Reference · Tempo of a loop or sample

How to find the BPM of a sample

Count the beats over a fixed stretch of time and scale to sixty seconds: 32 beats in 20 seconds is 96 BPM. Then play the same groove at half and at double that number, because both readings are musically true and the one chosen decides how everything after it is chopped and arranged. A one-shot has no tempo to find, because nothing repeats, so there is no interval to measure.

Session™ Key, described further down this page, is published by Capiscana, Inc., which operates this site. Facts about it are dated to the day the source was read.

Counting it, step by step

This costs nothing, works on a phone speaker, and is the method every other method is checked against.

1. Loop it cleanSet the loop so it starts on a downbeat and repeats with no gap and no stutter. A loop that hitches at the seam will make every count after it wrong.
2. Count quarter notes against a clockStart a stopwatch and count the pulse for twenty seconds. Tapping a finger on each beat while counting keeps the count honest, because the hand notices a dropped beat before the mouth does.
3. Scale up to sixty secondsTwenty seconds is one third of a minute, so multiply the count by three. Thirty-two beats in twenty seconds is 96 BPM. Forty seconds multiplies by one and a half instead.
4. Confirm against a clickSet a metronome to the number and play it over the loop. A click that stays locked for eight bars confirms the count, and a click that drifts apart means the count was off by a beat or the loop was cut short.

Precision follows the length of the count. One beat missed across twenty seconds shifts the result by three BPM, which is enough to hear as drift over a long section. Counting across forty seconds instead cuts that error in half.

The half-time and double-time trap

The common mistake is not a wrong number. It is the right number at the wrong scale. A groove counted at 75 is the same groove counted at 150, and both are correct descriptions of the same audio, because the pulse a listener follows is a choice about where the beat sits.

The choice is not cosmetic. It sets the grid the sample is chopped on, the note values every part after it is written in, and whether a hi-hat pattern reads as sixteenths or as eighths. Deciding it after an arrangement exists means rewriting the arrangement.

The test is to play the loop against a click at the counted tempo, then at half it and at double it. The one that matches how the body counts the groove is the one to build on. Every method falls into this trap, by ear and in software, because there is no measurement that settles it.

Warping it in a DAW

The other manual route is to drop the sample onto the timeline, guess a tempo, and adjust until the transients line up with the grid across the whole loop. It is exact, and it reads a tempo off material that is too dense or too sparse to count comfortably.

What it costs is order. The sample has to be inside a project before the project tempo is known, so a first guess gets committed to and then corrected. Lining transients up against the grid also inherits the same half-time question, since a grid at 75 and a grid at 150 both accept the same transients.

Reading the tempo from a Mac's audio output

Session™ Key is a macOS menu bar app that reads the BPM and the key of whatever is playing through the Mac's audio output and shows both in the menu bar. The source is the machine's own output, so a sample previewed in a browser, played from a library or bounced in a DAW is read the same way. Detection runs on the device and no audio is uploaded, recorded or stored.

Every read carries a confidence mark, and when a groove is equally defensible at half or double time both tempos are named rather than one being chosen silently. One keystroke then writes the tempo into Pro Tools, Logic Pro, Ableton Live, FL Studio or Studio One, and the key as well in every one of those except FL Studio, which has no project key field.

What it costs and what it runs on

Common questions

How is the BPM of a sample found without software?

Loop the sample, count the quarter notes across a fixed stretch of time, and scale the count to sixty seconds. Thirty-two beats in twenty seconds is 96 BPM, and tapping each beat while counting keeps the count honest.

Why does the same sample feel like two different tempos?

Because both readings are true. A groove at 75 BPM is the same audio as a groove at 150, half time against double time, and the choice decides the grid it is chopped on and the note values everything after it is written in.

Can the BPM of a one-shot be found?

No. Tempo is an interval between repeating events, and a single hit has nothing to repeat against. A number returned for a one-shot describes the length of the file rather than a tempo.

How accurate is counting by hand?

It follows the length of the count. Missing one beat across twenty seconds moves the answer by three BPM, and counting across forty seconds instead halves that error.

Is there a way to read a sample's BPM without uploading it?

Yes. Session™ Key reads the tempo from the Mac's own audio output while the sample plays. Detection runs on the device, so nothing is uploaded, recorded or stored.