Audio descriptions

AWS Elemental MediaConvert supports two different workflows for including broadcast audio descriptions in your output.

You can mix an audio description with other audio content if your input contains an audio description audio signal and an audio description data stream. An audio description audio signal is a spoken description of a video, made for people who cannot see the visual content. An audio description data stream contains fade and pan data used by an encoder. MediaConvert uses this data stream to temporarily lower the volume of other audio channels while an audio description is active.

If your input already has pre-mixed audio descriptions instead of an audio signal and data stream channel, you can signal it to downstream systems by writing audio description metadata in your output.

For more information about audio descriptions, see BBC WHP 198 and BBC WHP 051.

Requirements for mixing audio descriptions

When you mix audio descriptions, jobs have the following requirements:

Input

You need an input with an audio description audio signal on one audio channel and an audio description data stream on another.

Audio remixing

To mix audio descriptions across multiple audio selectors, for example if you have a sidecar audio file, enable Manual audio remixing in your output.

Otherwise, to apply audio description mixing to a single input audio selector, you can enable Input remix controls in your input instead. Keep in mind that this mixing will not apply across multiple audio selectors or across audio selector groups.

Mixing audio descriptions

The following sections describe how to configure your job settings to mix audio descriptions.

MediaConvert console

To mix audio descriptions in your output by using the MediaConvert console:

1. Open the Create job page in the MediaConvert console.

2. Add an input that has an audio description audio signal and an audio description data stream.

3. Add an output with at least one audio track.

4. In the output audio track, expand Advanced and enable Manual audio remixing.

5. Enter the channel that has your audio description audio signal in Audio description audio channel.

6. Enter the channel that has your audio description data stream in Audio description data channel.

7. Specify the total number of Input channels. For example, if your input has four input channels including left, right, audio description audio signal, and audio description data stream, select 4 .

8. Specify the total number of Output channels. For example, for a stereo output select 2.

9. Specify mixing levels under Channel mapping. For example, to mix an audio description audio signal from channel 3 across left and right channels in a stereo output, and to mute the data stream, enter the following:

   Channel mapping
   Inputs	Outputs
   	L (0)	R (1)
   1	0	-60
   2	-60	0
   3	0	0
   4	-60	-60

   Note

   When played over speakers, audio description data streams sound like noise. Mute the data stream in your output by setting its channel mapping to -60, as shown in the previous channel mapping example table.

API, SDK, or AWS Command Line Interface (AWS CLI)

The following is an excerpt of a job settings JSON that specifies audio description mixing for a stereo output. Note that the audio description audio signal is in input channel 3, and the audio description data stream is in input channel 4:

{
  "Settings": {
    "Inputs": [],
    "OutputGroups": [
      {
        "Name": "File Group",
        "OutputGroupSettings": {
          "Type": "FILE_GROUP_SETTINGS",
          "FileGroupSettings": {}
        },
        "Outputs": [
          {
            "VideoDescription": {},
            "AudioDescriptions": [
              {
                "CodecSettings": {
                  "Codec": "AAC",
                  "AacSettings": {
                    "Bitrate": 96000,
                    "CodingMode": "CODING_MODE_2_0",
                    "SampleRate": 48000
                  }
                },
                "AudioSourceName": "Audio Selector 1",
                "RemixSettings": {
                  "ChannelMapping": {
                    "OutputChannels": [
                      {
                        "InputChannelsFineTune": [
                          0,
                          -60,
                          0,
                          -60
                        ]
                      },
                      {
                        "InputChannelsFineTune": [
                          -60,
                          0,
                          0,
                          -60
                        ]
                      }
                    ]
                  },
                  "ChannelsIn": 4,
                  "ChannelsOut": 2,
                  "AudioDescriptionAudioChannel": 3,
                  "AudioDescriptionDataChannel": 4
                }
              }
            ],
            "ContainerSettings": {
              "Container": "MP4",
              "Mp4Settings": {}
            }
          }
        ]
      }
    ]
  }
}

Requirements for pre-mixed audio descriptions

If you have an input that already has pre-mixed audio descriptions, and does not contain an audio description audio signal or an audio description data stream, set Audio description broadcaster mix to Broadcaster mixed AD.

When you do, MediaConvert writes metadata in your output that signals to downstream systems that it contains broadcaster mixed audio descriptions.

Specifying pre-mixed audio descriptions

The following sections describe how to configure your job settings to write metadata in your output signalling it contains pre-mixed audio descriptions.

MediaConvert console

To write broadcaster mixed audio description metadata in your output by using the MediaConvert console:

1. Open the Create job page in the MediaConvert console.

2. Add an input that has pre-mixed audio description.

3. Add an output with at least one audio track.

4. In the output audio track, set Audio description broadcaster mix to Broadcaster mixed AD.

5. (Optional) If you enable Manual audio remixing, keep Audio description audio channel and Audio description data channel blank, as these channels will not be present in your input.

API, SDK, or AWS Command Line Interface (AWS CLI)

The following is an excerpt of a job settings JSON that specifies pre-mixed audio descriptions for a stereo output:

{
  "Settings": {
    "Inputs": [],
    "OutputGroups": [
      {
        "Name": "File Group",
        "OutputGroupSettings": {
          "Type": "FILE_GROUP_SETTINGS",
          "FileGroupSettings": {}
        },
        "Outputs": [
          {
            "VideoDescription": {},
            "AudioDescriptions": [
              {
                "CodecSettings": {
                  "Codec": "AAC",
                  "AacSettings": {
                    "Bitrate": 96000,
                    "CodingMode": "CODING_MODE_2_0",
                    "SampleRate": 48000,
                    "AudioDescriptionBroadcasterMix": "BROADCASTER_MIXED_AD"
                  }
                }
              }
            ],
            "ContainerSettings": {
              "Container": "MP4",
              "Mp4Settings": {}
            }
          }
        ]
      }
    ]
  }
}