A simple one, this. Firstly, get to where rhythmbox stores its data:

# cd ~/.local/share/rhythmbox

Then, do this:

# grep podcast-feed rhythmdb.xml -A 5 | grep location | sed -e 's/<location>//' | sed -e 's/<\/location>//'

You should end up with a list of URLs which your podcast feeds are served from.

2 thoughts on “find podcast feeds in rhythmbox’s database

  1. Bonjour from Paris,

    Indeed I am a french new user of the Rhythmbox application on Linux, and I found your answer regarding the way to generate a list including the URL’s which actually feed the Rhythmbox “Podacst” activity.
    But, would you mind to tell me how to also generate in addition, in this previous list, the titles and descriptions relevant to every record
    (actually the “Mp3” files from classic Radio station broadcasts)
    provided by each Podcast URL?

    Thank you in advance,


    PS: Merry Christmas!

    • Hi Patrice,

      Sorry my response comes so long after you commented.

      Do you mean extracting other data related to each Podcast feed? That could actually be quite simple – just omit everything after the first command, but increase the number of lines extracted to around 16:
      # grep podcast-feed rhythmdb.xml -A 16

      The result of that command gives me output like this:
      <entry type="podcast-feed">
      <artist>Dimitri Larmuseau</artist>
      <description>Sourcetunk will try to demystify the beautiful beast that is Open Source and show the listeners the more practical examples of Open Source and Free Software. It will discuss software for Linux, BSD, MacOSX and Microsoft Windows systems</description>
      <copyright>Dimitri Larmuseau</copyright>

      Then at least you could pick out the relevant information.

      An alternative way to extracting this could be just to open the rhythmdb.xml file into Firefox and do a text search for “podcast-feed”.

      Or … did you mean get Rhythmbox to display the title of each individual track that the radio station includes in its podcast mp3 file?

      Unfortunately, that won’t be possible, as all data seems to be aggregated into one file with only the single description.

      I’m afraid I don’t know everything about streaming protocols, but my understanding of this is that the radio station can optionally include this metadata in the stream that they “broadcast”.

      If you are looking to temporarily store MP3s that you like that a radio station plays, perhaps you should listen to a live feed instead and record that stream? A better media player in Linux for doing this is Exaile, which I believe has a stream recorder that can split incoming MP3s into separate files. This is the only viable approach to doing that, as far as I know (with a media player that supports recording live streams).

      However, there’s an older discussion on AskUbuntu where someone asks about scrobbling radio streams from Rhythmbox, which links to a discussion on GNOME’s Bugzilla, where the reply is “The shoutcast metadata provided by most streams only includes a single title string which cannot reliably be split into artist and song title fields.” (https://bugzilla.gnome.org/show_bug.cgi?id=376659).

      So I’m not quite sure how Exaile (or another media player with similar capability) actually does this recording trick, unless the metadata is of a more useful format!

      Hope that all helps!

Have your say!