This page contains reproducible DMR and AMBE+2 reference vectors for development, interoperability testing, and packet-format validation. The vectors are payload fields only; they are not complete DMRD, Motorola IPSC, or over-the-air call packets.
The following steady-state frames were generated by the IPSC3 OpenDMR software AMBE+2 encoder from a 1,000 Hz sine wave. The source was 8 kHz, signed 16-bit mono PCM with a peak level of -12 dBFS (amplitude 8,231).
| Parameter | Value |
|---|---|
| Source frequency | 1,000 Hz |
| PCM format | 8 kHz, signed 16-bit mono |
| Source peak | -12 dBFS |
| AMBE frame duration | 20 ms |
| Sequence duration | 60 ms, three AMBE frames |
16 30 81 0A 45 14 80
16 30 81 0A 45 15 80
16 30 81 0A 45 14 80
Each line contains 49 data bits packed most-significant-bit first into seven displayed octets. The low seven bits of the final octet are padding and are not part of the 49-bit codec frame.
16 3D 35 77 EC 22 14 8A 29
16 3D 35 77 EC 22 14 8A 2B
16 3D 35 77 EC 22 14 8A 29
This is the canonical A(24) + B(23) + C(25) frame order used at the IPSC3 software-vocoder boundary. It is not the bit order carried by an MMDVM DMRD voice burst.
66 1C 0C B4 54 AC DC 5E 43
66 1C 0C B4 54 AC DC 5E 53
66 1C 0C B4 54 AC DC 5E 43
The three lines form one 27-byte, 60 ms AMBE media sequence. They use the DMR over-the-air bit interleave carried inside MMDVM DMRD voice payloads. A complete DMRD voice burst must still supply its normal sync or embedded signalling, sequence information, and call metadata.
16 30 81 0A 45 14 85 8C 20 42 91 45 61 63 08 10 A4 51 48
This 19-byte field packs the same three 49-bit frames with the separator bits used by Motorola IPSC voice media. It is only the packed voice field, not a complete Motorola IPSC packet.
Repeat the complete three-frame sequence for a continuous test tone. A real call must still include valid call headers, source and destination IDs, timeslot signalling, stream sequencing, and a terminator appropriate to the selected transport.
AMBE+2 is a speech codec rather than a laboratory test-tone codec. A round-trip decode of this vector had a dominant component near 975 Hz, so it is useful for media-path and level testing but must not be treated as a precision 1,000 Hz frequency reference.