hamradiotools

Nexus 1.23.0

Released 5 November 2026. The current release is 1.24.0. Download it · Every version

Added

  • RTTY at 50 baud, and your RTTY tone pair saved (#459).

    Settings ▸ Digital ▸ RTTY ▸ Baud rate now offers 50 baud beside 45.45 and 75; it sets the transmit bit clock (AFSK and true FSK) and the decoder alike. The new Low tone (Hz) box beside it saves where RTTY sits: the decoder and the AFSK transmitter start on this tone and this plus the shift, so 1275 at 170 shift is 1275/1445 every time you open RTTY, with no waterfall click. It is the low tone rather than "mark" because which of the two is mark depends on the sideband and Reverse. A waterfall click still nets for the session; changing the box brings both back onto the saved pair. The licence check judges the pair you actually transmit on. 2125/2295 stays the default, so nothing changes until you set it.

  • PSK63 on the PSK screen.

    The sub-mode selector has PSK63 beside PSK31 and QPSK31: PSK31 at twice the speed (62.5 baud, about 125 Hz wide), to receive and to transmit, through the same TX switch, licence check, Stop and watchdog as PSK31. Contacts made on it log as PSK63, and a PSK63 station on the Needed board opens the PSK screen on PSK63.

  • The FT drive is remembered per band and mode.

    On FT8, FT4, FST4 and the other FT modes, the drive slider now remembers where you set it on each band, for each mode, on each radio. Change band, mode or radio and the drive you last used there comes back by itself, so an amplifier that needs less drive on 6 m than on 20 m no longer has to be trimmed by hand at every band change. A mode you have never used on a band takes the drive you last used on that band in any FT mode; a band you have never driven leaves the drive where it is. Only your own moves of the slider are remembered. The drive never changes while you are transmitting: a band change during an over brings the new band's drive back as the radio unkeys, before the next over. There is nothing to turn on, and Phone, CW, RTTY, PSK and JS8 are not affected. The memory is kept with each radio's settings and goes with a settings backup.

  • A dark map, whatever the theme (#444).

    The Flat map, the 2-D Globe and the Beam map have a "Map style" choice at the top of the Layers panel: Atlas (the light map they have had since 1.17, still the default) or Dark (dark land and sea, with the grid, the greyline, the TX/RX lines and the stations still easy to read). It is independent of the UI theme, so the Light theme can have a dark map too; on a dark map the stations and labels are drawn in the dark theme's colours, so they stay readable. A dark built-in theme (Midnight and the others) keeps its own map until you pick one, and then offers "Theme's own" to go back to it. The choice is kept per window and shared by the Connect map and the APRS map. The 3-D globe keeps its photo of the Earth.

  • Needle TX meters with peak hold (#343).

    Settings ▸ Appearance ▸ Workspace ▸ TX meters picks Bars (as before, and still the default) or Needle. Needle draws power, ALC and SWR as analog meter faces wherever Nexus shows the transmit meters: the FT8 and other FT screens and their larger meters window, Phone, CW, SSTV, JS8, Tempo and Settings ▸ Radio ▸ Audio, from the same readings as the bars: power scaled to your radio's rated output, the ALC face's amber and red arcs where your radio's own ALC zone puts them, and the SWR face's at 1.5:1 and 2:1 (none where Nexus cannot vouch for the radio's SWR scale, which the face marks SWR?). A bright peak mark jumps to the highest reading at once, holds it for a second, then falls back slowly. Between overs the needle rests and the peak mark and the number show the last over's peak, until you transmit again. (Bars, between overs, show the last reading instead.) Compression (with Phone's processor controls) and the temperatures stay bars beside the needles; the FT strip's small cell has room for the three faces only, and its compression reading is in the larger window. The meters are readouts only and the faces stay the same size whether you are transmitting or not. The choice travels with Settings ▸ App ▸ Backup & reset, with the rest of the look.

  • ALC beside the drive in SSTV, JS8, Tempo and Settings.

