How to Use This Page
This page covers the most common problems encountered when using the tempo detection tools on MusicTempoFinder.com — including failed microphone detection, inaccurate tap tempo results, audio BPM results that seem half or double the expected value, file upload issues, and mobile-specific problems. Work through the section that matches your situation. If your problem is not covered here, use the Contact page to submit a technical question.
For background on how each detection mode works and what a reliable result looks like, see the How It Works page.
Tap Tempo Problems
Problem: My tap tempo result varies by 5–10 BPM between sessions.
Large variation between sessions almost always comes from one of two sources:
Inconsistent tapping. Human timing variability is real — it is genuinely difficult to tap at a perfectly consistent interval, especially for longer periods. If you are drifting in tempo as you tap, the averaged result will reflect that drift. Try tapping sub-beats (eighth notes at twice the beat rate) and halving the displayed result — smaller intervals are easier to tap consistently.
Too few taps. With fewer than 6 taps, a single slightly off-tempo tap can shift the average significantly. Use 8 or more taps for a reliable result. The tap count is displayed during the session.
Problem: My tap tempo result resets before I finish tapping.
The session resets automatically after approximately 2–3 seconds without a new tap. If you are pausing between taps — for example, waiting for the next beat of a slow track — the session may reset mid-sequence. For very slow tempos (below approximately 60 BPM), the beat interval exceeds 1,000 ms, which is close to the reset threshold. Tap on every beat without pausing, or tap on every other beat and multiply the result by 2.
Problem: My tap tempo result is exactly double the tempo I expect.
You are likely tapping on sub-beats (eighth notes) rather than the beat (quarter notes). The tool measures the interval between your taps — it does not know which rhythmic level you are tapping on. If you tap on every eighth note at 120 BPM, the tool will return 240 BPM. To correct: tap on the quarter note pulse (every other tap), or halve the displayed result.
Microphone Audio Detection Problems
Problem: The microphone BPM tool will not start.
The three most common causes:
- Another application is using the microphone exclusively. Close all communication apps — Zoom, Teams, Discord, Google Meet — before loading the page. Reload and try again.
- Browser microphone permission is blocked. To reset:
- Chrome: Lock icon in address bar → Site settings → Microphone → Allow
- Firefox: Lock icon → Connection secure → More information → Permissions → Use the Microphone → Allow
- Safari: Safari → Settings → Websites → Microphone → find musictempofinder.com → Allow
- System-level microphone access is disabled.
- Windows: Settings → Privacy and Security → Microphone → ensure your browser is listed and allowed
- macOS: System Settings → Privacy and Security → Microphone → ensure your browser is ticked
Problem: The microphone tool detects no tempo or produces a very unstable result.
Music volume is too low. The music being played near the microphone must be loud enough to register clearly above the microphone noise floor. Play the track at a higher volume or move the microphone closer to the speaker.
Background noise is competing with the music signal. Ambient noise at similar frequencies to the kick drum or bass — HVAC, traffic, low-frequency hum — can confuse onset detection. Test in a quieter environment.
The track has sparse or inconsistent rhythmic content. Ambient music, classical pieces, and heavily swung jazz do not have a clear, regular beat onset pattern that onset detection can reliably identify. For this type of music, tap tempo is a more reliable approach.
OS audio processing is suppressing the signal. Automatic gain control and noise suppression applied to the microphone input can attenuate transients — exactly the signal components that onset detection depends on. Disable audio enhancements in your system audio settings and retry.
Audio BPM Result Seems Wrong
Problem: The result is exactly half or double what I expect.
This is the half-time/double-time ambiguity — the most common accuracy issue in audio BPM detection. It is not a tool error. It is an inherent limitation of autocorrelation-based beat period identification.
A track at 120 BPM and a track at 60 BPM produce autocorrelation peaks at related periods (500 ms and 1,000 ms respectively). The algorithm scores candidates and returns the highest-scoring one, but that scoring is heuristic — it can be wrong.
How to resolve it: Use the tap tempo mode on the same track. Tap along to what you feel is the beat and compare the tap tempo result to the audio analysis result. The tap tempo result reflects your perception of the beat level — which is the correct reference for production and performance use. If the audio result is double your tap result, halve it. If it is half, double it.
Problem: The result is not half or double — it seems completely unrelated to the actual tempo.
This typically indicates one of the following:
No clear beat pulse. Tracks without a regular, consistent rhythmic pulse — heavily improvised music, field recordings, spoken word, or music with a tempo below approximately 60 BPM — do not provide enough onset regularity for autocorrelation to identify a reliable beat period. Use tap tempo for this type of material.
The dominant onset is not the beat. In some productions, a high-frequency element — a rapid hi-hat pattern, a fast arpeggio, or a tremolo effect — produces more energy onset events than the kick drum or bass beat. The onset detector may lock onto this faster pattern rather than the underlying beat. Try analysing via file upload rather than microphone, which allows analysis of the full waveform rather than a live capture.
Very high tempo tracks. Tracks above approximately 200 BPM push the upper boundary of the tool’s autocorrelation lag range. For drum and bass, jungle, or other very fast genres, the tool may return a half-time value. Double the result if this is the case.
File Upload Problems
Problem: My uploaded file is not accepted.
The tool accepts MP3, WAV, AIFF, OGG, and FLAC files. Files in other formats — AAC (.m4a), WMA, or video formats such as MP4 — are not supported by the browser’s decodeAudioData() API in all browsers. Convert your file to MP3 or WAV using a free audio converter and retry.
Problem: The file upload analysis completes but returns an unexpected result.
File upload analysis uses the same onset detection and autocorrelation pipeline as microphone analysis but operates on the complete waveform rather than a live capture. The half-time/double-time considerations described above apply equally to file upload results. For files with sparse rhythmic content, very slow tempo, or complex rhythmic structure, tap along to the file after analysis to verify the result.
Problem: The file upload is very slow.
Larger files — particularly lossless WAV and FLAC files over 50 MB — require more time for the browser’s decodeAudioData() to process. The entire file is decoded before analysis begins. For large files, allow 10–30 seconds for the analysis to complete. If the analysis does not complete after 60 seconds, try uploading a shorter section of the file or converting to a compressed format such as MP3 to reduce file size.
Mobile-Specific Problems
Problem: Tap tempo works but audio BPM detection does not on my phone.
Mobile browsers have more restrictive Web Audio API implementations than desktop browsers. On iOS Safari, the AudioContext must be unlocked by a user tap before microphone capture begins. Tap the start button and then immediately play the music near the microphone. For file upload on iOS, the tool should work reliably in Safari and Chrome. For the most consistent audio analysis results, desktop use is recommended.
Problem: My tap tempo results on mobile are less accurate than on desktop.
On touchscreen devices, the tap latency — the delay between your touch and the registered event — varies by device and OS. Most modern smartphones have touch latency of 20–50 ms, which introduces an error of approximately ±1–2 BPM at 120 BPM. Using a hardware keyboard on a laptop or desktop eliminates this source of error. For critical tempo measurements on mobile, use 12 or more taps to average out touch latency variation.
Still Having Problems?
If none of the above resolves your issue, the FAQ covers additional questions about result interpretation and algorithm behaviour. For problems not addressed in either page, contact Hurst directly via the Contact page with the following information:
- The tool mode you were using (tap tempo, microphone, or file upload)
- Your browser name and version
- Your operating system and version
- A description of the result you are seeing and what you expected
- For file upload issues: the file format and approximate duration
Response time for technical issues: 24–48 hours.
Written by Hurst, founder of MusicTempoFinder.com. Last updated: June 2026.