Edgewise API
    Preparing search index...

    Interface VoiceReference

    Reference audio for a cloning model, with the speaker's consent.

    interface VoiceReference {
        consent: { attested: true; by?: string };
        reference: AudioLike;
        sampleRate?: number;
        saveAs?: string;
    }

    Hierarchy (View Summary)

    Index
    consent: { attested: true; by?: string }

    You confirm the speaker agreed to have their voice cloned. Required.

    reference: AudioLike

    3 to 10 seconds of clear speech from one speaker.

    sampleRate?: number

    Sample rate of reference when it is raw samples.

    saveAs?: string

    Keep the voice for later as voice: 'saved:<name>'.