Free Tool
Record ten seconds of your voice and find out whether your podcast or stream sounds rough because of your microphone, your levels, or the room you're sitting in. Runs entirely in your browser — nothing is uploaded.
Sit where you normally record. Read this out loud at your normal speaking volume — the pauses matter as much as the words.
The quick brown fox jumps over the lazy dog.
Pack my box with five dozen liquor jugs.
How much wood would a woodchuck chuck?
She sells seashells by the seashore. Stop.
Your microphone audio is analysed on your device and never sent anywhere.
Almost every microphone test online checks the same handful of things: is the mic connected, is the level moving, is it clipping. Those are worth knowing, but they very rarely explain why a recording sounds amateur. The usual culprit is the room.
Your voice reaches the microphone twice. Once directly, and then again a few milliseconds later after bouncing off a wall, a window, a desk, a bare floor. Your ears are extremely good at ignoring those reflections; a microphone is not. It records them at full strength, which is why a room that sounds perfectly normal to sit in can sound hollow and distant once it's recorded.
This is also why upgrading your microphone so often disappoints. A more sensitive, more detailed microphone in an untreated room simply captures the reflections more accurately. If your recordings sound roomy, the fix is almost never a better microphone — it's a bit of soft material in the right places.
When you stop speaking, the sound in the room keeps going for a moment as those reflections die away. The time that decay takes is what acousticians call reverberation time, and it is the single most useful number for a recording space.
Normally you'd measure it by playing a test tone or popping a balloon and recording the result. This tool instead uses a blind estimation method, which recovers the decay from ordinary speech by looking at the short gaps between your words. It's a maximum-likelihood estimator described by Ratnam and colleagues in the Journal of the Acoustical Society of America in 2003, and it means you need no test signal, no calibration and no special equipment.
Alongside that we measure your noise floor, how far your speech sits above it, your peak level and whether anything clipped. The score weights the room most heavily, because it is both the biggest contributor to how amateur a recording sounds and the thing people are most likely to misdiagnose as a gear problem.
Tested against speech mixed with simulated rooms of known reverberation time, the estimate lands within about 0.03–0.07 seconds when the recording is clean. Two honest limits are worth knowing. It cannot resolve below roughly 0.2 seconds, so a very well treated booth is reported as “under 0.25s” rather than given a precise figure. And background noise hides the tail of the decay, which pushes the estimate upwards — so if your noise floor is high, the tool tells you the reverb figure is probably an overestimate and marks the result low confidence. Fix the noise, then measure again.
Treat the band — dry, controlled, live, very live — as the reliable output, and the exact number as an indication.
For spoken word, aim for around 0.3 seconds or less. Broadcast voice booths sit nearer 0.2. A typical untreated spare bedroom lands between 0.4 and 0.6, and a living room with hard floors and big windows can easily exceed 0.8, which is where voices start to sound like they were recorded down a hallway.
You do not need to build a studio to get there. Two or three absorptive panels at the first reflection points, something soft behind the microphone, and a rug on a hard floor will move most rooms into the right range for well under the cost of a mid-range microphone.