Clap once and find out how echoey your room is, and what to do about it
Processed in your browser. Nothing is uploaded. Updated on 24 Oct 2026.Clapping once and measuring how long the sound takes to fade says how echoey a room is. Under 0.5 seconds suits recording; much longer and speech starts to smear. Runs on your own microphone, and the recording is never uploaded to anyone.
Press Allow the microphone and continue, and allow access when the browser asks.
Stand about a metre from the microphone in the middle of the room and keep everything else quiet.
Press Record a clap, then clap once, sharply, while the three-second countdown runs.
Read the RT60 in seconds and the verdict on screen; the clap is analysed on your device and then discarded.
Clap once, and this page measures how long the sound takes to die away in your room and reports it as RT60, the reverberation time. Everything is measured in the page from a recording that is analysed and then discarded; nothing is saved and nothing is uploaded.
Under 0.3 seconds the room is dry. Voices stay crisp on a call, a recording needs no cleaning up, and carpet and curtains have already done the work. Between 0.3 and 0.5 seconds is comfortable and describes most furnished living rooms. Between 0.5 and 0.8 seconds the room is live: listeners hear a hollow quality and automatic noise suppression starts chewing at the ends of sentences. Over 0.8 seconds is echoey, which means bare walls, a hard floor and probably a high ceiling, and recordings made there cannot be repaired afterwards.
The usual cure is soft material on two adjoining surfaces rather than one. A rug on the floor and a curtain across the hardest wall will take a surprising amount off; so will a bookshelf, a sofa and anything else that is neither flat nor hard. Moving closer to the microphone helps too, because it raises the direct sound relative to the reflections without changing the room at all.
RT60 is the time a sound takes to fall by sixty decibels once the source stops. Measuring the full sixty needs a clap far louder than the room, which rarely happens on a laptop, so the standard method in ISO 3382 fits a cleaner part of the decay and extends it: T20 fits from -5 dB to -25 dB, T30 from -5 dB to -35 dB. The tool picks whichever your recording supports and names it.
The decay itself comes from Schroeder backward integration, where each point on the curve is the energy still to come after it. That integral is smooth enough to fit a straight line through, and the integration stops where the decay meets the room noise, because integrating several seconds of hiss would flatten the curve and report a far longer time than the truth.
Under 0.3 seconds is dry and ideal for speech. Between 0.3 and 0.5 seconds is comfortable and covers most furnished living rooms. Above 0.8 seconds a room is echoey enough that consonants smear and noise suppression starts cutting words short, which no amount of editing afterwards will repair.
Put soft material on two adjoining surfaces rather than one: a rug on the floor and a curtain across the hardest wall will take a surprising amount off. Bookshelves, sofas and anything neither flat nor hard help too. Moving closer to the microphone raises the direct sound relative to the reflections without changing the room.
RT60 is the time a sound takes to fall by sixty decibels after the source stops. Measuring all sixty needs a clap sixty decibels above the room noise, which rarely happens indoors. ISO 3382 solves that by fitting a cleaner part of the decay, from -5 dB to -25 or -35 dB, and extending the line.
Refusal happens when the clap was not far enough above the background noise, which needs at least 35 decibels of headroom for a usable fit. Turn off fans and air conditioning, move away from a window, clap harder and closer to the microphone, then record again. Guessing from too little decay would invent the answer.
A clap is a rough but honest impulse and is what acousticians use for a quick check. A balloon burst or a starter pistol carries more low-frequency energy and gives a better reading in the bass. Expect a browser clap measurement to land within about ten per cent of a proper one in an ordinary room.
No. The few seconds of audio are held in memory only long enough to compute the decay curve, then discarded, and the microphone is released as soon as the recording ends. Nothing is written to a file and no audio is sent to any server at any point.
Still stuck? Ask about Room Echo Test (RT60) and anybody who knows can answer. No account needed.
Also works as: Reverb Time Calculator · RT60 Calculator · Room Acoustics Test.
Same job, other tools: Frequency Generator – Generate a steady test frequency for speakers, headphones or instruments · Mic Test – See your voice on a live waveform and hear it played straight back · Audiogram Test – Draw an audiogram-shaped chart from tones you say you could hear · High Frequency Hearing Check – Screen the top tones, 4 kHz and 8 kHz, ear by ear, then test higher with a tone generator · Audio Sweep Generator – Play a slow logarithmic sweep and hear every rattle in the room.
On the related tools: Remove Pauses from Audio – Cut every pause out of a voice recording completely, for tight podcast edits.
In the collection Audio Test Tools.