Nexus 1.0.1
Released 5 August 2026. The current release is 1.24.0. Download it · Every version
Added
-
Scroll any digit of the frequency dial to tune by that amount.
Hover the 100 Hz digit and one notch moves 100 Hz; hover the MHz digit and one notch moves a megahertz. It carries like a real VFO — 14.199 rolls up to 14.200 — and every digit is live, including MHz. The dial stops at the edge of the band rather than refusing to move, and it will not tune while the radio is transmitting. The tuning-step selector and the scope wheel are unchanged; this is in addition to them, not instead. On the main dial in Operate, Phone, CW, RTTY and SSTV.
-
4 m (70 MHz) is now a band you can pick in Settings ▸ Frequencies
, with FT8 at 70.154 MHz and the SSB/CW and AM/FM calling channels, alongside the MSK144, JT65 and WSPR channels already there. Frequencies are the IARU Region 1 plan. ⚠️ 4 m allocations vary a great deal by country — several are narrower than 70.0–70.5 MHz, and there is no US allocation at all, so a US station will see the band but cannot transmit on it. Check your own licence. No FT4 channel is offered because no 4 m FT4 frequency exists in any band plan we could find.
Fixed
-
The DXpedition board recommended 60 m for everything, on every card.
An operator with 60 m enabled saw all eleven WORK NOW cards read "Best shot: 60m" — the 12 m card, the 20 m card, the 160 m card, all of them. The windows differed per expedition, so the path modelling was running; only the band was stuck. Three separate faults, now all fixed. The model had no ceiling for 60 m. Every band carries a "how hard is real DX here even when the band is open" cap; 160/80/40 m had one and 60 m had been missed, so it alone was allowed a perfect score. That matters at night, when the model's other discriminators go quiet and the cap becomes the whole of the ranking — 60 m scored an identical 0.939 on paths from 2,214 km to 11,589 km and won them all. It now sits at 0.80, between 80 m and 40 m where it belongs, and a test refuses any future band added without a cap. The same fault was showing in the general band ladder in Connect, the globe and the map, where 60 m had been reading as the single best band for every operator on Earth at every hour of the day; 60 m no longer wins there either. What is honestly not fixed: that ladder still names one band far too often — 40 m now, where it used to be 60 m. The cause runs deeper than any one band and is described under Known below. 40 m at least earns DXCC credit and is a band expeditions actually run, so the recommendation is no longer actively misleading, but it is not yet carrying much information. A card now reports its own band. Each card covers one band, but the "best shot" line was ranking every band on the path — so a 20 m card would headline 60 m, a band that expedition may never have announced. Each card now shows the window for the band it is actually about, and says so plainly when that band has no opening. And nothing chasing DXCC leads with 60 m any more. 60 m earns no ARRL DXCC credit of any kind — not Mixed, not per-band, not the Challenge — and most DXpeditions do not run it: it is secondary, channelised, and a single-channel receiver rules out the split that runs a pileup. So the headline on a chase now goes to the best band that can actually earn you credit. 60 m is demoted, never hidden: it keeps its true score and its real window right beside the headline, and if it is genuinely the only thing open it still leads and says so. Where the question is not chasing — "which bands are open right now", in Connect and on the map — 60 m is left exactly where the physics puts it, because there it is a perfectly good answer.
-
Linux: the audio device list finally reads like your own machine, and shows all of it.
An operator with eight sound cards — an FTDX10 among them — was offered one card, named
hw:CARD=Device,DEV=0, while his system,aplay -l, KDE, pavucontrol and WSJT-X all called his radioUSB AUDIO CODEC. Nothing in the list matched anything he could see anywhere else, and his log filled with ALSA errors. Two causes, both in the audio library Nexus used to enumerate devices: it named each device by its raw ALSA PCM string and threw the human description away, and it tried to open every device to see whether it existed — so anything held by PipeWire, or by another application, silently disappeared. Nexus now reads ALSA's own device list directly. It opens nothing, so a card that is momentarily busy is still offered; it uses ALSA's own input/output flags instead of guessing; and it shows one entry per card, under the name the rest of your desktop uses. The error flood is gone because nothing is being probed. After updating, re-pick your audio devices once — the list now prefers the format-converting entry for each card, which is the forgiving one for a radio codec, and your saved entry may be the older strict one. Your existing choice keeps working and stays selected until you change it. -
Linux: Detect Rigs now fills in your radio's audio, which it never could before.