    The SSTV, JS8 and Tempo screens' transmit strips now show the transmit meters (SWR, ALC, PO, COMP) beside Tune and Stop TX, as Phone and CW do, on any window but the smallest, so you can set the Drive (or SSTV's RF power) against the radio's ALC; SSTV carries the picture in the audio, so watching ALC there keeps an overdriven picture off the air. Settings ▸ Radio ▸ Audio shows the same meters right under the Drive slider, so you can key Tune and set the drive without leaving Settings. They read while the radio transmits and hold the last readings after; they are shown for the radio in use, not while you edit another one. The meters need a radio that reports them over CAT.

  • The satellite transponder you work, remembered — and a ★ favorite (#458).

    On a satellite you work in one mode (the ISS FM repeater, say), you no longer have to open "show all", scroll and pick it again every pass. The transponder chooser remembers the one you picked last on each bird, marks it "last used" and keeps it on screen even when it is past the first rows, and "▶ Work this pass" picks it. Each workable card also has a ☆ you can star: a favorite wins over the last pick. Starring or remembering never tunes the radio by itself; only a pick does, as before. "Work this pass" still leaves alone a transponder already being worked on the pass and your "None — leave the dial to me". If SatNOGS no longer lists your saved transponder as working, the usual pick is used and a quiet line under the chooser says so; beacons can never be chosen. The choices are kept with your station data, so a settings backup and a computer driving your station from afar carry them.

  • WWFF beside POTA and SOTA (#363).

    The POTA/SOTA board has a WWFF tab that lists the World Wide Flora & Fauna activators on the air now, from WWFF's own Spotline service (spots.wwff.co, which publishes its spot list for programs like Nexus; Nexus asks it at most once every 30 seconds, as Spotline asks). It lists activators spotted within the last hour. The tab that showed both programmes is now All and shows POTA, SOTA and WWFF together. HUNT on a WWFF spot works as it does for a park: the radio moves to the spot, the log strip fills in WWFF and the reference (KFF-1234, DLFF-0001), and the contact is logged with it. The log strip's program list has WWFF too, for a reference you type yourself. A WWFF contact is exported with SIG=WWFF, SIG_INFO and the dedicated WWFF_REF field, imported from either, and counted on the Awards page's WWFF line. Live WWFF activators get a WWFF chip on the Needed board, with a WWFF filter chip beside POTA and SOTA. A remote station's board still shows POTA and SOTA only.

  • Other programs can log contacts into Nexus.

    Keep Nexus as your log while another program runs the radio: the FreeDV contact port (Settings ▸ Radio ▸ Transmit limits & sharing ▸ Accept contacts other programs log, 127.0.0.1:2239, on by default) now takes the contacts WSJT-X, JTDX, MSHV, JTAlert and GridTracker send (their QSO Logged and Logged ADIF messages), N1MM's contact broadcast, and plain ADIF over UDP as Ham Radio Deluxe and Log4OM take it. Turn on "Accept from devices on my network" for a device on your home network such as the S.A.T. satellite tracker (on the S.A.T., set QSO LOG TYPE to Log4OM or HRD, QSO LOG IP to this computer, port 2239); only private network addresses are taken, and you can name the one address allowed (it narrows who may send but is not a password: anything on your network that can send to the port can log contacts, so turn it on only on a network you trust). Each contact is logged only under your own station call, is not written twice when you also logged it in Nexus within five minutes, takes your running POTA or SOTA activation when it names no park of its own, says "Logged K1ABC from WSJT-X", and its contact card says which program it came from (a device that sends no name of its own, such as the S.A.T., is "a device on your network"; its address is only in the connection log, and which program a contact came from stays in your log and goes to no logging service). A satellite contact from WSJT-X keeps its satellite name: Nexus waits a moment after WSJT-X's first message for the fuller second one, which names the bird. A date that does not exist is refused. It goes to QRZ, Club Log, eQSL, HRDLog, Cloudlog/Wavelog, WRL and N3FJP like your own contacts (held while contest mode is on), and is not sent back out to WSJT-X UDP, HRD, N1MM or DXKeeper. A port Nexus itself sends to is refused, with the reason. N1MM's edits and deletes are not read yet, and the S.A.T.'s DXKeeper and ACLog targets (TCP) are not taken: set it to Log4OM or HRD. The switch, the port, the network switch and the allowed address are set at the station only, never from a computer driving it. Receive only: nothing sent to the port can key or tune the radio. With two radios in two windows, the first window takes the contacts; the second no longer reports another program holding the port, and its connection log says the other Nexus window takes them (#431). Something else on the port is still reported as before.

