February 2026 may go down as the month the ham radio digital world got its fastest-ever weak-signal mode — and immediately fractured over who owns it. Welcome to the story of FT2.
The Birth of a New Mode: Made in Italy, Heard Worldwide
On February 16, 2026 at 22:47 UTC, a small group of Italian radio amateurs quietly made history. Gathered across sites in Campania, the island of Capri, Sardinia, and Turin, stations IZ8VYF, IZ8XXE, and IC8TEM completed the very first verified FT2 QSOs in history — dozens of contacts on the 40m and 80m bands, each completed in a jaw-dropping 7 to 11 seconds.
The man behind this achievement is Martino Merola, IU8LMC, a member of the ARI Caserta amateur radio club in southern Italy. In what he himself describes as a “world first,” Merola developed FT2 as a dramatically time-compressed evolution of the popular FT8 and FT4 digital protocols. Strikingly, Merola has openly acknowledged that Anthropic’s Claude AI was used as a development tool to help modify the WSJT-X source code — a candid and increasingly common disclosure in modern amateur radio software development.
In barely two weeks, FT2 has gone from a tightly controlled private beta shared via a WhatsApp group to a globally discussed, hotly debated phenomenon appearing on PSKReporter maps, YouTube channels, and ham radio forums from Europe to India.
What Exactly Is FT2?
FT2 is an ultra-fast, weak-signal digital mode designed for HF amateur radio operation. To understand why it matters, consider where it fits in the FT-family timeline:
The headline numbers are remarkable: FT2 is four times faster than FT8 and twice as fast as FT4, capable of a theoretical throughput of 240 contacts per hour under ideal conditions. The fastest verified QSO on air so far? Seven seconds on the 80m band.
The Technical Architecture
At its core, FT2 uses the same proven codec as FT8 and FT4 — a 77-bit payload, LDPC (174,91) forward error correction, and 8-GFSK modulation. The critical and singular difference is aggressive time compression: shorter symbols, a cycle reduced to just 3.8 seconds, and a bandwidth footprint of approximately 150 Hz.
The trade-off is real and explicitly documented. FT2’s sensitivity floor sits around –12 to –14 dB SNR, compared to FT8’s extraordinary –20 dB. This means FT2 is not a replacement for FT8 in marginal propagation conditions. It is a specialized tool for scenarios where signals are strong and speed is the limiting factor — contesting, DXpeditions, pile-ups, and brief VHF/UHF propagation openings.
One critically important operational requirement: PC clock synchronization to ±50 ms accuracy. FT8 tolerates a generous ±200 ms; FT2’s compressed cycles demand four times tighter timing. Tools like Meinberg NTP are strongly recommended.
The Software: Decodium 3 and the Open-Source Fork
Decodium 3 by IU8LMC
The original FT2 implementation runs on Decodium 3, a fork of the WSJT-X v3.0.0-rc1 codebase. Installation is relatively painless for anyone already running FT8 — you install a standard WSJT-X base, then overwrite the core executable files with Decodium’s modified binaries. The software preserves the familiar WSJT-X waterfall interface while adding FT2 mode, integrated Auto CQ, and an Auto Log function. Downloads are available at hampass.com/ft2.
Early releases were Windows 64-bit only, but by February 24 a new beta installer supporting Windows 32-bit, Linux, and macOS was released — a significant moment, since much of the amateur radio community runs older dedicated “shack computers” with 32-bit architecture. No ham should be left behind due to hardware age.
WSJT-X Improved 3.1.0 by DG2YCB
Ten days after Decodium’s first on-air tests, developer DG2YCB released WSJT-X Improved v3.1.0, a separate open-source implementation of FT2 built from scratch without using any of IU8LMC’s code. This version is fully open-source and available on SourceForge, with packages for Windows (32 and 64-bit), Linux, and Raspberry Pi.
WSJT-X Improved 3.1.0 describes FT2 as “an ultra-fast 77-bit mode with TR periods of 3.75 seconds (= 1/2 of FT4)” and traces the conceptual lineage back to early experiments by the original WSJT development team — Joe Taylor K1JT, Steve Franke K9AN, and the late Bill Somerville G4WJS.
The FT2 Schism: When Two Modes Share One Name
Here is where the story takes a dramatic turn, and where the ham radio community finds itself in the middle of what commentators are already calling the “FT2 War“ or the “FT2 Schism.”
The two FT2 implementations — Decodium’s and WSJT-X Improved’s — are technically incompatible. Despite occupying the same conceptual space and even carrying the same name, they speak entirely different digital languages. A direct comparison reveals the incompatibility:
The result: Station A (WSJT-X Improved) cannot decode Station B (Decodium), and vice versa. Both camps acknowledge this. A station receiving only garbage decodes from the other software is the current reality.
