Description #

The audio segmentation operator is used to extract audio from audio or video files and split the audio into multiple segments according to specified rules, outputting them to a user-defined storage path.
This operator is primarily used for structured processing scenarios involving long audio or video files, such as audio preprocessing, data segmentation, and training data construction. It supports batch concurrent processing and flexible output path organization.
The current version uses TOS storage addresses as both input and output formats. Support for protocols such as HTTP/HTTPS will be added in future versions.

Key features #

1. Audio/video extraction and segmentation

• The input can be an audio file or a video file.
• When the input is a video file, the audio track is automatically extracted.
• Supports splitting audio into multiple segments based on specified duration rules.
• Agnostic to upstream data types and uniformly outputs multiple audio segments.

2. Supports custom segmentation rules

• Supports configuring the segmentation duration (unit: seconds).
• Suitable for segmenting long audio and long video files.
• If segmentation duration is not specified, the server can process using default strategies.

3. Supports custom output audio formats

• Supports specifying the output audio format.
• Can be configured to common audio formats such as mp3, flac, wav, and more.
• The range of supported formats depends on the underlying FFmpeg codec capabilities.

4. Supports output path templates

• Supports users defining the output path for segmented audio using path templates.
• Output path templates support placeholders (such as {index}) for generating multiple audio segment paths.
• Suitable for organizing segmentation results by sequence, segment number, or business rules.
• Returns a list of all generated output_paths.

5. Supports extended audio encoding parameters

• Supports passing through FFmpeg-level extended parameters.
• Can be used to control audio quality, encoding method, or sampling strategy.
• Meets scenarios with refined requirements for audio processing.

• Billing standards

  Sub-items	Billing standards description
  Billing item	Billing is based on usage calculated from the duration of input audio.
  Billing type	Usage-based billing, unit: CNY/minute, billed hourly based on actual usage.
  Unit price	0.003 CNY/minute

• Billing details
  Billing formula: Total cost = unit price * usage

Interface description #

Call the las_audio_extract_and_split operator to extract audio from a video and split it.

Request parameters #

Response parameters #

Example #

Request example #

# Set INPUT_PATH to the TOS path of the video file saved under this account
export INPUT_PATH="tos://xxxx/sample.mp4"
# Set OUTPUT_PATH_TEMPLATE to a writable TOS path under this account
export OUTPUT_PATH_TEMPLATE="tos://xxxx/{index}/test.wav"
# Initiate an operator service request
curl --location "https://operator.las.ap-southeast-1.volces.com/api/v1/process" \
--header "Content-Type: application/json" \
--header "Authorization: Bearer $LAS_API_KEY" \
--data '{
    "operator_id": "las_audio_extract_and_split",
    "operator_version": "v1",
    "data": {
        "input_path": "$INPUT_PATH",
        "output_format": "wav",
        "output_path_template": "tos://$OUTPUT_PATH_TEMPLATE/{index}/test.wav",
        "split_duration": 15
       }
}'

Response example #

{
    "metadata": {
        "task_status": "COMPLETED",
        "business_code": "0",
        "error_msg": "",
        "request_id": "c7b29d78a99f88beda5497753ed60816"
    },
    "data": {
        "output_paths": [
            "tos://xxx/0/test.wav",
            "tos://xxx/1/test.wav"
        ],
        "metrics": [
            {
                "name": "duration_ms",
                "value": 26633
            }
        ]
    }
}

Error codes #