如果有多个通道，则合并，然后从音频文件中获取采样长度并将其保存到s3(If multiple channels, merge then take sample length from audio file and save it to s3)

 我正在使用transloadit从视频文件中提取音频，然后将其保存到S3。 这很好，但我想知道如何： 
 
 
 
 检查文件是否有多个频道，然后挤压一个按照https://transloadit.com/demos/audio-encoding/merging-multiple-audio-streams/ - 我需要检查这个或做我默认为使用这个机器人？ 
 
 
 从音频文件中提取一个小样本 - 并将其另存为一个单独的文件。 
 
 
 例如，我有一个2h的音频文件，我想从中获取5％的长度并将其保存为sample.mp3 
 
 在ffmpeg中，我可以剪切： 
 
ffmpeg -ss 0 -t 30 -i original.mp3 sample.mp3
 
 但我不确定如何链接这个工作流程，这是我到目前为止： 
 
const opts = {
  params: {
    notify_url: `${ process.env.SELF }/services/trans/${ jwToken }`,
    steps:      {
      import: {
        robot:  '/s3/import',
        use:    ':original',
        bucket: process.env.S3_INGEST,
        path:   ingest.key,
        key:    process.env.AWS_ID,
        secret: process.env.AWS_SECRET,
      },
      encode: {
        robot:        '/audio/encode',
        use:          'import',
        ffmpeg_stack: 'v2.2.3',
        preset:       'aac',
        ffmpeg:       {
          ab:  '128k',
        },
      },
      export: {
        robot:   '/s3/store',
        use:     'encode',
        bucket:  s3Export,
        path:    `${ prefix }/${ token }.m4a`,
        headers: {
          'Content-Type': 'audio/mp4',
          'x-amz-server-side-encryption': 'AES256',
        },
        key:    process.env.AWS_ID,
        secret: process.env.AWS_SECRET,
      },
    },
  },
};
 
 在文档中， https://transloadit.com/docs/conversion-robots/我看不到如何做到这一点？ 
 
 任何意见是非常感谢。 

I am using transloadit to extract the audio from a video file, which is then saved to S3. This works great, but I wanted to know how to:
 
 
 
check if the file has multiple channels and then squash it inot one as per https://transloadit.com/demos/audio-encoding/merging-multiple-audio-streams/ - do I need to check for this or do i default to use this robot?
 
 
extract a small sample from the audio file - and save this as a separate file.
 
 
For example, I have a 2h audio file from which I want to take 5% of the length and save this as sample.mp3
 
In ffmpeg, i can cut:
 
ffmpeg -ss 0 -t 30 -i original.mp3 sample.mp3
 
but I am unsure how to chain this workflow, here is what i have thus far:
 
const opts = {
  params: {
    notify_url: `${ process.env.SELF }/services/trans/${ jwToken }`,
    steps:      {
      import: {
        robot:  '/s3/import',
        use:    ':original',
        bucket: process.env.S3_INGEST,
        path:   ingest.key,
        key:    process.env.AWS_ID,
        secret: process.env.AWS_SECRET,
      },
      encode: {
        robot:        '/audio/encode',
        use:          'import',
        ffmpeg_stack: 'v2.2.3',
        preset:       'aac',
        ffmpeg:       {
          ab:  '128k',
        },
      },
      export: {
        robot:   '/s3/store',
        use:     'encode',
        bucket:  s3Export,
        path:    `${ prefix }/${ token }.m4a`,
        headers: {
          'Content-Type': 'audio/mp4',
          'x-amz-server-side-encryption': 'AES256',
        },
        key:    process.env.AWS_ID,
        secret: process.env.AWS_SECRET,
      },
    },
  },
};
 
in the docs, https://transloadit.com/docs/conversion-robots/ i can't see how to do this?
 
any advice is much appreciated.

原文：https://stackoverflow.com/questions/43006931