Record one clear pitched source
Place the microphone close enough for a healthy signal, avoid clipping and reduce room reflections. Isolation matters more than extreme sample rates.
In-browser · private · no upload
WAV to MIDI Converter
Turn uncompressed WAV recordings into editable MIDI, MusicXML, PDF notation, ABC and chord charts.
Drop a WAV file here
WAV · up to 10 minutes · 3 minutes on mobile
Or try a sample:
Preserve the detail in an uncompressed WAV recording
WAV is a strong transcription source because it normally stores uncompressed PCM samples. Clear attacks, quiet notes and natural decays reach the decoder without the masking decisions used by MP3 or AAC.
More data is not automatically more accuracy. Tonera converts the decoded source to 22,050 Hz mono because that is the input expected by Basic Pitch. A clean 16-bit/44.1 kHz solo recording is usually more useful than a 24-bit/96 kHz file containing room noise and several overlapping instruments.
High sample rates are handled for you
The browser decodes the WAV and Tonera prepares the mono 22,050 Hz signal expected by the model. There is no need to resample manually.
Use thresholds to tame resonance
Sustain pedals, guitar sympathetic vibration and room echo can create note tails. Minimum duration and frame controls help separate musical notes from residue.
What matters in a WAV before transcription
WAV describes a container, not a guarantee of a clean recording. Bit depth, sample rate, channel layout and the sound in the room all influence what the model can infer.
16-bit/44.1 kHz versus 24-bit/96 kHz
For finished music, 16-bit/44.1 kHz already covers the audible range and gives Basic Pitch enough source detail after resampling. A 24-bit recording provides more headroom during recording and mixing, but the extra precision does not directly become more MIDI notes.
Use 24-bit when it helps capture a clean signal without clipping. Do not upsample a 16-bit file: it creates more samples, not new musical information.
Mono, stereo and room sound
Tonera combines channels to mono for inference. Wide stereo effects can partially cancel or smear a centered melody when collapsed, so a dry instrument bus or mono microphone is often better than a mastered stereo mix.
Room resonance produces energy after the played note ends. Shorten the minimum note duration carefully, increase the frame threshold and compare the visual candidates rather than applying one preset to every recording.
DAW export settings
Export the instrument or vocal bus without a limiter if you can. Leave enough headroom to avoid clipping, keep effects modest and render the exact passage you need. Normalization is optional; a clear signal matters more than maximum loudness.
If your WAV is part of a full arrangement, solo the target track before export. That is more reliable than asking a pitch model to untangle the mastered mix later.
Prepare, transcribe and refine a WAV recording
1
Choose the isolated take or bus
A solo track with moderate headroom gives the model clearer note boundaries.
2
Let Tonera prepare the signal
Upload the WAV directly; browser decoding and model resampling happen automatically.
3
Correct resonance before export
Adjust onset, frame and minimum-duration controls, then export the cleaned notes.
Lossless does not mean separated
WAV preserves the recording, including its problems. Clipping, fan noise, reverb and overlapping instruments remain present. Use a dry isolated source or high-precision stem separation when the target part is buried in a mix.