The split escalated quickly because three sensitive issues collided simultaneously: claims of development priority (Decodium was first on air), questions about GPL license compliance (WSJT-X is GPL3, which requires derivative works to publish source code), and the deeply held collaborative norms of ham radio digital culture, where community trust runs deep.
IU8LMC’s supporters point to working on-air QSOs as proof of concept and priority. DG2YCB’s supporters emphasize open-source transparency and the established WSJT-X development philosophy. Neither camp shows signs of immediate convergence.
Adoption: A Global Groundswell in Just Two Weeks
Despite the schism, the broader community response to FT2 has been striking. Within days of the first on-air tests:
- PSKReporter began listing “Decodium 3 FT2 v3.0.0-rc1 (mod by WM8Q)” among its reporting software, with 389 FT2 spots logged in just two hours as of February 19, 2026 — and rising.
- MSHV, another popular weak-signal digital mode application developed by LZ2HV, was being tested for FT2 compatibility by its developer Hristo, with early reports of successful decoding.
- JTDX beta builds were also noted to show some compatibility with the WSJT-X Improved FT2 message formats.
- Early adopters from France (F4JKH, F6KKR), Austria, Czech Republic, Slovakia, India, and many other countries reported successful QSOs within the first week.
French operator F4JKH, testing both from his home station and the F6KKR club station, reported: “As FT8 and FT4 are often overpopulated, especially during contests, FT2 may take away some of that traffic.” This observation touches on a genuine and pressing problem — the 14.074 MHz FT8 window is often a wall of overlapping signals, and a faster mode that can spread activity more efficiently across the band has real practical appeal.
QSOs have been confirmed uploading to both LOTW (Logbook of The World) and QRZ.com, though the latter currently categorises FT2 contacts under the “MFSK” mode label — a stopgap until formal mode recognition is granted.
Where FT2 Shines: The Use Cases
FT2 is explicitly not designed to compete with FT8 for weak-signal DX, EME, or grey-line propagation work. FT8 will remain the undisputed king for those scenarios. FT2’s sweet spot is a distinct set of operating conditions:
High-Rate Digital Contesting: With a theoretical throughput of 240 QSOs per hour, FT2 is faster than RTTY while occupying far less bandwidth and using robust error correction. Contest operators looking to maximize contact rates under good conditions have an immediately compelling argument for FT2.
DXpeditions and Pile-Up Management: A 7-second QSO versus a 60-second FT8 QSO changes the mathematics of a DXpedition pile-up dramatically. More stations can be worked per hour, reducing the frustration that characterises the often chaotic FT8 DXpedition experience.
VHF/UHF Sporadic-E and Tropo: Propagation openings on 6m, 2m, and higher bands are often brief and unpredictable. The ability to complete a QSO in 11 seconds rather than 60 means more stations can be worked in a single opening. Some experimenters have also proposed potential utility for aircraft scatter contacts on 1.3 GHz and above, where signal windows are measured in seconds.
Special Event Stations: High-volume special event operations, where the goal is to give out as many contacts as possible in a limited time, stand to benefit substantially.
Controversy and Criticism
Not everyone is enthusiastic, and the criticisms deserve fair hearing.
The automation concern is real. Decodium’s integrated Auto CQ feature, which allows fully automated operation without any human interaction, represents a long-standing tension in the FT8 community now carried over to FT2. Critics argue that “zombie mode” — where a station can rack up hundreds of contacts without any operator present — fundamentally changes what amateur radio means. Supporters counter that this capability exists in other WSJT-X forks and is ultimately a matter of operating ethics.
The DXpedition claim has been challenged. One experienced digital operator noted pointedly that existing FT8 Superfox and Fox-and-Hound modes are specifically designed for DXpedition pile-up management and outperform FT2 for that application because they handle multiple simultaneous contacts. The claim that FT2 improves on FT8 for DXpeditions, they argue, reflects unfamiliarity with the existing toolkit rather than genuine technical advantage.
The SNR trade-offis significant and non-trivial. At –12 dB sensitivity versus FT8’s –20 dB, FT2 requires roughly eight times more signal power for equivalent reliability — a meaningful barrier for QRP operators and marginal propagation paths.
The fragmentation riskis the most serious systemic concern. Two incompatible implementations with the same name create confusion, waste adoption energy, and risk leaving the mode stranded with insufficient critical mass on either side to sustain long-term activity. The history of amateur radio digital modes is littered with technically promising protocols that failed to achieve the network effect they needed.
What Comes Next?
Despite the turbulence, several developments in the coming weeks and months will be decisive:
Protocol convergence or coexistence: Will IU8LMC and DG2YCB find a path to interoperability? The WSJT-X Improved developer has been direct — the two protocols are not just incompatible, they are fundamentally different. Convergence would require one or both sides to change their encoding, which neither has indicated willingness to do. Coexistence under different names is one possible outcome.
