pipertts - v1.3.0
    Preparing search index...

    Interface PiperConfig

    Normalized voice config with defaults applied. Port of PiperConfig in config.py.

    interface PiperConfig {
        defaultSpeakerId: number;
        espeakVoice: string;
        hopLength: number;
        lengthScale: number;
        noiseScale: number;
        noiseWScale: number;
        numSpeakers: number;
        numSymbols: number;
        phonemeIdMap: Record<string, number[]>;
        phonemeType: PhonemeType;
        piperVersion?: string;
        sampleRate: number;
        speakerIdMap: Record<string, number>;
        vowelClusters: Set<string> | null;
    }
    Index
    defaultSpeakerId espeakVoice hopLength lengthScale noiseScale noiseWScale numSpeakers numSymbols phonemeIdMap phonemeType piperVersion? sampleRate speakerIdMap vowelClusters
    defaultSpeakerId: number

    Fallback speaker id when none is requested.

    espeakVoice: string

    espeak-ng voice identifier (for example, "fr").

    hopLength: number

    STFT hop length used for chunk sizing.

    lengthScale: number

    Default speech-rate multiplier (>1 slower, <1 faster).

    noiseScale: number

    Default voice variability.

    noiseWScale: number

    Default timing variability.

    numSpeakers: number

    Number of speakers (1 for single-voice models).

    numSymbols: number

    Number of phoneme symbols.

    phonemeIdMap: Record<string, number[]>

    Phoneme string to ONNX id list mapping.

    phonemeType: PhonemeType

    Phonemizer back-end used by this voice.

    piperVersion?: string

    Piper version that trained the model, when recorded.

    sampleRate: number

    Output sample rate in Hz.

    speakerIdMap: Record<string, number>

    Named speaker to id mapping (multi-speaker models).

    vowelClusters: Set<string> | null

    Vowel clusters treated as single phonemes, if any.

    Settings
    Member Visibility

    On This Page
    Properties