VideoUltra API
    Preparing search index...

    Interface Transcriber

    Turns speech into timed words. Implement this to use your own speech service.

    interface Transcriber {
        languages?: readonly string[];
        takesVocabulary?: boolean;
        check?(language: string): Promise<TranscriberCheck>;
        dispose?(): void;
        transcribe(request: TranscribeRequest): Promise<TranscribeResult>;
    }
    Index
    languages?: readonly string[]

    Languages it hears (ISO 639-1), plus 'auto' if it can detect the language. The Captions panel offers only these, and shows no choice when there's one. Default: the panel's own list and Detect.

    takesVocabulary?: boolean

    It uses TranscribeRequest.vocabulary: the Captions panel offers a "Names to spell right" field.

    • Optional: whether and how language can be transcribed here, before starting.

      Parameters

      • language: string

      Returns Promise<TranscriberCheck>

    • Frees its model / worker (optional).

      Returns void

    • Parameters

      • request: TranscribeRequest

      Returns Promise<TranscribeResult>