riffgen/auto-tune ✓❓🖼️🔢 → 🖼️

▶️ 128 runs 📅 Jul 2025 ⚙️ Cog 0.15.9

About

AI pitch correction that auto-tunes any voice.

Example Output

Output

Performance Metrics

121.45s Prediction Time

187.75s Total Time

All Input Parameters

{
  "plot": true,
  "scale": "auto",
  "audio_file": "https://replicate.delivery/pbxt/NIGQ5eDhlVV9rXOcm2RqUcYmjYygQSkorysqKKPFIoWupdld/Falling%20in%20Love%20-%20Rough.m4a",
  "output_format": "wav",
  "adaptive_strength": true,
  "smooth_transitions": true,
  "correction_strength": 1
}

Input Parameters

plot Type: booleanDefault: true: Generate a pitch correction visualization plot.
scale Default: auto: Musical key for correction. 'auto' will detect the key, 'closest' uses chromatic correction.
audio_file (required) Type: string: URL or path to the audio file to be pitch-corrected
output_format Default: wav: Output format for generated audio.
adaptive_strength Type: booleanDefault: true: Adapt correction strength based on voice confidence.
smooth_transitions Type: booleanDefault: true: Smooth transitions between voiced/unvoiced segments.
correction_strength Type: numberDefault: 1Range: 0 - 1: Strength of pitch correction (0.0 to 1.0). Higher values = stronger correction.

Output Schema

Output

Type: array • Items Type: string • Items Format: uri

Example Execution Logs

🎤 Automatically detecting key...
/src/predict.py:17: UserWarning: PySoundFile failed. Trying audioread instead.
y, sr = librosa.load(filepath, sr=None)
/usr/local/lib/python3.11/site-packages/librosa/core/audio.py:183: FutureWarning: librosa.core.audio.__audioread_load
Deprecated as of librosa version 0.10.0.
It will be removed in librosa version 1.0.
y, sr_native = __audioread_load(path, offset, duration, dtype)
⚠️ Key detection failed: No librosa attribute key. Falling back to chromatic correction.
/src/utils.py:151: UserWarning: PySoundFile failed. Trying audioread instead.
y, sr = librosa.load(str(filepath), sr=None, mono=True)
/usr/local/lib/python3.11/site-packages/librosa/core/audio.py:183: FutureWarning: librosa.core.audio.__audioread_load
Deprecated as of librosa version 0.10.0.
It will be removed in librosa version 1.0.
y, sr_native = __audioread_load(path, offset, duration, dtype)
/usr/local/lib/python3.11/site-packages/librosa/core/convert.py:1095: RuntimeWarning: divide by zero encountered in log2
midi: np.ndarray = 12 * (np.log2(np.asanyarray(frequencies)) - np.log2(440.0)) + 69
/usr/local/lib/python3.11/site-packages/numpy/lib/function_base.py:1452: RuntimeWarning: invalid value encountered in subtract
a = op(a[slice1], a[slice2])

Version Details

Version ID: b8198afc90cb2702cad5330980220ad6fff1b465157e71ab2325d4275f58c8a2
Version Created: July 3, 2025

Run on Replicate →