PHONE RECORDING WORKFLOW

Voice Memo to MIDI

Save the memo as an audio file your browser can decode, then upload it here. A dry recording with one hummed, sung, or played note at a time is the clearest input. Do not rename a file extension to make it look supported. If the browser cannot decode the memo, export a real MP3 or WAV copy from the recording app, then retry.

File: use a real supported audio format
Performance: one note at a time
Room: reduce music and echo
Retry: change one control only

Voice memo workflow from phone recording through browser decode, note preview, and MIDI download — The file must decode before pitch detection begins; a renamed extension does not change the audio encoding.

PROCESS ON THIS DEVICE

Upload or record one melody

Choose a saved memo, or record a new clear melody with the microphone button. The audio is processed in this browser.

INPUT MP3 / WAV / M4A / FLAC / OGG → OUTPUT .MID

Drop one audio file here

or click to choose audio from your device

Choose audioMP3, WAV, M4A, FLAC, or OGG · one file · up to 50 MB / 5 minutes

Or capture one clear melody with your microphone.

Your audio stays on your device. All processing happens in this browser. No file is uploaded to a server.

No sign-upOne file at a timeNo watermark

What happens	Check first	Next controlled action	Stop condition
The memo cannot be opened	Confirm its real format, not only its filename	Export a genuine MP3 or WAV copy from the source app	The browser audio player can load and play the file
Background music becomes notes	Listen for accompaniment under the voice	Use a dry solo memo or an isolated recording	The preview follows the voice rather than the background
Breaths create short fragments	Locate fragments beside audible breaths	Raise threshold or Minimum note, one at a time	Breath fragments disappear and sung note attacks remain
Soft intended notes disappear	Compare the missing spot with the waveform playback	Lower threshold one step	The intended note returns without extra room noise
Several voices overlap	Identify moments with simultaneous pitches	Record or isolate one melodic line	Only one intended pitch dominates each moment

Voice memo signal check separating the intended melody, breath noise, room noise, and background music — Listen to the original memo and name the unwanted source before changing a detection control.

FILE GATE

Decode comes before detection

The browser must decode the audio file before this script can analyze it. Changing a filename extension does not change the underlying encoding.

LOCAL PATH

The file stays in the browser

The page reads the selected file into the browser audio context. It does not upload the memo to a server or require a user account.

VOICE SOURCE

One clear note at a time

Humming or singing a single melodic line gives the detector one intended pitch per moment. Harmony, backing tracks, and room playback add competing patterns.

RETRY RULE

Name the symptom before changing a control

Use threshold for quiet unwanted sounds and Minimum note for brief fragments. Hold the memo fixed and revert unsuccessful changes.

QUICK ANSWERS

Voice memo-to-MIDI questions

Can I turn a voice memo into MIDI?

Yes, if the browser can decode the file and the memo contains a clear single-note melody. The result is detected MIDI note events, not the original voice sound.

Does this page upload my voice memo?

No. The selected audio is decoded and analyzed in this browser. The page does not send the file to a server.

Why will my M4A voice memo not open?

The extension is accepted, but browser decoding still depends on the actual encoding. Export a real MP3 or WAV copy from the recording app rather than renaming the file.

How do I stop breaths from becoming notes?

Use the same memo and test one slightly higher silence threshold. If fragments are very brief, revert and test Minimum note separately.

Can I hum a melody directly into the page?

Yes. Use Start recording, allow microphone access when you choose to do so, and hum one clear note at a time away from background music.

Download the voice memo checklist (.csv)