Tap along with the track and read its tempo. A rolling average with the raw last gap beside it, a steadiness score for your own timing, and the millisecond values the number is worth once you have it.
Tap the pad, click anywhere on it, or press the space bar in time with the track.
Measured as the spread of your gaps around their own average, so it means the same thing at 70 BPM as it does at 170. It grades your tapping, not the track.
Steadiness needs at least 4 taps (three gaps) before it says anything honest.
Newest first, up to the 8 gaps in the rolling window. The BPM column is what each single gap would have read on its own; watch how far one of them can sit from the average.
No gaps yet. Two taps produce the first one; the window then keeps the last 8.
Click any figure to copy it into a delay time, a pre-delay, or a gate release.
No microphone, no audio, no upload. The page times your own key presses with the browser's monotonic clock, and nothing leaves this tab.
No tempo measured yet.
Play the tempo you just tapped as a click, with meters, accents, and swing, and hear whether it really matches the track.
Take the number further: every note value in milliseconds and hertz, dotted and triplet columns, and reverb pre-delay and decay.
Once the tempo is settled, get the pitch right: sustained reference tones with a movable A4 for tuning by ear.
Counting bars against a stopwatch is slow and gives you one figure with no idea how much to trust it. Tapping gives you the same answer in a few seconds, and, because every gap is measured separately, it can also tell you how steady the tapping was, which is the difference between a tempo you can commit to and a guess.
Click the pad or press the space bar (any character key works) on the pulse you would clap to. Nothing appears on the first tap, because a tempo is the gap between two taps, not a single moment.
From the second tap the rolling average takes over, and the raw last-gap figure sits beside it so you can see how much one tap moved the number. Six to eight taps is usually where it stops wandering. Stop for two seconds and the run closes, holding the reading until your next tap starts a fresh one.
Copy the whole, half, or exact BPM, halve or double it if you tapped the wrong layer of the groove, nudge it by a tenth to match a loop, then send it to the metronome or read the note lengths in milliseconds.
The headline number averages up to eight intervals and drops the fastest and slowest once there are five, while the last single gap is shown untouched, the clearest way to see whether you are settling or drifting.
The spread of your gaps is measured against their own average, so it reads the same at 70 BPM as at 170. You get the average deviation in milliseconds, the percentage spread, your fastest and slowest gap, and a plain-language verdict.
Tapping the backbeat of a fast track halves the reading and tapping eighths on a slow one doubles it. Both are one button away, so a misread groove does not mean tapping the whole thing again.
Three copyable readings for three situations: the exact average for a sampled loop, the whole number a produced track was written at, and the half-BPM step that hardware sequencers and older transports accept.
Every interval in the rolling window is listed newest first with the tempo it implies on its own and how far it sits from the average, with the outliers marked. It is a practice tool as much as a measuring one.
Quarter, eighth, dotted eighth, sixteenth, and full-bar lengths in milliseconds are calculated from the tapped tempo, ready to paste into a delay or a gate, with a one-click handoff to the metronome.
Because one tap is a moment, not a tempo. A tempo is a distance between two moments, so the first reading can only appear on the second tap; that gap, converted with 60,000 divided by its length in milliseconds, is the raw BPM. Every tap after that refines it.
Four gets you a steadiness figure, six to eight gets you a tempo you can act on. A single gap carries all of your reaction-time error; averaging eight of them divides that error down, and dropping the fastest and slowest removes the one tap where you sneezed. Past about a dozen taps the number stops improving; if it is still moving, you are tapping something other than a steady pulse.
Human tapping scatters by roughly 10 to 30 ms even when it feels locked. At 120 BPM one gap that is 20 ms long reads 115 BPM instead of 120, a whole tempo out on a single mistake. The average of eight gaps shrinks that scatter by about two thirds, which is why the headline figure here is the average and the last gap is shown only for comparison.
Almost certainly nothing: you tapped a different layer of the same groove. Drum and bass sits at about 174 BPM but its backbeat feels like 87, and a slow hip-hop beat at 85 invites you to tap the hi-hats at 170. Both readings are correct descriptions of what you tapped. Use the half-time and double-time buttons to move between them, and sanity-check against genre: most club music lives between 120 and 130, most rock and pop between 90 and 140.
Tap every beat. BPM counts beats, so tapping only the first beat of each bar in 4/4 gives you exactly a quarter of the tempo. Tapping the downbeat is useful for finding where a bar starts, which is a different job; the beat dots on the pad count you round a bar of four so you can see which beat you are on while you tap.
It depends what you tapped. Anything sequenced to a grid (most pop, electronic, and modern hip-hop) was written at a whole or half BPM, so if you land on 127.6 the honest answer is 128 and the decimals are your own timing. Anything played by people without a click, which includes most jazz, most classical, older rock recordings, and live sets, genuinely has no single tempo, and a figure like 118.4 is a fair average of something that breathed.
You can measure any one stretch of it, but not the whole thing. A rubato passage, a live band pushing into a chorus, or a production with a written tempo map has a different tempo in different bars, and averaging across them describes a tempo the music never actually played. Tap a stretch that feels level, note the figure, then tap another stretch and compare. If the two disagree by more than a couple of BPM, the track is moving and no single number will fit it.
The timing comes from the browser's monotonic clock, which is precise to well under a millisecond, but it is deliberately made slightly fuzzy in most browsers to frustrate timing attacks, and your keyboard, mouse, or touchscreen adds a few milliseconds of its own. All of that is far smaller than human tapping scatter, so the limit here is you, not the page. That is what the steadiness score is for.
No. There is no microphone access, no audio, and no upload: the page only times your own key presses and clicks, in this tab. Automatic BPM detection from audio is a different technique with different failure modes, and it is not what this tool does.
A two-second gap is treated as you stopping rather than as a very slow beat, so the next tap starts a fresh measurement instead of poisoning the average with the pause. The reading you had stays on screen until then, and the run indicator tells you it has closed. Two seconds is also the practical floor of this tool: sustained tapping slower than 30 BPM cannot be measured this way.
Explore the growing collection for calculations, documents, writing, and everyday work.