  • LP on the Rotor box, the long way round (#338).

    Beside a cockpit, the Rotor box's line for the call in your log entry now has LP beside Point, as the rotator strip in the cockpit header already does: Point turns the antenna to the station the short way, LP to the reciprocal heading. Two buttons, never a switch, so the path you pressed is the path you get, and the message after the turn says which it was. Neither appears when there is no call or Nexus cannot place the station. On the desktop only, as before.

  • Follow the Moon with your rotator (EME).

    With your grid square set, the Rotor box shows where the Moon is from your station (azimuth and elevation, worked out for your own spot on the Earth's surface, within a hundredth of a degree of NASA JPL's figures) and a Follow the Moon button. It turns the rotator to the Moon and keeps it there, moving it each time the Moon passes the rotator's deadband; the elevation follows too on a rotator that has an elevation axis (an azimuth-only rotator is turned in azimuth only, and the box says so). While the Moon is below the horizon the antenna stays where it is and the box says so, and a rotator that stops answering is asked again and followed again when it answers. Stop following, ■ STOP, turning the rotator yourself, pointing it at a station (Point, LP, → CALL or the Needed board's Go) or starting a satellite pass ends it: one tracker at a time, so starting it also ends a running satellite pass. Beside a cockpit whose station's grid Nexus knows, the box also gives the next EME window with that station, when the Moon is above the horizon at both ends. It moves the rotator only: no transmitting, no tuning, no Doppler. See Rotor Box.

  • Replay the last six hours of the sun and a flare.

    The flare row of the map's and the 3-D globe's Layers panel has a "⟲ replay 6 h" button. It opens a replay bar at the top of the chip stack, starting at the highest X-ray reading of the last six hours: slide it, and the sun, the moon, the night side and the flare's absorption and band rings move together to that time. It is drawn only from the GOES X-ray readings your station already holds, so the flare it shows is always marked "estimated from GOES" (NOAA's own absorption map is not kept for past times), and a time with no reading draws no flare and says so. The bar always says REPLAY, the UTC time and how long ago, with a dashed blue edge, and that stations, spots and every other layer are still live; the flare chip for now steps aside until you press "back to now". Nothing about a replay is saved.

  • Space-weather chips on a narrow map: two, and "+N more".

    On a big-event day the map's and the 3-D globe's space-weather chips (the flare or replay bar, the proton chip, the aurora chip and NOAA's storm watch) could cover much of a narrow map, as at 1024×768. When the map is too narrow for the Layers panel and the band conditions side by side, it now shows the first two chips and one "+N more" chip for the rest; click it, or Tab to it and press Enter, to see those chips whole in a list, and Escape closes it. The replay bar is always one of the two. A wider map shows every chip, as before.

  • Where on the sun: NOAA's active regions.

    Unfold the flare chip and a small sun disc shows NOAA's numbered active regions (solar north up, east on the left, as NOAA's images show it), each listed with its NOAA location and whether it is near the east or west edge; hover the sun on the map for the same list. They come from NOAA's daily region list, drawn where NOAA observed them that day and labelled with that day. A list more than three days old, dated ahead of the station's clock, or not held at all draws no region and says why. Nexus asks NOAA for the list at most once an hour, only while a map shows the sun or the flare layer.

  • The polar caps and the aurora oval, with their edges.

