This article shows how, and, just as important, when not to trust the number.

RoomLab measures RT60 per band from one clap, free on Android

Why a phone mic is acceptable here

A phone microphone is coloured, uncalibrated, and processed. For an absolute frequency response measurement that disqualifies it. RT60 is different: it is a time measurement. You are not asking “how loud is 500 Hz”, you are asking “how long does 500 Hz take to die”. Colouring shifts the level, not the clock. Two things do break the clock, and both are avoidable:

  1. Voice processing. Android’s automatic gain control and noise suppression will happily “clean up” your decay tail, turning a 0.8 s room into a 0.3 s guess. The measurement has to run on the raw, unprocessed microphone source. (RoomLab requests exactly that, and its Signal chain screen shows what your device does to sound.)
  2. A weak impulse. A hand clap has almost no energy below 250 Hz, so the low bands cannot be read from a clap at all. Honest tools leave those bands empty rather than extrapolating.

The clap method, step by step

  1. Placement. Phone on a hard, level surface: a table, a stool. Not on a bed, not in your hand. Mic edge (usually the bottom) unobstructed.
  2. Distance. About a metre from where you clap, away from walls and corners: against a wall you measure the wall’s reflection, not the room.
  3. Silence, then one sharp clap. Cupped palms, single impact. The room’s answer, band by band, is the measurement.
  4. Repeat in two or three spots. One position is an anecdote. The spread between the centre and the corner is usually the real finding: the same room can swing by half a second.

How the number is actually read (and why you should get to see it)

Under the hood the recording becomes a Schroeder decay curve: the energy of the tail, integrated backwards, per octave band. A line is fitted to the fall from −5 to −25 dB (T20) or −5 to −35 dB (T30) and extrapolated to 60 dB, per ISO 3382. That fit is only meaningful if the decay stands far enough above the background noise: T20 needs roughly 35 dB of peak-to-noise, T30 about 45 dB.

This is why a tool should show you the curve and a confidence rating, not just a number. A kitchen at night and a kitchen with a fridge humming give the same clap two very different peak-to-noise ratios, and only one of them deserves your trust. If an app hides the curve, it is asking you to take its word for it.

When you need the sweep instead

For the 125 and 250 Hz bands, or any room where the low end is the question (mixing rooms, drum rooms), a clap physically cannot deliver the energy. The standard answer is an exponential sine sweep played through a speaker, recorded, and deconvolved into an impulse response (Farina’s method). That needs an external speaker a metre or more from the phone: a phone playing its own sweep mostly measures itself.

Practical rule: clap for mids and highs and quick comparisons; sweep with a speaker when the low bands or a report for someone else are on the line.

What to do with the number

  • Under 0.3 s in the speech bands (500 Hz to 2 kHz): fine for recording voice.
  • 0.4 to 0.7 s: audible room in every take; treat first reflections and add absorption.
  • Above 0.8 s: the room dominates; measure a few spots and treat the worst surfaces first, or record somewhere else.
  • Before and after. The only way to know whether panels did anything is the same measurement in the same spot, before and after. The difference per band tells you whether you bought the right thickness: thin foam moves 2 kHz and does nothing at 250 Hz.

Measured with RoomLab, which shows the decay curve behind every bar, maps multiple positions in a rotatable view, and can play your own room’s response back to you. Free to measure on Android; the treatment plan in square metres per material is a one-time Pro unlock.