WebVTT to CutMarks Operation

ID: webvtt-to-cutmarks

Description

This operation parses a WebVTT Subtitle File and generates a JSON of CutMarks based on the Subtitle Timestamps. It's primary intention is to be used as a silence detection based on available subtitle information.

With the CutMarks to SMIL workflow operation this JSON can be converted to a SMIL that can be used in the Video Editor for cutting.

Parameter Table

Tracks are assumed to start at 0.

Configuration Keys	Default value	Example/Other Possible values	Description
source-flavor	None, must be set	captions/vtt+en	The Flavor of the WebVTT Subtitle File to parse, used for creating the CutMarks.
target-flavor	None, must be set	cut-marks/json	The Flavor of the output JSON, which contains the CutMarks based on the subtitle information.
min-time-silence-in-ms	0	5000	Time (in ms) between two subtitles for them to be considered seperate cutting sections, otherwise they are merged. Basically the minimum length of a silent section for it to be cut out.
buffer-time-around-subtitle	0	1500	How much buffer time (in ms) to add before and after a subtitle/non-silent section.
track-flavor	Optional/Empty	presenter/source	The flavor of the track related to the WebVTT, used for determining the end time of the video.
start-treatment	IGNORE	USE_FOR_MIN_TIME/ALWAYS_INCLUDE	How to treat the beginning of the video in relation to the subtitle sections.
end-treatment	IGNORE	USE_FOR_MIN_TIME/ALWAYS_INCLUDE	How to treat the end of the video in relation to the subtitle sections. (If not IGNORE, Needs track-flavor to be set)

Operation Example

  - id: webvtt-to-cutmarks
    description: Use WebVTT as a silence detection
    configurations:
      - source-flavor: captions/vtt+en
      - target-flavor: cut-marks/json
      - min-time-silence-in-ms: 5000
      - buffer-time-around-subtitle: 1500
      - track-flavor: presenter/source
      - start-treatment: USE_FOR_MIN_TIME
      - end-treatment: USE_FOR_MIN_TIME