    During a solar proton event the polar caps on the map and the globes now show where they end. While NOAA's own D-RAP absorption map is current, the caps are NOAA's, with NOAA's real proton cutoff, and the chip says "NOAA D-RAP" with its time; otherwise they are Nexus's estimate from the GOES proton reading (NOAA's D-RAP method), with the cap's edge drawn as a dashed violet line labelled "cap edge (approx.)", because that edge is Nexus's approximation, and only when a measured Kp places it (without one the chip says the edge is unknown). The proton chip gives NOAA's S level, the strongest absorption at 30 MHz, the proton reading's time (and its age once it is more than 15 minutes old), and how long it should last by NOAA's rule of thumb from the proton level ("Lasts ~24 h"), which it leaves out when the reading is old. Only a proton event NOAA would call one is drawn: protons at S1 or above, from a reading under 30 minutes old that carries its time (and not dated more than two minutes ahead of the station's clock). Below S1, or with a reading 30 minutes old or more or without a time, no caps, no edge and no chip are drawn, NOAA's absorption map is not read as protons, and the layer says which (no proton event, how old the reading is, that it is dated ahead of the station's clock, that it has no time, or that the station does not say). A proton event stops being drawn 30 minutes after its reading even if the station stops answering in between. During a proton event Nexus also asks NOAA for its absorption map, at most once a minute. The aurora oval now draws its equatorward edge: NOAA OVATION's own line where its chance of aurora reaches 10 %. When OVATION is out of date (its forecast time more than 10 minutes gone), carries no forecast time (so it cannot be shown to be current), or is missing, no oval and no edge are drawn at all, and the Layers panel and the aurora chip say which, with the time the old forecast was for when there is one; nothing is drawn in its place. The Kp and NOAA's Kp outlook stay on screen as before. The aurora layer turns itself on while Kp is 5 or more (NOAA's measured or estimated figure, whichever is higher) or a NOAA G watch is in force, and off again after; the Layers panel and a chip say it is on for the storm, and if you untick it during the storm it stays off until the storm is over, on every map in that window (the Flat map, the Globe and the 3D globe, on any screen, and after you leave and come back). Your own choice for the layer is not changed.

  • What a flare or a proton event does to your path, right now.

    When NOAA's X-ray satellite shows an M- or X-class flare, or its proton sensor shows a proton event (S1 or above), the selected station's path is marked where it is hit: the map and both globes draw the stretch of each route the flare weakens (orange) or takes out (red) on the band shown, and a proton event's polar-cap absorption (magenta), over the short and the long path alike. The Path card's sentence adds how much, for example "~9 dB of flare absorption on 20m (estimated from NOAA's D-RAP grid 12:47Z), recovery ~13:20Z (NOAA's own estimate)", and Path over the day hatches the current hour of each band that is hit, without changing the modelled colours. With no station selected, the Band outlook and the band conditions on the map start with a "now" line: "Flare X1.3 (R3): 60m and below out on the dayside until ~13:20Z", an S2 proton event over the polar caps, a NOAA geomagnetic watch with the time NOAA issued it and when it starts. Every figure says where it came from and when: NOAA's own D-RAP map while it is current, otherwise an estimate from the GOES X-ray reading, and NOAA's recovery time only when NOAA gives one (otherwise NOAA's rule of thumb, named as such). During a proton event (S1 or above, from a reading under 30 minutes old), the absorption NOAA's map adds over the polar caps for protons is never counted as the flare's: there the flare figure uses the GOES estimate and says so, and the proton part is counted once, as the proton event's, when a Kp reading places the edge of the polar cap (without one it is not counted at all, and the path's details say so). The figures are estimates laid over the propagation model, which does not include flares; the model's own numbers, and the path grades of your logged contacts, are never changed by them. Nothing is said when the readings are missing or old.

  • A big Operate/Standby button in the Amplifier box.

