Best Settings for Intelliscore Ensemble MP3 to MIDI Conversion
Converting MP3 to MIDI with Intelliscore Ensemble can yield clean, usable MIDI files when you pick the right settings and follow a careful workflow. Below are practical, step-by-step recommendations and explanations so you get the best possible results for single-instrument and multi-instrument audio.
1. Prepare the source audio
- Use a high-quality MP3 or WAV — prefer WAV (44.1 kHz, 16-bit) if available; higher bitrate MP3s (320 kbps) work better than low-bitrate files.
- Trim silence and noise — cut leading/trailing silence and remove background noise to improve note detection.
- Isolate the part you want — if you only need the piano or guitar track, crop to that section or use an audio editor to reduce other instruments.
2. Project settings in Intelliscore Ensemble
- Input format: Load WAV if possible; otherwise use the highest-quality MP3.
- Sample rate: Keep original sample rate (44.1 kHz) — do not resample unless necessary.
- Channel: Use mono for single-source instruments; stereo is OK for full mixes.
3. Instrument and polyphony settings
- Choose instrument type: Select the closest match (e.g., Piano, Guitar, Bass). Intelliscore tailors detection to the selected instrument.
- Polyphony level:
- For single-melody monophonic sources (vocals, single lead), set to monophonic.
- For piano or full chords, enable polyphonic and set polyphony depth to match complexity (start at 4–8 voices).
- Sensitivity: Start at a medium sensitivity and adjust: higher sensitivity detects quieter notes but may add false positives; lower sensitivity reduces noise but can miss soft notes.
4. Note detection, threshold, and smoothing
- Onset/threshold: Set a moderate onset threshold so brief noise spikes are ignored. If you get many spurious notes, increase the threshold; if notes are missed, decrease it.
- Minimum note length: Set a short minimum for fast passages (e.g., 50–100 ms) and longer (150–300 ms) for sustained instruments to reduce chopping.
- Smoothing / legato detection: Enable smoothing or legato options if available to avoid splitting sustained notes into multiple shorter notes.
5. Tempo, key, and pitch settings
- Tempo: If the track has a steady tempo, set it manually for better alignment. For variable tempo, allow auto-detection but review results.
- Key/scale: Select the correct key if known — this can improve pitch quantization and reduce octave errors.
- Pitch quantization: Use a small quantize grid (1/16–1/8 note) for rhythmic accuracy; for free rubato performances, avoid aggressive quantization.
6. Post-conversion cleanup
- Manual editing: Expect to open the MIDI in a DAW or MIDI editor to:
- Correct misdetected pitches and octaves.
- Merge or split notes that are incorrectly combined.
- Remove spurious short notes and tighten timing.
- Instrument routing: Assign appropriate instruments to tracks (piano, strings, drums) and tweak velocities.
- Humanize: Add slight timing/velocity variations if the MIDI sounds too mechanical after quantizing.
7. Recommended presets for common scenarios
- Solo piano (complex chords): WAV input, polyphonic 8 voices, medium sensitivity, smoothing ON, min note length 120–200 ms, manual tempo.
- Acoustic guitar (strummed chords): WAV input, polyphonic 4–6 voices, slightly higher sensitivity, smoothing ON, min note length 80–150 ms.
- Vocal melody: Mono input, monophonic, high sensitivity, min note length 50–100 ms, key set if known.
- Low-quality MP3s or mixes: Increase threshold, lower sensitivity, focus on short segments, and plan for heavier manual editing.
8. Troubleshooting common problems
- Many extra notes / noise: Lower sensitivity, raise onset threshold, increase minimum note length, denoise audio.
- Missing soft notes: Raise sensitivity, lower threshold, or normalize audio to increase quiet parts’ level.
- Wrong octaves: Enable octave correction