Creating audio-only outputs

Focus mode

Creating audio-only outputs - MediaConvert

You can use AWS Elemental MediaConvert to create outputs that contain only audio, without video. With audio-only outputs, MediaConvert supports a more limited number of codec and container combinations for input and output files.

The restrictions and procedures in this chapter apply to outputs that don't have video in the container. This includes the following:

Outputs in File output groups that don't have video included
Streaming HLS output groups that contain only audio outputs
Streaming DASH output groups that contain only audio outputs

When you set up streaming output packages that contain audio, video, and captions, you create separate outputs for each element inside the output package. These are not audio-only outputs as described in this chapter. For more information about setting up streaming outputs, see Creating outputs in ABR streaming output groups.

You set up an audio-only output in the same way that you set up an output that contains video, except that you don't include video or captions.

MediaConvert generates the following files for audio-only outputs:

File output groups: One separate audio-only file for each output.
HLS output groups: A single rendition in the ABR stack for each output.
DASH ISO output groups: A single rendition in the ABR stack for each output.

Note

For AAC streaming outputs, the initial segment is longer in duration than the others. This is because, with AAC, the initial segment must contain silent AAC pre-roll samples before the audible part of the segment. MediaConvert accounts for these extra samples in the timestamps, so the audio plays back correctly.

To create an audio-only output (console)

To confirm that MediaConvert supports your input files, consult the input table in Supported output formats for audio-only workflows.
Set up your job as described in Tutorial: Configuring job settings, but with the following differences:
- Remove the Video tab in the Encoding settings section of your output.
- Choose a supported output container and audio codec from the output table in Supported output formats for audio-only workflows.
- Include only one group of audio settings per output. That is, don't choose Add audio to create an Audio 2 tab under Encoding settings.
If your outputs are in an HLS output group, choose a container for the audio-only output.

Under Output settings, in the Audio section, for Container for audio-only output, choose MPEG-2 Transport Stream to create a file in an MPEG2-TS container. Keep the default value Automatic to create a raw audio-only file with no container.

To create an audio-only output (API, SDK, and AWS CLI)

To confirm that MediaConvert supports your input files, consult the input table in Supported output formats for audio-only workflows.
Set up your JSON job specification. Either manually edit your JSON file, or use the console to generate it as follows:
1. Follow the previous procedure for the console.
2. In the Job pane on the left, under Job settings, choose Show job JSON.
The JSON job specification for audio-only jobs differs from standard jobs as follows:
- Exclude each instance of VideoDescription and its children from the Outputs portion of your job JSON.
- For each output, include only one child group of audio settings under AudioDescriptions.
- For audio-only outputs in an HLS output group, specify a container for the audio-only output. Under Outputs, OutputSettings, HlsSettings include the property HlsAudioOnlyContainer. Set it to M2TS to create a file in an MPEG2-TS container. Set it to AUTOMATIC to create a raw audio-only file with no container. AUTOMATIC is the default behavior.