    The Amplifier box (on Conditions, in the dashboard rail and in a cockpit's boxes) now has a large button that shows and switches your amplifier between Operate and Standby. It is red with "OPERATE · Active" and a lamp symbol while the amplifier is in Operate, and green with "STANDBY · Not active" and a pause symbol while it is in Standby, so the state reads from the words and the shape as well as the colour. The button shows what the amplifier itself last reported: clicking it sends one Operate/Standby press, and the button changes when the amplifier reports the new state, usually within a second or two. From the click until then the button is held and says "Sent — waiting for the amplifier", so a second click cannot switch it straight back; if the amplifier reports no change within five seconds the button frees up again, still showing what the amplifier reported. It is locked, with a dashed edge, while you are transmitting (whichever of the radio or the amplifier says so) and while the amplifier is not answering, when it shows "—" and the box says why. If Nexus could not send the press at all, "Not sent" appears under the button. Nexus drops a press without a message, and the button stays as it was, whenever it cannot be sure the radio is not transmitting. With an Elecraft KPA500 or KPA1500, which does not report transmitting itself, that means every press is dropped unless Nexus is connected to your radio over CAT and reading its PTT: without that CAT link the button says "Sent — waiting for the amplifier" and frees up five seconds later having done nothing. A press is also dropped when the amplifier answering is not the model set in Settings. Standby does not stop a transmission (the radio keeps transmitting and its power passes straight through to the antenna), and the button's tooltip says so. The Operate button in the cockpit header is unchanged.

  • "My station" profiles: Home, a portable kit, a club call, picked from the top bar.

    Settings ▸ Station ▸ My station profiles holds named setups of your station, each with a call, grid, state, county, CQ and ITU zones, IOTA, POTA park, power, rig and antenna. Once you have made one, your callsign in the top bar becomes the picker (a small ▾ beside it; click the call, or Tab to it): pick a profile from its list, which shows each one's full name and call; hover the call to see the name of the profile you are on. It takes no extra room, so the top bar keeps its height and Stop TX and Tune stay where they are. The profile's call and grid become the ones Nexus sends on the air, and new contacts carry its details (MY_GRIDSQUARE, MY_STATE, MY_CNTY, MY_CQ_ZONE, MY_ITU_ZONE, MY_IOTA, MY_ANTENNA, MY_RIG when no radio is set up (a configured radio's model names the rig, as before), and TX_PWR where the radio did not measure the power) wherever you have not filled them yourself. Contacts already in your log never change. The profile changes only between contacts: while a run, a contact or a transmission is under way, a JS8 message is queued, a finished contact waits for Log QSO, or JTTY asks whether to log one, the picker waits and says why, so one call is never sent on the air while another is logged. In that time the callsign in Settings ▸ Station also stays as it is (the field says why), and a settings restore or reset waits too. As a last check, on every station, with profiles or without: if you change your callsign (by a profile or in Settings) while a message is waiting to go out, Nexus drops that message instead of sending your old call, and the screen says why. This covers a queued JS8 message, an RTTY Auto over or a queued RTTY, PSK or CW text, an SSTV picture, an APRS beacon and a JTTY over. The same call typed in other capitals or with spaces around it is not a change. In FT8, FT4, the other FT modes and contests, saving a new callsign in Settings makes the waiting messages again under the new call, as WSJT-X does when you press OK in its Settings: the next over goes out under the new call and counts as sent, the contact goes on, and nothing made under your old call is sent. An over already on the air finishes as it was keyed, and is not switched partway through to a message under the new call. While a Field Day or other contest entry is running, a Settings save that changes the station's callsign first asks whether to change it anyway, since an entry is operated under one call and its log and Cabrillo stay under that call; changing the operator at the key never asks. As in WSJT-X, a reply addressed to your old call is taken only when it has the same base call as the new one (K1ABC and K1ABC/P), so after a change to a different call a station still answering the old one is not worked. If the call changes just as an over is being prepared, that over is not sent and counts as not sent: the contact is not logged or uploaded, and neither "Disable Tx after sending 73" nor a CW ID follows it; it goes out under the new call at the next slot instead. A late reply Nexus still owed a station you had already worked, made under your old call, is dropped rather than held, and the screen says so, so your next CQ under the new call goes out on time. A contact is logged under the call its latest over went out under, so a contact finished on the air under your old call is logged under that call even when you changed the call before it was logged. Free text in the FT free-text box, which Nexus addresses with your call, goes out with the call you have when it is sent, including one that reads as a standard message (73, RR73, a report), which is made again under the new call like Nexus's own 73. In JS8 only the messages queued under your old call are dropped; one you queue after the change still goes out. Text you type yourself (Tx 5, the free-text box, a logger's free text) is held back only when it names a call you no longer use as the sender; a message such as "W1AW 5NN 73" goes out as you typed it. A contact the run set aside for Log QSO is logged under the station it was made from, and Settings left open while you switch profile shows the profile you are on; a Settings save only ever edits the profile you are on, and a settings file whose chosen profile is missing opens on Home. Starting a POTA activation of a profile's park switches to that profile, and Stop switches back, also after a restart; during a run or a contact the activation starts and the profile stays, and a Stop that cannot switch back says which profile to pick in the top bar once it has ended (the next activation of that park, stopped when idle, goes back by itself). Editing a profile you are not on does not interrupt a transmission planned for the next slot. The park goes on a contact only from a running activation of it, never from the profile alone; the activation's park, grid and state still win, and a profile that names a park puts its county, zones and IOTA only on that park's contacts (one with no park, on contacts with no park). A contact held for your confirm keeps the profile it was made under. As with your grid and state today, the county, zones and IOTA Nexus fills in are left out of the LoTW upload when TQSL signs with a Station Location, so it signs from that location as before; signing from the location in the ADIF, they go as stored on the contact. A profile under another licensee's call (a club call) does not transmit while a licence class is set in Settings ▸ Station, because Nexus cannot know that call's privileges; a portable form of your own call (/P) transmits as usual. Profiles are switched and edited at the station only, not from a computer driving it: from there a POTA Start or Stop never changes the call, though a Settings save from there can still change the callsign of the profile you are on. Apart from that last check, a station that makes no profile sees no change.

