Audio to MIDI · online tool and model guide

Input

Full songs, instrumental ensembles, or multi-instrument recordings; runs the MIROS MusicFM multi-decoder route

Workflow

1. Upload the recording to transcribe at its original speed and rhythm.
2. Compare each MIDI part with the recording for missing notes, wrong pitches and instrument assignments.
3. Import the result into your DAW to fix velocity, quantization, and tracks

Parameter guide

• Six-stem pre-split: BS-RoFormer separates the audio before each stem is transcribed. It returns individual and merged MIDI, without downloadable separated audio.

What to check

• Keep audio with a stable tempo and avoid long silent stretches
• Switch to YourMT3+ to compare transcriptions of the same recording.
• For solo piano, switch to a piano-specialist model instead

Output

MIROS-generated MIDI for multi-instrument transcription review and DAW cleanup

Input

Solo piano, practice recordings, or a single keyboard performance.

Workflow

1. Upload solo piano audio and run the selected piano model.
2. Download the MIDI and review pitches and note durations in a piano roll or score editor.

Parameter guide

• Solo piano: these models are intended for clean piano or keyboard recordings.

What to check

• Best for solo piano; not recommended for complex ensembles
• Heavy pedal or reverb may need more cleanup afterwards

Output

Piano MIDI whose note offsets follow the No Pedal Ext/key-release target; use it to inspect physical note boundaries rather than pedal-extended acoustic duration.

Input

Clear recordings of solo piano or a single keyboard instrument.

Workflow

1. Upload solo piano audio and run the selected piano model.
2. Download the MIDI and review pitches and note durations in a piano roll or score editor.

Parameter guide

• Solo piano: these models are intended for clean piano or keyboard recordings.

What to check

• Best for solo piano; not recommended for complex ensembles
• Heavy pedal or reverb may need more cleanup afterwards

Output

Piano MIDI generated by TransKun V2 Aug, for comparison with V2 results.

Input

Clear recordings of solo piano or a single keyboard instrument.

Workflow

1. Upload clean piano audio with clear dynamics
2. Compare the MIDI with the recording, checking bass notes, chords and repeated-note lengths.
3. Import into your DAW or notation software for fine editing

Parameter guide

• Solo piano: these models are intended for clean piano or keyboard recordings.

What to check

• Best for clean piano recordings; avoid noisy environments
• Heavy pedal or reverb may need more cleanup afterwards
• If other instruments are mixed in, separate them first before transcribing

Output

Piano MIDI

Input

Solo piano recordings where pedal timing matters; powered by the ByteDance piano transcription path

Workflow

1. Upload clear solo piano audio
2. Generate MIDI with the ByteDance pedal-aware model
3. Inspect pedal controller events in your DAW or notation software

Parameter guide

• Solo piano: these models are intended for clean piano or keyboard recordings.
• Generates piano MIDI with sustain-pedal events (CC64).

What to check

• Use clean solo piano sources for the most reliable pedal events
• Dense reverb or room noise can still need manual cleanup
• Best for solo piano; not recommended for complex ensembles

Output

Generates piano MIDI with sustain-pedal events (CC64).

Input

Supports inputs in formats such as mp3 / wav / flac / ogg / m4a / aac / aif / aiff / mp4 / mov / mkv

Workflow

1. Upload a full mix and select instruments or automatic detection.
2. Compare each MIDI track with the recording, correct missing or wrong notes, then download.

Parameter guide

• Choose a MIDI tempo setting: follow changes in the original song or use a single BPM. Note timing stays unchanged when quantization is off.
• Quantize grid: controls beat-snapping precision.
• Post-processing: quantization, dedupe, velocity smoothing, and polyphony limits all affect cleanliness.

What to check

• More complex arrangements need more manual note correction afterwards

Output

MIDI

Input

Upload solo singing, humming or an isolated lead vocal. If you only have a finished song, choose “Full song” and the lead vocal will be separated before transcription.

Workflow

1. Choose the upload type, then add solo singing, a vocal stem or a full song. Full songs are separated first.
2. Start with the default model; use the four-version comparison when one result is hard to judge.
3. Switch between the original vocal and each piano roll, correct the chosen MIDI, then download it.

