Pocket decision
Design cadence against the chosen count: allocate sustained vowels, rests, and ad-libs at lower BPM, or partition dense language into smaller pockets as tempo rises.
Tempo and Key Finder with Private Browser Analysis
Analyze one or more audio files locally for BPM, musical key, and Camelot position. Your browser does the work; the files are not uploaded to AI2Rap.
Tempo and key are estimates, especially with changing harmony, weak percussion, or dense layers. Confirm uncertain tracks by ear.
Find BPM and key without uploading the track
Batch input · MP3, WAV, FLAC, M4A, AAC, or OGG · 100MB maximum per file
On-device analysis keeps every selected file inside this browser
Analyze BPM and key without sending the audio away.
Choose the files, let two rhythm trackers and harmonic templates estimate the coordinates, then verify the result by ear.
- 01
Select one or more files and keep the tab active while the Web Audio API decodes them locally. The audio is never posted to an AI2Rap endpoint or added to an account record.
- 02
Essentia.js evaluates harmonic evidence against EDMA templates and runs Multifeature and Degara rhythm strategies. Key strength and rhythm agreement answer different questions, so inspect both before trusting the coordinates.
- 03
Evaluate the primary BPM with its half- and double-time readings, then inspect key, scale, match strength, and Camelot position before using the estimate.
Route the Detected BPM and Key into the Next Decision
Carry the verified BPM and key into AI Rap Generator as production constraints, then add subject, section plan, mood, and vocal direction before creating candidates.
Tempo gives the lyric brief a measurable rhythmic target. Lower values permit longer phrase envelopes; higher values demand tighter syllable placement. Correct the detected grid by ear before locking the brief.
Apply the verified grid to a rap briefTonal coordinate
Treat key as a handoff for melodic roles, not a writing rule. Give the estimate to collaborators adding a sung hook, bass, or harmony, and use Camelot only to shortlist references for listening.
Listening check
When estimators conflict, place the primary, half-time, and double-time grids against transients in your DAW. Live timing, sectional changes, and modulation are valid reasons to reject the automatic result.
Brief transfer
Record the human-verified BPM, tonic, mode, and Camelot code with the project. A four-field technical note reduces drift when arranging vocals, selecting samples, or rebuilding a session later.
Map the track with tempo, key, and Camelot code
01
One pulse shown on primary, half, and double grids
Inspect the primary tempo beside its half-time and double-time equivalents when beat tracking and musical counting disagree by a factor of two.
02
Estimated tonic and mode
The result pairs tonic and mode with EDMA template-match strength, exposing weak harmonic evidence instead of presenting every estimate as equally certain.
03
Camelot coordinate
AI2Rap converts the estimated tonic and scale into a Camelot coordinate, giving DJs a compact compatibility reference to verify by ear.
Read BPM as a choice of counting level
Beats per minute expresses how many pulses occupy sixty seconds, but musical counting is hierarchical. The same groove may sensibly read at 70 or 140 depending on subdivision, so the result preserves half-time and double-time interpretations beside the primary estimate.
Interpret the estimated tonic and mode
The estimated key names the tonic and major or minor scale whose pitch pattern best explains the analyzed audio. It summarizes a tonal center; it does not assert that every chord, sample, or later section remains inside that scale.
Use Camelot to shortlist harmonic transitions
Camelot notation converts each detected major or minor key into a compact coordinate such as 8B or 8A. Adjacent positions help DJs narrow transition candidates quickly, but the map cannot account for every modulation, detuned sample, or contextual clash.
How the browser estimates rhythm and tonal center
Inside the browser, Essentia.js KeyExtractor compares the track's harmonic pitch-class distribution with EDMA major and minor templates. Rhythm is estimated separately: the Multifeature RhythmExtractor2013 strategy combines several rhythmic cues, while Degara follows candidate beat positions. AI2Rap selects the displayed integer BPM from those rhythm estimates and reports key strength as template similarity.
Treat every detection as a coordinate, not a musical verdict
The alternative tempos are mathematical views of the detected pulse, not additional analyses. Likewise, key strength reports similarity to one EDMA template rather than universal accuracy. Verify changes, ambiguous harmony, and unstable rhythm by listening across the track.
What stays inside this tab
Data flow stops at the current tab: Web Audio decodes the selected bytes, a local worker measures them, and closing or reloading removes the working result. The audio and coordinates are neither uploaded to AI2Rap nor attached to an account.
Production Decisions Informed by BPM and Musical Key
Reach for BPM, key, and Camelot analysis when timing or harmonic compatibility will change the next production decision.
Align written cadence with the beat’s measured pulse
Analyze the beat before designing cadence and phrase length. Compare primary, half-time, and double-time grids when the pocket supports more than one valid counting level.
Shortlist transitions before auditioning the blend
Reduce a DJ transition search to compatible Camelot candidates, then listen through the actual blend to catch modulation, detuning, or harmonic context the coordinate cannot encode.
Establish coordinates for a remix or production pass
Document a reference track's tempo and tonal center as production coordinates before composing. Use those measurements for orientation while keeping melody, structure, and arrangement original.
Match candidate beats to a vocal pocket
Process several candidate beats sequentially and compare their result cards in one view when matching material to a vocal range, verse pocket, practice set, or writing session.
How local BPM and key analysis behaves
What does the key match percentage actually measure?
Treat the percentage as similarity to the selected EDMA key template, not ground-truth accuracy. Stable harmony gives stronger evidence; modulation, detuned samples, noise, and dense layers can weaken it. Check several sections by ear before writing the key into a project brief.
Why does the result show half-time and double-time BPM?
The same pulse can be counted at half or double speed, so a groove near 70 BPM may also make musical sense at 140. Compare each displayed grid with kicks, snares, phrase lengths, and your intended cadence before choosing the production count.
Can one BPM and key describe a song that changes sections?
Analyze a stable representative passage when possible. A single result cannot describe every key change or tempo drift across a track. If sections disagree, record the relevant coordinates per section or verify the full arrangement in a DAW rather than forcing one label.
How should a DJ use Camelot results without trusting them blindly?
Use the code to shortlist same-number or neighboring Camelot positions, then audition the actual transition. The coordinate cannot account for modulation, tuning differences, sectional harmony, or the energy of the blend, so it narrows candidates rather than approving a mix.
What leaves my browser during local tempo and key analysis?
The Web Audio API decodes each selected file and a worker in the current tab performs the calculation. AI2Rap receives no audio and writes no account record. Only the pinned Essentia.js engine may load from jsDelivr when it is not already cached.
How should I prepare a large batch for browser analysis?
MP3, WAV, FLAC, M4A, AAC, and OGG files up to 100MB are accepted and processed sequentially. Files above 50MB can take longer because decoded audio uses substantial memory. If decoding exceeds the safe budget, try a shorter or lower-sample-rate copy.