DMR vs C4FM, D-STAR and NXDN: how the digital modes differ

Amateur digital voice on VHF/UHF isn't a single protocol but four competing standards that don't talk to each other directly: DMR (Digital Mobile Radio), Yaesu System Fusion (also known as C4FM), Icom D-STAR and NXDN. All four squeeze the voice with similar codecs, push a digital stream over the air — and yet you still can't hear someone on a different mode; the radio simply stays silent. For a newcomer, picking a first system out of this zoo is hard. Let's lay it out point by point: how the modes differ at the physical level, who is open and how much, what the hardware costs, what ecosystem stands behind each — and why DMR became the most widespread among hams.

The four modes in a nutshell

All four do the same thing: they digitize voice, compress it with a vocoder, add forward error correction (FEC) and modulate it onto the carrier with a narrowband 4-level scheme. The difference is in the details and, above all, in who invented them and for whom.

• DMR — an open ETSI standard born in commercial radio (professional radios, taxis, security). It reached the amateur world "from below," via cheap mass-market hardware. It uses TDMA with two time slots.
• Yaesu System Fusion (C4FM) — designed by Yaesu specifically for hams. Simple setup, pleasant audio on "native" hardware, gentle coexistence of analog and digital on one frequency (AMS) and WIRES-X rooms.
• Icom D-STAR — the pioneer of amateur digital (a JARL standard, early 2000s). The oldest of the four, with routing by callsign and reflectors. Its vocoder is the oldest too.
• NXDN — a joint development by Icom and Kenwood for commercial use (the IDAS and NEXEDGE brands). The most narrowband mode (there's a 6.25 kHz variant on par with D-STAR, while at 12.5 kHz it runs at 9.6 kbit/s), but it is the rarest among hams because affordable radios are scarce.

The vocoder: exactly how the voice is compressed

The heart of any digital mode is the vocoder (the voice codec), which turns speech into a stream of a few kilobits. In all four modes this is the AMBE family from DVSI, but the versions differ, and that matters.

• D-STAR uses the old AMBE (AMBE+, with roots in the AMBE-2020 chip, 1996–2002): a full payload of 3600 bit/s — of which 2400 bit/s is voice and 1200 bit/s is FEC. It sounds noticeably more "digital" and thinner than its neighbors.
• DMR, System Fusion and NXDN use the newer AMBE+2 (roughly 2005). In NXDN, for example, of those same 3600 bit/s, 2450 bit/s goes to voice (AMBE+2) and 1150 to FEC. The audio is subjectively cleaner and more intelligible.

TDMA versus FDMA: the key technical difference

The most noticeable difference in practice is the channel access method. Here DMR stands apart.

• DMR — TDMA (Time Division Multiple Access). One frequency is split in time into two independent time slots (Slot 1 and Slot 2). On a single repeater frequency two different conversations run at the same time. The stream rate is 4.8 kbit/s × 2.
• D-STAR, NXDN and Fusion — FDMA (Frequency Division Multiple Access). One conversation = one frequency; you don't get a second slot "for free."

TDMA brings DMR some nice bonuses: twice as efficient use of spectrum (and the bands keep getting more crowded), and noticeably longer battery life — a handheld's transmitter physically works only during "its" half-cycle, that is, about half the time.

Standard openness and hardware price

This is where DMR really wins. It is described in the open standard ETSI TS 102 361 (parts 1–4). That means radios are made not by one vendor but by dozens: Hytera, Motorola, AnyTone, TYT, Retevis, Baofeng, Radioddity and others. Competition has driven prices down to almost absurd levels.

• DMR — usually the cheapest of all. A simple Baofeng/TYT handheld costs a few dollars worth of pocket change; a top dual-band AnyTone D878 with a callsign database and Bluetooth runs around 130–250 dollars at global prices.
• System Fusion (C4FM) — almost all hardware from a single brand (Yaesu). Mid-range price, fewer model choices, but everything works "out of the box" and is friendly to newcomers.
• D-STAR — as a rule, the most expensive: Icom handhelds and base stations (ID-52, IC-9700) cost a fair bit, and there are almost no alternative vendors.
• NXDN — the standard is open, but amateur radios are extremely rare; mostly it's commercial Icom/Kenwood gear that enthusiasts drag into our bands.

The ecosystem: networks, hotspots, firmware

A mode isn't just a radio, but also where it connects. And here DMR has the richest infrastructure.

• BrandMeister — the world's largest amateur DMR network, with thousands of talkgroups, the one almost every amateur hotspot points to by default. Alongside it live DMR+, TGIF and other networks.
• MMDVM (Multi-Mode Digital Voice Modem) — an open modem platform on which cheap personal hotspots are built. The firmware is open, the board costs next to nothing, and you can assemble a node in an evening from a Raspberry Pi and an MMDVM shield.
• System Fusion relies on WIRES-X (the Yaesu network) and YSF reflectors; D-STAR on reflectors and callsign routing; NXDN on the few NXDN reflectors there are.

The beauty of MMDVM is that one and the same hotspot can handle several modes at once (DMR, YSF, D-STAR, NXDN, P25) — in software. So physically a node can receive any of them; the only question is whether you have a radio for that mode and where the bridge leads next.

Why DMR won with hams

If you boil it down to a few points, DMR's popularity rests on four pillars:

1. Price. An open standard = dozens of manufacturers = the cheapest radios on the digital market.
2. Spectrum and battery. Two-slot TDMA saves frequencies and extends a handheld's runtime.
3. The network. The giant BrandMeister and a wealth of talkgroups — there's someone to talk to from day one.
4. Hotspots. A cheap MMDVM with open firmware gets you onto the global network even where there isn't a single repeater nearby.

To be fair: Fusion is often praised for the most "human" audio and its simplicity, and D-STAR for its elegant callsign routing. But on the combined score of "cheap + network + hardware + slots," DMR has pulled away from everyone. Newcomers are usually advised to start with DMR or Fusion, and to take up D-STAR — let alone NXDN — deliberately, for a specific local infrastructure.

Incompatibility and the role of bridges

A key and often non-obvious fact: the modes are mutually incompatible over the air. A DMR radio won't hear D-STAR, Fusion won't understand NXDN — even if both sit on the same frequency. The reasons stack up: different access methods (TDMA versus FDMA), different modulations (4FSK, C4FM, GMSK), different vocoder versions and different packet formats. There's physically nothing to "hear" a neighbor on another mode with.

The solution is bridges and transcoding in the network, not on the air:

• A bridge connects to two networks of different modes and transcodes the voice: it unpacks one vocoder and repacks it into another. This is exactly what hardware AMBE vocoders (or their emulation) are needed for — because of the different AMBE+2/AMBE versions you can't "just move the bytes across."
• An MMDVM hotspot on receive works in one chosen mode, but supports several in software — and through software like MMDVMHost/Pi-Star it can link them to different networks.
• The commercial world has its own incompatibility: NXDN/IDAS/NEXEDGE radios talk to each other, but not to DMR, D-STAR, P25 or TETRA.