About Practice Tuner Home

Audio Analysis Settings

Pitch Detection Model

Use AI Model (CREPE)

AI model provides more accurate pitch detection but uses more resources

Volume Peak Threshold

Minimum volume to detect as peak (0.001-0.1)

Onset Threshold

Minimum spectral flux for onset detection (0.1–1.0)

Analysis Window (ms)

Time window for analysis (1000–15000ms)

Onset Refractory Period (ms)

Minimum time between onsets (50–500ms). Higher = fewer false positives.

Volume Gate Multiplier

Onset only fires if volume > background × this value (1.0–10.0). Higher = quieter notes ignored.

Calibration Duration (ms)

How long to record ambient noise when calibrating (3000–30000ms)

Noise Subtraction Fraction

Fraction of noise floor subtracted from incoming audio (0–1). 0.8 recommended; 1.0 can cause artifacts.

Look Ahead Time (ms)

How far ahead to show notes (1000-10000ms)

Look Behind Time (ms)

How far back to keep notes visible (1000-15000ms)

Clarity Threshold (Tuner)

Minimum clarity to show tuner marker (0.01-0.2)

EXTRA SETTINGS

Smoothing Factor (Alpha)

Higher values = smoother volume tracking (0.1-1.0)

Max Volume

Maximum expected volume level (0.1-2.0)

Min Peak Separation (ms)

Minimum time between detected peaks (10-500ms)

Min Note Duration (ms)

Shortest valid note duration (50-1000ms)

Threshold Multiplier

Flux threshold = mean + (this × std dev). Higher = less sensitive (1.0–5.0)

FFT History Frames

Number of previous FFTs to display (1-20)

FFT Frame Separation

Frames between saved FFTs (1-10)