Skip to content

Traktor Broadcast

Native Instruments Traktor Pro does not expose a track API or database that third-party software can poll. Instead, it has a built-in broadcasting feature intended for live streaming to SHOUTcast and Icecast servers. Now Playing runs a local broadcast receiver on your machine and reads track metadata out of the stream. The integration is implemented in the traktor-connect package.

Traktor’s broadcast output is a standard OGG Vorbis audio stream delivered over TCP. The format includes Vorbis comments — metadata tags that Traktor populates with the current track’s information as it plays.

  1. Now Playing starts a TCP server on port 8000 (configurable).
  2. In Traktor, broadcasting is configured to point at 127.0.0.1:8000 with the format set to Ogg Vorbis.
  3. When Traktor starts broadcasting, it opens a TCP connection and begins streaming OGG frames.
  4. Now Playing parses the OGG container and decodes the Vorbis comments from each metadata packet.
  5. When the track changes, Traktor sends a new Vorbis comment block and Now Playing emits a track event.

Vorbis comments are free-form key/value strings. Traktor sends:

FieldVorbis Comment
TitleTITLE
ArtistARTIST
AlbumALBUM
BPMBPM
GenreGENRE
DurationEncoded in the Vorbis stream header
  • Preferences → Broadcasting
    • Server: 127.0.0.1
    • Port: 8000
    • Format: Ogg Vorbis
    • Mount point: any value (Traktor requires it; Now Playing ignores it)

Because Traktor does not expose a track database or a network protocol, the broadcast output is the only stable, officially supported way to read the current track. It is the same mechanism DJs have used for years to auto-update SHOUTcast metadata on their streams.

The broadcast stream reports track metadata only. Controller state (faders, EQ, crossfader) is not in the stream. If you want on-air detection to work during transitions, enable the MIDI Bridge so the mix processor has fader data.

The broadcast stream contains actual audio as well as metadata. Now Playing discards the audio packets — only the metadata is used.