If D-STAR was built from a blank sheet for amateur radio, P25 is the opposite: a mode hams never designed and never asked for, borrowed wholesale from the people who key up when the sirens go. Project 25 was built for police, fire, and EMS, by a government-backed standards body, and it reached the ham bands secondhand — one surplus public-safety radio at a time. So why does it belong on a site about ham digital voice? Because the very waveform at the heart of Yaesu System Fusion was proven here first, and the same open-framework-closed-voice tension that runs through every mode on this site runs straight through P25 too. (For the people who built the open-source side of this story, see With Thanks: The Hams Who Built Digital Voice.)
P25 began not in a ham shack but in a meeting of public-safety agencies. In October 1989, APCO — the Association of Public-Safety Communications Officials — joined with state telecommunications directors and several federal agencies (NTIA, the National Communications System, the NSA, and the Department of Defense) in what became the APCO/NASTD/FED agreement. APCO assigned it the next available project number: 25. A steering committee formed in 1990, a memorandum of understanding with the Telecommunications Industry Association followed in 1992, and the standard was published as the TIA-102 series — today running to dozens of separate documents. Federal grant money funded the program for years. The goal was interoperability: police, fire, and EMS agencies whose incompatible radios couldn't talk to each other in an emergency.
This is the first thing to understand about P25. It is a public-safety standard, governed by a public-safety process, designed for mission-critical use. Everything about its presence in amateur radio is a hand-me-down.
P25 comes in two generations. Phase 1 uses FDMA in a 12.5 kHz channel at 9600 bps, carrying the IMBE vocoder (Improved Multi-Band Excitation) at 7200 bps — 4400 bps of voice plus 2800 bps of forward error correction. Its over-the-air modulation is C4FM, four-level continuous-phase FM (with a CQPSK variant used for simulcast systems). Phase 2, which reached the market around 2012, switches to two-slot TDMA to fit two voice paths into the same 12.5 kHz — a "6.25 kHz equivalent" — and replaces IMBE with the more efficient half-rate AMBE+2 codec.
That C4FM detail matters more than it looks. The four-level waveform Yaesu later adopted for System Fusion was pioneered in P25 Phase 1 in the 1990s, years before Fusion existed. The most popular "ham" digital voice mode runs on a modulation first proven for public safety.
Here is the asterisk D-STAR and DMR operators will recognize on sight. Both of P25's vocoders — IMBE and AMBE+2 — come from the same company, Digital Voice Systems, Inc. (DVSI), and both are proprietary and licensed. It's the identical pattern that sits inside D-STAR (AMBE), DMR (AMBE+2), and NXDN (AMBE+2): an open or openly published framework wrapped around a closed voice codec from a single vendor. The one small mercy is that the original IMBE patent has since expired — but in practice P25 voice still means buying DVSI's technology in hardware you can't read.
P25 entered amateur radio the way a lot of good gear does: on the surplus market. Decommissioned commercial and public-safety radios — Motorola's XTS and XTL series, the ASTRO and later APX families, and equivalents from other vendors — can be programmed onto the 2 m and 70 cm ham bands. These are rugged, professionally built radios, and the first reaction most hams have on hearing P25 is how good it sounds. They are also not cheap, which is part of why P25 has always drawn a smaller, more technical crowd than DMR's flood of inexpensive handhelds.
For years the flagship amateur P25 system was P25NX, the P25 Network Exchange — a repeater-linking project clever enough to be shown off in the TAPR booth at the Dayton Hamvention. Rather than classic reflectors, it used a Cisco-style dynamic multipoint VPN and multicast to link repeaters with almost unlimited on-demand talkgroups, an unusual design for ham linking of its day. As the MMDVM era arrived, the center of gravity shifted: Jonathan Naylor (G4KLX) wrote a P25 gateway and reflector, hotspots running Pi-Star and WPSD gained a P25 mode (standard port 41000), and Steve Miller (KC1AWV) — under the handle "nostar" — maintained the open DVReflectors that much of the network now runs on. Worldwide and regional talkgroups (TG 10100 worldwide, 10200 for North America, and others) gave the mode a global meeting place, and DVSwitch bridges let P25 reach the other digital modes. Like NXDN and M17, P25 remains a quiet, niche corner of the hobby — a busy night is a handful of voices.
P25 never had a large amateur following and never will. But its real fingerprint on ham radio isn't a user count — it's a waveform. The C4FM modulation that defines Yaesu System Fusion, the most widely used digital voice mode in amateur radio, was first engineered and proven in P25 Phase 1. Every Fusion QSO is, in a sense, riding technology that was built for first responders and quietly handed down to the bands.
P25 is the mode amateur radio didn't build, didn't design, and didn't pay to develop — it borrowed it from public safety, surplus radio by surplus radio. It brings superb audio and genuinely rugged hardware, along with the same closed-codec compromise as the rest of the digital family and a user base small enough that most nights are silent. Its lasting legacy on the bands is upstream of all that: the C4FM waveform it pioneered became the backbone of the most popular ham digital voice mode of all.