The main components provided by RAAR, enclosed by the gray box.
- Importer: Monitors a directory for new recordings and imports them into the archive.
- Downgrader: Reduces the quality of the archive files after defined periods of time.
- Rest API: Provides a JSON API to browse, listen to and configure the archive.
- Database & Filesystem: Persistence to store the files and the corresponding metadata.
These external services are required to run RAAR and must be integrated additionally.
- Recorder: An recorder that writes audio files into a given directory.
- Airtime: The database of an Airtime instance containing the metadata of the shows and broadcasts.
- Web UI: An arbitrary client accessing the Rest API to provide a nice user interface for internal and external listeners as well as admins configuring RAAR.
RAAR distinguishes two key words: Shows and Broadcasts. In the naming of RAAR, a show is the recurring container for singular broadcasts. So a broadcast is the transmission of a certain show for a single specific date and time. E.g. there may be the 'One and only morning show', with broadcasts every weekday from 8 o'clock to 11 o'clock.
A broadcast may have multiple Audio Files with different formats/bitrates. Each audio file contains the entire broadcast (the duration of the audio file equals the duration of the broadcast).
Played Tracks may be archived additionally, independent of Broadcasts but associated to them for convenience. Currently, they are not imported from Airtime, but may be managed with the API.
The archive configuration mostly defines various audio formats to create when importing a broadcast and when they are downgraded. An audio format is defined with three values:
- codec: The audio encoding, e.g. mp3, ogg, wav, ...
- bitrate: The bitrate in kbps, if applicable.
- channels: Number of audio channels, 2 for stereo, 1 for mono.
The configuration of the different formats to archive the various shows/broadcasts is bundled in Profiles. We suggest to only provide a small amount of different profiles. Each show is assigned exactly one profile.
Three models contain the actual configuration: ArchiveFormat with DowngradeAction and PlaybackFormat.
Defines which audio formats should be created when importing a broadcast assigned to a given profile. Only one archive format may be defined per codec and profile.
Additionally, a maximum public bitrate may be defined. If set, this value restricts API access to audio files of the given codec with a higher bitrate to users with a login or access code.
Defines the bitrate and channels the audio files should be downgraded to after a number of months for a given archive format. Multiple downgrade actions with different months/bitrates/channels may exist for a single archive format.
If the bitrate is set to nil, the audio files will be entirely deleted after the defined number of months.
Additionally to the archive formats, a set of playback formats may be defined. They are supposed to provide uniform audio qualities for archive listeners across different shows/broadcasts. As long as there is an archive format for a given codec in the show's profile, audio files for all corresponding playback formats (with equal or lower initial bitrates) are created when a broadcast is imported.
However, when a downgrade action takes place for a given codec, it does not stop from the playback format and does remove all audio files with higher bitrates. When a certain playback format file is requested and not available, the audio file with the same codec and the highest bitrate below the playback format bitrate is served.
The entities described above are represented by the following database model:
