VoxaTrace/com.musicmuni.voxatrace.tona.model

Package-level declarations

Types

data class ContourExtractorConfig(val preset: PitchPreset = PitchPreset.BALANCED, val algorithm: PitchAlgorithm = PitchAlgorithm.SWIFT_F0, val sampleRate: Int = 16000, val hopMs: Int = 10, val cleanup: PitchProcessingConfig = PitchProcessingConfig.SCORING, val voiceType: VoiceType = VoiceType.Auto, val quietHandling: QuietHandling = QuietHandling.NORMAL, val strictness: DetectionStrictness = DetectionStrictness.BALANCED)

Configuration for batch pitch contour extraction (ADR-001).

DetectionStrictness

enum DetectionStrictness : Enum<DetectionStrictness>

How strict to be about pitch detection.

HistogramConfig

data class HistogramConfig(val numBins: Int? = null, val density: Boolean = true, val foldOctaves: Boolean = false, val mode: HistogramMode = HistogramMode.DURATION, val smoothSigma: Float? = 5.0f)

Configuration for pitch histogram computation.

HistogramMode

enum HistogramMode : Enum<HistogramMode>

How histogram bins are weighted.

InterpolationMethod

enum InterpolationMethod : Enum<InterpolationMethod>

Interpolation method for silence gap filling.

LabellingConfig

data class LabellingConfig(val windowSeconds: Float = 0.15f, val hopSeconds: Float = 0.03f)

Configuration for mean-pitch labelling of contour segments.

LinearFitConfig

data class LinearFitConfig(val windowSeconds: Float = 1.5f, val breakThresholdSeconds: Float = 1.5f, val hopSeconds: Float? = null)

Configuration for piecewise linear segment fitting.

NoteEvent

data class NoteEvent(val tStartSeconds: Float, val tEndSeconds: Float, val freqHz: Float, val label: String)

A transcribed note: a time-bounded span labelled with a svara and its frequency.

NoteTranscriptionConfig

data class NoteTranscriptionConfig(val vicinityCents: Float = 50.0f, val minDurationSeconds: Float = 0.05f)

Configuration for com.musicmuni.voxatrace.tona.PitchAnalysis.transcribeNotes (ADR-001: companion presets + Builder + .copy()).

OctaveCorrectionConfig

data class OctaveCorrectionConfig(val removeSpuriousJumps: Boolean = true, val snapBackCorrection: Boolean = true, val snapBackThresholdCents: Float = 150.0f, val referencePitchHz: Float = 0.0f, val correctBoundaries: Boolean = true, val boundaryWindowMs: Float = 50.0f, val hopMs: Int = 10)

Configuration for octave correction (ADR-001).

PeakDetectionConfig

data class PeakDetectionConfig(val method: String = "hybrid", val peakAmpThresh: Float = 5.0E-5f, val valleyThresh: Float = 3.0E-5f, val lookahead: Int = 20, val avgIntervalHint: Float? = null, val minPeakAreaFraction: Float = 0.0f)

Configuration for peak detection in a pitch histogram.

PeakStats

data class PeakStats(val referenceInterval: Float, val peakPosition: Float, val peakAmplitude: Float, val mean: Float, val median: Float, val stdDev: Float, val variance: Float, val coeffOfVariation: Float, val skewness: Float, val kurtosis: Float, val pearsonSkew2: Float)

Statistical summary of pitch distribution around a detected histogram peak.

PeakStatsCollection

data class PeakStatsCollection(val statsByReferenceInterval: Map<Float, PeakStats>) : Iterable<PeakStats>

Collection of PeakStats keyed by reference interval (cents).

PeakStatsConfig

data class PeakStatsConfig(val maxPeakwidthCents: Float = 50.0f, val minPeakwidthCents: Float = 25.0f, val symmetricBounds: Boolean = true)

Configuration for per-peak statistics computation.