Automatic rig-audio pairing matches on names like "USB AUDIO CODEC" — and on Linux it was only ever shown
plughw:CARD=CODEC,DEV=0, so it matched nothing and left audio blank on every Linux station since the feature shipped. It now sees the real card name. Where a machine has more than one generic USB audio device, the more specific match wins rather than whichever came first in the list, so a radio codec is preferred over a plain USB sound adapter (this part applies on Windows and macOS too). -
A radio audio device that will not open now says so instead of quietly using the wrong one.
If you had explicitly chosen your rig's codec and it was unavailable — powered off, unplugged, or held by another program — Nexus fell back to the system default without a word. On a laptop that means receiving from the built-in microphone and sending transmit audio to the speakers, while PTT still keys the radio over CAT: a dead unmodulated carrier on the air that looks like everything is working. You now get a banner naming the device that failed. Nexus keeps running on the system default so the radio is not dead in the water, and keeps retrying your chosen device every couple of seconds — so a rig you switch on after Nexus, or a codec another program was holding for a moment, is picked up on its own within a few seconds and the banner clears itself. "System default" is still a real choice and still behaves exactly as before. If you have a device saved that is not one of the ones offered, the picker now says saved, not in the list rather than claiming it is absent — on Linux the list is pruned to one entry per sound card, so a device can be perfectly usable and still not be one we offer. Whether it actually opens is answered where it can be answered honestly: at open time, by name, in the banner.
-
FT8 decodes the station answering your CQ when you transmit off your receive frequency.
WSJT-X gives its deepest decoding two frequency windows, not one: around where you listen, and around where you transmit. The second exists because a caller normally answers on your transmit frequency, and with "Hold Tx Freq" on those are different places. Nexus only ever had the first, so the one signal in the slot you most want was the one decoded with the least help. Measured on a caller 900 Hz away from the receive marker: 12 of 12 recovered with the window, 0 of 12 without. Nothing changes for operators who transmit and receive on the same frequency — the two windows land on top of each other, which is what WSJT-X computes too.
-
The FT8 early decode no longer costs twice what it should.
WSJT-X runs the peek at 11.8 seconds deliberately cheap — a higher sync threshold and none of the a-priori passes — because the real decode re-reads the same audio at the end of the period anyway. Nexus was running the full-price decoder there. On the reference off-air recording that peek took 540 ms where it now takes 265 ms, and it returned the same thirteen signals either way. That matters beyond CPU: an early pass still running when the period ends makes the whole period's decode get dropped, which on a slower machine reads as a dead band. What the cheaper pass does give up is the a-priori recoveries — a station finishing a QSO with you, a few dB below the ordinary threshold, now arrives at the end of the period instead of three seconds early. The authoritative decode at the period boundary is unchanged, and it is the one the log and the roster are built from.
-
The FT waterfall.
It looked, in the operator's words, "so 8 bit" — and that turned out to be four separate causes stacked on one surface, all now fixed. The intensity axis was linear in amplitude against a reference that moved every 20 ms, which left the noise floor about 15 of 256 shades to work with and reserved the whole colourful middle of the palette for signals that are rarely there; it is now dB against a fixed reference, and an ordinary run of FT8 signals is spread across the palette instead of crowded into one dark blue. Five of every six spectra were being thrown away rather than averaged, so the floor boiled frame to frame; each drawn row is now the mean of every frame since the last one. Bins were point-sampled onto pixels, painting each one as a hard rectangle, and the accumulated history repainted that way every time you touched the palette picker or resized a pane — both now interpolate. The FT waterfall also keeps its own palette, defaulting to Turbo, so a choice made in another mode no longer follows it. What this does not do: it does not reach WSJT-X's smoothness. Ours integrates ~171 ms per row against their ~2.5 s, which is the price of a display that updates 12× faster.
-
The waterfall palette picker in Phone, CW, RTTY and SSTV said it applied to every mode.
It no longer reaches the FT waterfall, and now says so.
-
A capture underrun could blend one waterfall row across the gap, mixing audio from before the stall with audio from after it.