ADIF Certification — Already Confirmed: In a major milestone announced on February 25, 2026, the ADIF Development Group officially accepted FT2 for inclusion in the ADIF Specification. Within weeks, FT2 will be formally certified as a recognized digital mode — paving the way for full integration into ARRL Logbook of The World (LoTW). This is the fastest any new digital mode has achieved ADIF recognition, and it removes one of the biggest barriers to mainstream adoption. Award hunters and DX chasers can now plan FT2 operations with confidence that contacts will count.
Official WSJT-X inclusion: The mainstream WSJT-X project, maintained by the original developers at Princeton, has not adopted FT2. Inclusion in the official release would be the single most powerful driver of mass adoption — as it was for FT8 and FT4. Whether that happens depends on the protocol question being resolved and on the WSJT-X team’s own assessment of FT2’s value.
Contest sponsor recognition: For FT2 to fulfil its contesting potential, major contest sponsors — ARRL, CQ Magazine, RSGB, and others — would need to formally recognise it as a valid mode category. That step typically requires a stable, well-documented protocol and meaningful community adoption. Neither condition is fully met yet.
Logbook recognition: Formal mode entries in ADIF (Amateur Data Interchange Format) and recognition by LoTW and eQSL would legitimise FT2 contacts in award programmes, a significant incentive for many operators.
Getting Started with FT2 Today
If you already run FT8 from your shack, getting on FT2 requires minimal additional equipment — just a software change and a clock sync check.
- Download the latest Decodium 3.0 “Raptor” from ft2.it (Windows 32/64-bit, Linux, or macOS builds available). Alternatively, use WSJT-X Improved 3.1.0 from SourceForge if you prefer the fully open-source path.
- Ensure your PC clock is synchronised to ±50 ms — install Meinberg NTP if needed.
- Use a separate WSJT-X profile (via the
--config FT2command-line flag) to protect your existing FT8 settings. - Set your transceiver to one of the provisional FT2 frequencies in USB mode.
- Tune your audio levels and dive in.
The learning curve is shallow if you know FT8. The experience, by all early accounts, is genuinely different — contacts that feel almost conversational in their speed after years of staring at 15-second countdown timers.
Final Thoughts
FT2 is, in every meaningful sense, a genuine technical achievement. The idea of compressing an FT-family QSO to under 4 seconds per cycle, maintaining structured message integrity, and achieving verified on-air decodes down to –12 dB SNR is not trivial engineering. That it was accomplished by a regional amateur radio club in southern Italy, with AI assistance, in a matter of weeks, speaks to both the creativity of the amateur radio community and the democratisation of software development tools.
Whether FT2 becomes the next FT8 — ubiquitous, dominant, transformative — or fades into a footnote as an interesting experiment, depends on questions that go beyond its technical merits: community trust, open-source governance, and the unglamorous work of standardisation.
What is already certain is that February 2026 added a new chapter to the long story of amateur radio’s relentless drive to communicate further, faster, and more efficiently. The bands are alive with something new.
Related Posts
- FT2 – A New Digital Mode Hams Are Starting To Spot (February 18, 2026)
- FT2: The Fastest Digital Mode in Amateur Radio (February 19, 2026)
- FT2 Digital Mode: Speed King With Serious Growing Pains (February 20, 2026)
- Low-Loss Feedline on a Budget: Building 450 to 600 Ohm DIY Ladder Line from Common Materials (March 3, 2026)
- Sporadic E Propagation: The Science Behind Long-Distance FM Radio Reception (February 26, 2026)
Thanks for quoting myself in the article. Did not ask for it, did not get asked for being quoted – but it is OK, my words have not been too twisted around that I would feel misinterpreted or angry. It’s just fine.
FT2 means for me actually three things – in spite of having the two different software versions for the modulation scheme:
(1) Finally we have a third official frequency for narrowband digital transmissions in the different HF bands, this means relief that you can feel in FT8 and FT4 frequency usage. I have been so much looking forward to that.
(2) The usage of FT4 and FT2 has been defined for contesting. If you have a close look what is happening today, then you realize that they are used mainly for normal communication as it is happening in FT8. FT4 and FT2 should not have the stigma of being there only for contesting. The use of these modulation schemes is way too important for the Digital Narrowband transmission environment.
(3) Fast transmissions means also fast reception and ultimate time synchronisation needs. At the end, you will find out in large numbers of FT2 QSOs that this mode is not a lot faster for medium SNR QSOs in FT4 due to unsuccessful demodulation and subsequent retransmissions.
Conclusion:
I am so happy that FT2 exists. No matter what format of FT2. It works, many people use it around the clock and complete / confirm QSOs. New frequencies have been defined for Narrowband Digital inside the framework of band use without stealing from CW or SSB voice assignments. The success of Narrowband Digital is richer due to FT2. However, HAM radio ethics and respect must be adhered to, your license code everybody of use adhered to must be obeyed. This is not a free lunch, playing with others for the own success is against HAM spirit.