Method 1: count the bars
If the loop is a clean loop — starts on a downbeat, ends where it repeats — this gives you the exact tempo with arithmetic and nothing else.
BPM = (beats in the loop × 60) ÷ length in secondsIn 4/4: 1 bar = 4 beats, 2 bars = 8, 4 bars = 16, 8 bars = 32.
A 2-bar loop that is 3.75 seconds long:
(8 × 60) ÷ 3.75 = 128 BPMThe method's real strength is the sanity check built into it: an implausible answer tells you your bar count was wrong, which is information an automatic reading never gives you.
Where it breaks: a loop with reverb spilling past the loop point, or one trimmed a few milliseconds short. Both make the file longer or shorter than the musical length. If you get 127.4 in a pack where everything else is 128, trust 128 — you are looking at a trim artefact.
Method 2: tap it
Every DAW has a tap-tempo button. Tap along for eight bars and read the number.
Good for: getting within a couple of BPM in five seconds. Not good for: anything you intend to stretch. Human tapping lands within about ±2 BPM, and 2 BPM is a beat of drift every thirty seconds.
Use it to find the neighbourhood, then confirm.
Method 3: warp it
Import the loop, let the DAW's warp engine find the grid, read the tempo it settled on.
Good for: loops you are going to warp anyway. Watch out for: the DAW guessing the wrong metrical level and stretching your loop to half or double speed. It sounds obviously wrong and is a two-second fix once you notice.
Method 4: analyse it
Drop the file in and read the number. DropWave reports tempo to two decimal places alongside the key, the Camelot position and the loudness measurements, from one drop.
The decimals matter more than they look. Stretching a loop to a session using 128 when the file is 127.5 puts you a beat out every four bars — exactly the drift you notice at bar 33 and then spend twenty minutes hunting.
Why the BPM you got is "wrong"
Almost always, it is not. It is a different metrical level.
Half time and double time
A drum & bass tune at 174 BPM has a half-time feel at 87. A trap beat at 140 is frequently written as 70 with double-time hats. Both grids genuinely fit the audio; which one is correct is a question about how you count the track.
If your analyser reports 87 and you expected 174, both are right. Halve or double it and move on.
The behaviour worth insisting on is that the alternative gets named. DropWave prints "could also be 206.32 BPM" when the two were close, because that is a real property of the audio rather than an error to hide.
The genuine errors
There is one family that is not a metrical preference: 103 reported as 137. That is a grid at an awkward ratio that lines up with a plausible subset of the hits and is simply incorrect.
It happens most where the kick is sparse and the rest of the percussion carries the pulse — which is also why a tempo reading that only listens to the low end falls apart the moment the kick drops out. Reading the pulse across the whole spectrum is what keeps the grid alive through a breakdown.
When there is no BPM to find
- Vocals and a cappellas. Soft transients, and often no beat present at all. A key reading on a vocal is reliable; a tempo reading usually is not. See how to find the key of a vocal.
- One-shots. Nothing periodic.
- Ambient, drones, pads. Same.
- Rubato and live playing. The tempo moves, so there is no single number.
A confidence band is the useful output here. A tool reporting "unsure" on a vocal is telling you something true.
A practical routine
- Analyse the file. Note the BPM and the confidence.
- If it is exactly half or double what you expected, it is a metrical level. Pick the one that matches how you count it.
- If it is neither half nor double — 137 against 103 — count bars manually to confirm.
- If confidence is low and the material is a vocal, a pad or a one-shot, there is probably no tempo. Do not chase it.
- Put it in the filename.
break_128_Am.wavbeats any database, because it travels with the file.