One row in roughly 450, but it is now a clean restart.
-
MSK144 decodes about 90× faster
— 2.55 s down to 0.028 s for a 15-second period. The frequency tolerance was being derived from the receive passband, which made it 1350 Hz where WSJT-X uses 50; MSK144 sits at one fixed centre, so that width never meant anything. The same mistake let the decoder accept a signal a kilohertz off frequency, which is not an MSK144 contact at all. This is the decode that could make the whole app appear to hang.
-
An FT1 over could run past the end of its own slot
and transmit into the period it should be receiving in. The transmit deadline had no slot bound, so any delay in keying the radio — a slow CAT link can spend a full second — pushed the end of the over out with it. WSJT-X clamps the window to the period and can never cross it; Nexus now does the same, for every mode. FT8 and FT4 were never affected in practice: they carry over two seconds of slack. FT1 carries none at all — its transmit buffer fills its whole 4-second period, so every FT1 over now ends exactly at the slot boundary. FT1's 4-second timing is unchanged. (An earlier draft of this entry said FT1 had 214 milliseconds of slack; that was measured from the tones rather than the buffer actually sent, and was wrong.)
-
TempoFast (FT1) starts its transmission 100 milliseconds earlier in the slot
, which leaves 164 milliseconds of quiet at the end of the over instead of 64. That trailing quiet is what protects the end of your signal: transmit stops at the slot boundary, and your sound card is always running a little behind, so if its buffer is longer than the quiet you have left, the last of your tones never reaches the air. Buffers of 20 to 100 milliseconds are completely ordinary, so the old margin was inside the range where that happens. What it costs: the lead-in is also TempoFast's tolerance for a station whose clock is FAST, and the two come out of the same fixed budget. A station more than 0.3 seconds ahead of UTC will now be missed where up to 0.4 seconds was tolerated before. Measured across signal levels, that tolerance is a cliff rather than a slope — inside it, decoding is unaffected — and the tolerance for a station running LATE improves by the same 100 milliseconds. If you work someone whose PC clock is badly fast, this is the trade you are on the wrong side of; the measurement that chose it is kept in the source so it can be revisited.
-
The QRZ link on the callsign card
only appeared once a callbook lookup had already succeeded, which withheld it in exactly the cases you want it — no QRZ subscription, no credentials, a lookup that failed, or a callsign QRZ has never heard of. Opening the page only ever needed the callsign. Separately, if the browser refused to open, nothing said so anywhere; every QRZ and DXpedition link now reports the failure instead of doing nothing.
-
SSTV gained the radio controls the other cockpits already had
— power, tune and stop from the header. Drive matters more here than anywhere else, because the picture rides in the audio and pushing past ALC visibly wrecks the image at the far end. The drop-image box no longer reserves 160 pixels before you have loaded anything.
Known
-
The band recommender still favours one band too strongly.
After the 60 m fix above it usually says 40 m. The reason is in the model rather than in any band's settings: at night the prediction for every band below about 9.6 MHz collapses onto a fixed per-band ceiling, so the 24-hour "best" figure the recommendation uses is a constant rather than something that varies with the path. Two paths 5,000 km apart can get the same answer. Fixing it properly means reworking how the night-time prediction is calculated — a bigger change with its own testing, deliberately not folded into this release. What this does and does not affect: the per-band cards, their colours and their sort order are computed from the CURRENT hour and are unaffected — they were always sound. It is the single "best shot" summary line that is worth less than it looks.
