Audio to MIDI Converter
Turn a recording into editable notes. Upload WAV, FLAC, AIFF, OGG, M4A or MP3 and get a MIDI file with the pitches, timings and velocities — ready to drop into your DAW.
Powered by basic-pitch (Spotify Research, open source).
What is an audio to MIDI converter?
An audio to MIDI converter transcribes a sound recording into note data: which pitches were played, when each note starts and stops, and how hard it was struck. The output is a standard MIDI file — not audio, but instructions any DAW or notation app can play through any instrument you choose.
The hard part of the problem is polyphony. Tracking one melody line is old technology; hearing a chord and separating it into its component notes is what needed neural transcription models. This converter is polyphonic, so piano parts, guitar chords and layered lines all come back as multiple simultaneous notes rather than a single guessed pitch.
The practical use is separation of performance from sound: a take that was played right but recorded wrong, an idea captured on a phone, a loop you want to re-voice — all become editable notes. Transcription is estimation rather than extraction, so expect a strong first draft to correct in your DAW, not a perfect score.
Audio to MIDI converter features
Polyphonic transcription
Chords and overlapping lines are detected as simultaneous notes, not collapsed into one pitch — the difference between a transcriber and a tuner.
Six input formats
WAV, FLAC, AIFF, OGG, M4A and MP3 are all accepted and decoded before analysis. The less lossy your source, the cleaner the transcription.
Timings and velocities kept
The MIDI carries note starts, durations and velocities, so the feel of the performance survives — not just the pitches.
Standard MIDI file out
The download is a plain .mid file. Drag it into any DAW, hand it to notation software, or load it in the MP3 to MIDI page's sibling tools.
No installs, no DAW required
Conversion runs from the browser page — nothing to download or configure. You only need a DAW afterwards, to actually play the notes.
Free to use
Upload, convert and download without an account. If a conversion misses notes, try a less compressed copy of the same recording.
Which audio format should you convert from?
All six formats below are accepted and all are decoded before analysis, so the converter never sees the container — only the waveform. What differs is how much of the original waveform survived encoding, and that is what limits transcription accuracy. If you have a choice, convert from the least-compressed copy you own.
| Format | Type | What it means for transcription |
|---|---|---|
| .wav | Uncompressed | The reference case. Nothing has been discarded, so quiet notes, overtones and fast passages are all intact for the analyser to work from. |
| .aiff | Uncompressed | Equivalent to WAV in content — same uncompressed samples, different container. No accuracy difference between the two. |
| .flac | Lossless compressed | Smaller than WAV but bit-for-bit identical once decoded. There is no transcription penalty for using FLAC over WAV. |
| .mp3 | Lossy | Encoding discards quiet content masked by louder content — exactly the material a transcriber needs for inner voices. Usable, and the higher the bitrate the better; 128 kbps loses noticeably more than 320. |
| .m4a | Lossy (AAC) | Generally holds up better than MP3 at the same bitrate, particularly in the high end where cymbals and string harmonics live. |
| .ogg | Lossy (Vorbis) | Comparable to AAC in practice. Fine as a source, but a lossless copy of the same recording will always transcribe more cleanly. |
Who converts audio to MIDI
Producers reusing a recorded part
A guitar or piano take that plays right but sounds wrong becomes a MIDI part you can hand to any instrument. The performance survives; the tone gets replaced.
Songwriters capturing an idea
The phone recording of a riff turns into notes on a grid. Faster than transcribing by ear, and it keeps the timing feel of how you actually played it.
Sample flippers
Converting a loop to MIDI gives you the notes without the recording — a starting point you can re-voice, re-time and build on rather than chopping the audio itself.
Anyone scoring a part
MIDI imports into notation software, so a recorded performance becomes a first-draft score you correct rather than a blank page you fill.
How audio to MIDI conversion works
1. Upload a supported audio format
WAV, FLAC, AIFF, OGG, M4A or MP3. The file is decoded to raw audio first, so the format only matters insofar as it affects what survived encoding.
2. Notes get detected
A neural transcription model estimates which pitches are sounding at each moment and where each note starts and stops. Unlike simple pitch tracking, it handles several notes at once.
3. Download the MIDI
You get a standard MIDI file with pitches, start times, durations and velocities. Drag it into any DAW or notation app and edit it like anything else.
Audio to MIDI FAQ
How do I convert audio to MIDI?
Upload the audio file here and the transcription runs automatically, returning a downloadable MIDI file. There's nothing to install and no DAW required — though you'll want one to actually use the result, since MIDI carries note data rather than sound.
Does it work on chords, or only single notes?
Chords work. The model is polyphonic, so it estimates several simultaneous pitches rather than tracking one line. Dense full-band mixes are still the hardest case — an isolated instrument transcribes far more cleanly than a finished master.
What's the difference between this and the MP3 to MIDI page?
Same engine, different focus. This page covers all six accepted formats and when each one costs you accuracy. The MP3 to MIDI page is specifically about converting from lossy MP3, which is the most common and most compromised source.
Why does the MIDI have wrong or extra notes?
Transcription is estimation, not extraction. Overtones can register as real notes an octave or a fifth up, reverb tails stretch note lengths, and quiet inner voices get missed. Treat the output as a strong first draft to correct in your DAW rather than a finished transcription.
Does converting to MIDI keep the instrument's sound?
No, and that's the point. MIDI stores which notes were played and when — not the audio. When you load it in a DAW it plays through whatever instrument you assign, which is what makes it useful for changing the sound while keeping the performance.
Is drum audio worth converting?
Not with this tool. Percussion is unpitched, and pitch-based transcription has nothing to lock onto. For a drum groove you want onset detection instead — the Beat Finder gives you the beat positions to rebuild the rhythm from.
Turn any recording into editable notes.
WAV, FLAC, AIFF, OGG, M4A or MP3 in — a MIDI file out.
More AI music tools
One place to generate, convert, detect, and remix.
Turn a prompt into a full song with vocals.
Write full song lyrics from a theme or a line.
AI rap diss with lyrics + beat in seconds.
Stage name ideas by scene — trap, drill, classic.
Convert an MP3 into editable MIDI notes.
Add a slowed + reverb tape vibe, free in your browser.
Find any track's tempo by tap or upload.
Find whether a song counts in 4 or in 3.
Detect the musical key of any song.
Check your pitch accuracy note by note.
Get key, Camelot code and tempo from one upload.
Batch-key up to 25 tracks into one table.
Find the key through your mic — no file needed.
Get every beat position, bar by bar.
Tap out a tempo, get delay times in ms.
Every note and chord in any key.
Find the note and frequency of any sample.
Live chromatic tuner in your browser.
Turn a melody into note names and timings.
Measure your vocal range low to high.
Tape warmth, wobble and vinyl crackle on any track.
Guitar, bass, ukulele and more — notes, Hz and a tuner.
BPM to ms for every note value.