Parameter guide

• Choose “Solo vocal / vocal stem” for a cappella or already separated vocals. Choose “Full song” to separate the lead vocal first.
• Leave this blank to use GAME's official language ID 0. This is not automatic language detection; select a language when you know what is being sung.
• GAME boundary-segmentation confidence threshold.
• GAME boundary search radius in seconds.
• Starting noise time used by GAME’s official D3PM schedule.
• GAME denoising steps: 8 by default; accepts 1–128.
• GAME note-estimation confidence threshold.
• Analyze the original beats after transcription and write the tempo to MIDI, preserving note timing.
• ROSVOT note-boundary threshold: 0.85 by default; accepts 0–1.
• ROSVOT word-boundary threshold: 0.5 by default; accepts 0–1.

What to check

• This tool follows the sung lead melody; it does not rebuild drums, chords or a complete arrangement.
• GAME exposes language and segmentation controls. ROSVOT's released checkpoint is mainly trained on Mandarin singing.
• The four-version comparison runs four separate conversions and returns four independent MIDI files; it does not fuse the models.

Output

MIDI

Input

Supports inputs in formats such as mp3 / wav / flac / ogg / m4a / aac / aif / aiff / mp4 / mov / mkv

Workflow

1. Upload a recording containing just one instrument or voice.
2. Generate MIDI with Basic Pitch and adjust onset, minimum note length and pitch range to reduce false notes.

Parameter guide

• Basic Pitch onset confidence threshold.
• Basic Pitch frame-level activation threshold.
• Shortest note retained by Basic Pitch, in milliseconds.
• Discard Basic Pitch notes below this frequency; leave blank for no limit.
• Discard Basic Pitch notes above this frequency; leave blank for no limit.
• Keep separate pitch-bend curves for simultaneous Basic Pitch notes.
• Enable Basic Pitch’s official Melodia-inspired post-processing.
• Analyze the original beats after transcription and write the tempo to MIDI, preserving note timing.

Output

MIDI / CSV

Input

Supports inputs in formats such as mp3 / wav / flac / ogg / m4a / aac / aif / aiff / mp4 / mov / mkv

Workflow

1. Full mixes go through the model’s drum-extraction pipeline first. Isolated drums go directly to transcription or drum-stem separation.
2. Download drum MIDI and review kick, snare, hi-hat and other kit parts in a drum editor.

Parameter guide

• Enable for a full mix to separate the drums first; disable for an isolated drum track.
• Drum onset detection threshold for ADTOF Plus; leave blank to use the model default.
• Adjust detection thresholds separately for kick, snare, hi-hat, toms and cymbals.
• Time resolution, 100 FPS by default; changing it may affect transcription quality.
• Analyze the original beats after transcription and write the tempo to MIDI, preserving note timing.

Output

MIDI

Input

Supports inputs in formats such as mp3 / wav / flac / ogg / m4a / aac / aif / aiff / mp4 / mov / mkv

Workflow

1. Upload solo guitar and choose a model for its acoustic or electric tone.
2. Download tablature and MIDI, then check strings and frets against the recording.

Parameter guide

• Generates playable fingering from CrispASR predictions, with a maximum hand span of 5 frets.
• Analyze the original beats after transcription and write the tempo to MIDI, preserving note timing.

What to check

• If other instruments are mixed in, separate them first before transcribing

Output

MIDI / JSON / CSV

Input

Supports inputs in formats such as mp3 / wav / flac / ogg / m4a / aac / aif / aiff / mp4 / mov / mkv

Workflow

1. Upload song audio.
2. Identify the chord progression with ChordMini BTC / ChordMini 2E1D / Omnizart and download chord MIDI and a timestamped chord table.

Parameter guide

• Analyze the original beats after transcription and write the tempo to MIDI, preserving note timing.

Output

MIDI / CSV / JSON

Input

Existing .mid/.midi files with text meta events that need encoding cleanup

Workflow

1. Select an existing .mid or .midi file in the browser
2. Keep auto-detect on or choose the source text encoding manually
3. Choose the target encoding, convert locally, and download the rewritten MIDI file

Parameter guide

• Source encoding: used when auto-detect is off; applies only to MIDI text meta events.
• Target encoding: browser-local mode writes the selected target encoding; UTF-8 remains the recommended default for modern DAWs and editors.
• Auto-detect: chooses the likely source text encoding from the MIDI text meta-event payloads.

What to check

• Only MIDI text meta events 0x01-0x07 are re-encoded
• Notes, tempo, controllers, timing, velocities, SysEx, and audio are not changed
• If you need MIDI from audio, use a transcription tool instead

Output

Browser-local MIDI output in the selected encoding; notes, timing, tempo, velocity, and controllers are kept unchanged