Every version
- 1.24.0 11 Nov 2026
- 1.23.0 5 Nov 2026
- 1.22.0 1 Nov 2026
- 1.21.0 26 Oct 2026
- 1.20.0 22 Oct 2026
- 1.19.0 16 Oct 2026
- 1.18.0 12 Oct 2026
- 1.17.0 7 Oct 2026
- 1.16.0 3 Oct 2026
- 1.15.0 26 Sept 2026
- 1.14.0 20 Sept 2026
- 1.13.0 16 Sept 2026
- 1.12.0 14 Sept 2026
- 1.11.1 8 Sept 2026
- 1.10.3 4 Sept 2026
- 1.10.2 3 Sept 2026
- 1.10.1 1 Sept 2026
- 1.10.0 31 Aug 2026
- 1.9.2 29 Aug 2026
- 1.9.1 26 Aug 2026
- 1.9.0 25 Aug 2026
- 1.8.1 23 Aug 2026
- 1.8.0 23 Aug 2026
- 1.7.6 21 Aug 2026
- 1.7.5 20 Aug 2026
- 1.7.0 18 Aug 2026
- 1.6.1 17 Aug 2026
- 1.6.0 16 Aug 2026
- 1.5.0 16 Aug 2026
- 1.4.0 15 Aug 2026
- 1.3.0 13 Aug 2026
- 1.2.0 10 Aug 2026
- 1.1.0 10 Aug 2026
- 1.0.5 8 Aug 2026
- 1.0.2 6 Aug 2026
- 1.0.1 5 Aug 2026
- 1.0.0 4 Aug 2026
- 0.28.2 3 Aug 2026
- 0.28.0 3 Aug 2026
- 0.27.0 2 Aug 2026
- 0.25.0 1 Aug 2026
- 0.24.0 31 Jul 2026
- 0.21.5 29 Jul 2026
- 0.21.0 29 Jul 2026
- 0.20.0 28 Jul 2026
- 0.19.7 27 Jul 2026
- 0.19.6 26 Jul 2026
- 0.19.4 26 Jul 2026
- 0.18.0 25 Jul 2026
- 0.17.22 25 Jul 2026
- 0.17.21 25 Jul 2026
- 0.17.20 25 Jul 2026
- 0.17.19 25 Jul 2026
- 0.17.18 25 Jul 2026
- 0.17.17 25 Jul 2026
- 0.17.16 25 Jul 2026
- 0.17.15 25 Jul 2026
- 0.17.14 25 Jul 2026
- 0.17.13 25 Jul 2026
- 0.17.12 25 Jul 2026
- 0.17.11 25 Jul 2026
- 0.17.6 25 Jul 2026
- 0.17.5 25 Jul 2026
- 0.17.4 25 Jul 2026
- 0.17.3 25 Jul 2026
- 0.17.2 24 Jul 2026
- 0.17.1 24 Jul 2026
- 0.16.4 24 Jul 2026
- 0.16.3 24 Jul 2026
- 0.16.2 24 Jul 2026
- 0.16.1 24 Jul 2026
- 0.16.0 24 Jul 2026
- 0.15.24 24 Jul 2026
- 0.15.23 24 Jul 2026
- 0.15.22 24 Jul 2026
- 0.15.21 24 Jul 2026
- 0.15.20 24 Jul 2026
- 0.15.18 24 Jul 2026
- 0.15.17 24 Jul 2026
- 0.15.16 23 Jul 2026
- 0.15.15 23 Jul 2026
- 0.15.14 23 Jul 2026
- 0.15.1 22 Jul 2026
- 0.15.0 21 Jul 2026
- 0.14.0 21 Jul 2026
- 0.13.0 19 Jul 2026
- 0.12.0 18 Jul 2026
- 0.11.1 18 Jul 2026
- 0.11.0 18 Jul 2026
- 0.10.0 17 Jul 2026
- 0.9.7 17 Jul 2026
- 0.9.6 16 Jul 2026
- 0.9.5 16 Jul 2026
- 0.9.4 16 Jul 2026
- 0.9.3 16 Jul 2026
- 0.9.2 15 Jul 2026
- 0.9.1 15 Jul 2026
- 0.9.0 15 Jul 2026
- 0.8.9 15 Jul 2026
- 0.8.8 14 Jul 2026
- 0.8.7 14 Jul 2026
- 0.8.6 14 Jul 2026
- 0.8.5 14 Jul 2026
- 0.8.4 13 Jul 2026
- 0.8.3 13 Jul 2026
- 0.8.2 13 Jul 2026
- 0.8.1 12 Jul 2026
- 0.8.0 12 Jul 2026
- 0.7.1 12 Jul 2026
- 0.7.0 12 Jul 2026
- 0.6.0 11 Jul 2026
- 0.5.2 11 Jul 2026
- 0.5.1 10 Jul 2026
- 0.5.0 10 Jul 2026
- 0.4.1
- 0.4.0
- 0.3.0
- 0.2.0 3 Jun 2026
- 0.1.0