Audio
Prime Video allows discrete audio file delivery of localized dubs or descriptive audio of a program’s content globally. Each language requires delivery of a separate audio asset, and all assets must conform to the video mezzanine delivered for that title. Content partners have the option to send discrete audio to enable additional languages. This is dependent on the original video being semi-textless. When you deliver assets to accompany previously delivered video mezzanines, all additional assets must use the same identifiers as the previously delivered sources. If you require a low-resolution proxy file of the mezzanine previously delivered to Prime Video to conform audio or timed text in additional languages, contact your operations account manager.
Technical guidelines for audio files and audio channel mapping
Prime Video accepts the following audio source formats:
- PCM (ipcm) audio in a QuickTime .mov container
- PCM (ipcm) audio in an MPEG-2 transport stream
- AAC audio in an .mp4 container
- PCM (ipcm) audio in a self-contained .wav file
Prime Video also accepts .wav files for stereo-only sources, due to the 4-GB size limit of this format.
Only deliver fully mixed program tracks. Additional tracks, like music and effects (M&E) and silent (MOS), must be removed prior to delivery.
Audio validation
To minimize delivery of out-of-spec audio assets, you can perform manual or automated quality control testing on audio files prior to delivery. Some of the common issues are as follows:
- One-hour offset in audio files
- Too many audio channels
- Too few audio channels
- Mismatched video and audio durations (times must be within 2.0 seconds of each other)
- Non-standard sample rates
- Long silent periods
- Multiple languages across audio tracks
- Non-program audio included in audio mixes. Music and Effects (M&E), silent tracks (MOS), and commentary tracks must be removed.
Descriptive audio
In our effort to further support blind and visually impaired customers, Prime Video fully supports descriptive audio or narration. Descriptive audio consists of a narrator talking through the presentation, describing what is happening on the screen or stage during the natural pauses in the audio, and sometimes during dialogue if deemed necessary. (For more information about how to write quality audio description, download the Audio Description Style Guide.)
Important: If you deliver a descriptive audio file for a title, Prime Video requires you to provide dialog audio in the same language. This can be delivered either via audio embedded in the video file, or as a discrete dubbed asset. If you deliver descriptive audio without a matching dialog audio file in the same language-locale, Prime Video blocks you from publishing that title. For example, if the main mezz is en-US but the descriptive audio is en-GB, the descriptive audio will be blocked and any subsequent deliveries of any assets will also be blocked.
To deliver descriptive audio, content partners must adhere to the following instructions for MMC packaging.
Audio in Media Manifest Core (MMC)
Type should be Narration in the inventory section. Presentation section should reference audio track as usual.
<manifest:Audio AudioTrackID="md:trackID:org:prodtest:SKU_xxxx:feature.audio.narration.en-US">
<md:Type>Narration</md:Type>
<md:Language dubbed="true">en-US</md:Language>
<manifest:ContainerReference>
<manifest:ContainerLocation>file://TitanFilms_2589723_Feature.mp4**>
<manifest:Hash method="MD5">2d93c7d99f99229b7c08cbf7ab58b94e</manifest:Hash>
</manifest:ContainerReference>
</manifest:Audio>
