Private research repository: https://github.com/Speexvocon/dm32-p25. Goal: assess and implement conventional P25 Phase 1 C4FM voice TX/RX on the DM32, with a selectable IMBE codec alongside the existing DMR path.
Current status: modulation registers, speech PCM and PHY interfaces mapped; host IMBE encode/decode exercised. Continuous flat RX and P25-compliant RF are unverified. No working native P25 build is claimed here.
- Roadmap and acceptance gates
- Emulation and replay strategy
- Baseline validation
- Hardware, modulation and codec evidence
- Firmware provenance
- Third-party notices
PowerShell, Python 3, Git and g++ are required for host work. Ghidra 12.0.2,
JDK 21 and the C-SKY module described in research/README.md are needed for
binary analysis. Firmware inputs and installed tools are local, ignored files;
the research report records their origins and hashes.
git clone https://github.com/Speexvocon/dm32-p25.git
cd dm32-p25
.\scripts\clone-upstreams.ps1
.\research\run_codec_probe.ps1The codec probe uses the tracked, pinned source snapshot. external/ holds real
Git clones for exploring/building upstream projects; commit pins are versioned in
research/references/UPSTREAMS.json. Do not edit the snapshot in place when making
an implementation: add project code separately and record upstream changes explicitly.
main holds reviewed milestones; use codex/<topic> branches for development.
Commit source, fixtures, small reports and provenance. Keep bulk captures, firmware
downloads, executables and Ghidra databases outside Git. Capture manifests should
record SHA-256, sample rate/format, settings, hardware revision and firmware commit.
Tag releases only after the corresponding roadmap gate has passed; no firmware
release tag exists yet.