VoxaTrace/com.musicmuni.voxatrace.calibra/CalibraVAD/Companion/create

create

fun create(config: VADConfig, modelProvider: VADModelProvider): CalibraVAD

Create VAD with config and model provider.

val config = VADConfig.Builder()
    .preset(VADConfig.SPEECH)
    .threshold(0.4f)
    .build()
val vad = CalibraVAD.create(config, VADModelProvider.Speech { ModelLoader.loadSpeechVAD() })

Return

Configured CalibraVAD instance

Parameters

config

VAD configuration

modelProvider

Type-safe model provider that supplies the model

fun create(modelProvider: VADModelProvider): CalibraVAD

Create VAD with model provider (convenience, uses default config for backend).

Backend is inferred from the provider type. Each backend has appropriate defaults baked in. Use VADConfig.Builder for customization.

// GENERAL backend (no model required)
val vad = CalibraVAD.create(VADModelProvider.General)

// SPEECH backend (Silero)
val vad = CalibraVAD.create(VADModelProvider.Speech { ModelLoader.loadSpeechVAD() })

// SINGING_REALTIME backend (SwiftF0)
val vad = CalibraVAD.create(VADModelProvider.SingingRealtime { ModelLoader.loadSingingRealtimeVAD() })

Return

Configured CalibraVAD instance

Parameters

modelProvider

Type-safe model provider that determines backend