136-174 / 400-470 MHz
DMR
Who called whom on which talkgroup, on both time slots, with speech through a source build carrying the ambe feature.
DMR puts two conversations in one 12.5 kHz channel by splitting it in time, and every burst names the caller, the group and the slot. That metadata is the useful half: a fleet becomes a list of radio ids and the groups they talk on, without a word of speech being decoded.
Speech itself needs the AMBE vocoder, whose algorithms are patent encumbered, so the published builds leave it out. A build from source with the ambe feature decodes it. Without that feature the call still appears with its caller, its group and its length.
Checked against off-air captures, asserted against what the transmission itself says rather than against another decoder.
What comes out
- Source and destination radio id, and whether it is a group or private call
- Talkgroup, colour code and time slot
- Call length, and whether the call is enciphered
- Speech, with a source build carrying the ambe feature
Speech needs a source build with the ambe feature; the published binaries decode everything but the audio.
$ waveshark --tune 439.4 --callsRead next
Something did not decode?
Record the burst with --record and open an issue with the capture. A capture that decodes becomes a test fixture.
Open an issue