PitchAlgorithm

enum PitchAlgorithm : Enum<PitchAlgorithm>

Pitch detection algorithm backend.

PitchContour

data class PitchContour(val samples: List<PitchPoint>, val sampleRate: Int = 16000, val hopSize: Int = 0)

A sequence of PitchPoints representing a pitch contour over time.

PitchDetectorConfig

data class PitchDetectorConfig(val algorithm: PitchAlgorithm = PitchAlgorithm.YIN, val bufferSize: Int = 1024, val hopSize: Int = 160, val tolerance: Float = 0.15f, val minFreq: Float = 80.0f, val maxFreq: Float = 1000.0f, val amplitudeGateDb: Float = -40f, val confidenceThreshold: Float = 0.75f, val enableSmoothing: Boolean = false, val enableOctaveCorrection: Boolean = false, val smoothingWindowSize: Int = 5, val octaveThresholdCents: Float = 150.0f, val swiftF0BatchSize: Int = 2560)

Configuration for pitch detector.

PitchHistogram

data class PitchHistogram(val binCenters: FloatArray, val values: FloatArray, val tonicHz: Float, val isDensity: Boolean, val isFolded: Boolean, val mode: HistogramMode)

Pitch histogram with musical metadata. Wraps Histogram from common.

PitchPoint

data class PitchPoint(val pitch: Float, val confidence: Float, val timeSeconds: Float = 0.0f)

A pitch detection result or timestamped pitch sample.

PitchPreset

enum PitchPreset : Enum<PitchPreset>

Pre-configured pitch detection presets.

PitchProcessingConfig

data class PitchProcessingConfig(val fixOctaveErrors: Boolean = true, val removeSpuriousJumps: Boolean = true, val octaveThresholdCents: Float = 150.0f, val referencePitchHz: Float = 0.0f, val fixBoundaryOctaves: Boolean = true, val boundaryWindowMs: Float = 50.0f, val smoothPitch: Boolean = true, val smoothingWindowSize: Int = 7, val removeBlips: Boolean = true, val minimumNoteDurationMs: Float = 80.0f, val hopMs: Int = 10)

Configuration for batch pitch processing pipeline (ADR-001).

QuantizationConfig

data class QuantizationConfig(val slopeThresholdCentsPerSec: Float = 150.0f, val maxDeviationCents: Float = 50.0f, val medianFilterWindowSamples: Int = 7, val applyMedianFilter: Boolean = true, val minSegmentDurationMs: Int? = null)

Configuration for pitch quantization (snapping to target intervals).

QuietHandling

enum QuietHandling : Enum<QuietHandling>

How to handle quiet audio.

RealtimePitchProcessingConfig

data class RealtimePitchProcessingConfig(val enableSmoothing: Boolean = true, val smoothingWindowSize: Int = 5, val enableOctaveCorrection: Boolean = true, val octaveThresholdCents: Float = 150.0f, val referencePitchHz: Float = 0.0f)

Configuration for realtime pitch processing (ADR-001).

SvaraTemplate

data class SvaraTemplate(val cents: FloatArray, val symbols: List<String>)

A raga's measured intonation grid: the cents (relative to tonic) of each svara the raga uses, with its label.

SvaraTemplateConfig

data class SvaraTemplateConfig(val smoothSigmaCents: Float = 15.0f, val peakToleranceCents: Float = 50.0f, val octaveSpan: Int = 1)

Configuration for com.musicmuni.voxatrace.tona.PitchAnalysis.computeSvaraTemplate (ADR-001: companion presets + Builder + .copy()).

TonalSegment

data class TonalSegment(val startSeconds: Float, val endSeconds: Float, val label: String? = null, val meanCents: Float? = null, val slopeCentsPerSec: Float? = null, val interceptCents: Float? = null)

A labelled time-span from melodic transcription.

VoiceType

sealed class VoiceType

Voice type presets for different singing styles and voice ranges.