  • A station profile can name its TQSL Station Location.

    Settings ▸ Station ▸ My station profiles ▸ Edit has a "TQSL Station Location" box. A LoTW upload signs each contact with the Station Location of the profile it was logged under, found by the call and the 4-character grid square stored on the contact (a contact made in another subsquare of the profile's square is still the profile's; only between two profiles of one call in one square does the whole grid decide); contacts of different profiles in one upload are signed in one TQSL run per location, and each contact gets its own run's result. A profile with the box left blank, a contact no profile fits, and a contact two profiles with different locations both fit, are signed with the Station Location in Settings ▸ Services ▸ Confirmations, as before; when a profile names a location, the upload's message says how many contacts were signed that way and why (two profiles fit, or none). With "sign from ADIF location" on, those contacts are signed from the call and grid they carry, and the message says so. If TQSL cannot run for one location after others were signed, the message says what went and how many contacts were left for the next upload. If TQSL did sign one but Nexus could not confirm the upload mark was saved to disk, the message says those contacts went to LoTW, rather than counting them as left; any whose mark was not saved go again with the next upload, which LoTW ignores. A run TQSL refused is shown as refused, never as gone. A Station Location name that starts with "-", in a profile or in Settings, is refused with a message instead of being passed to TQSL, which could read it as an option. A Station Location in Settings is then needed only when a contact of the upload uses it. Changing a profile's location restarts an automatic LoTW upload that a location error had paused.

  • Sort the stations hearing you.

    The Getting Out box ("Am I getting out?") has a sort bar above its list: Distance (the most distant first, as before, and still the default), Call A–Z, Newest (the most recent report first) and SNR (the strongest first; a report without an SNR goes last). Stations that tie keep a steady order, by distance and then by call, so the list does not shuffle as new reports arrive. Each Getting Out box remembers its own choice, on Conditions, in the dashboard rail and beside a cockpit, and keeps it when Nexus restarts.

  • Satellite, Propagation and Park name columns in the Logbook

    (#423, #437). Turn them on under Columns; they are off until you do. Satellite and Propagation show the satellite name and propagation mode each contact already keeps (the ones LoTW gets), so a satellite contact can be checked at a glance. Park name shows the name of the POTA park in the Park column, read from the park list on the POTA/SOTA screen (Download or Import CSV there; without that list the column stays empty, and its header says where to get one). A two-fer shows every park name the list knows. The name is never saved with the contact and never goes into an ADIF file, since ADIF has no field for it. SOTA summits have no name list, so their contacts show none. Each column sorts like the others.

  • 12 or 16 macro keys on the CW, RTTY and PSK screens.

    Wide windows have room for more than eight, so each of the three modes now has a Macro keys choice in Settings: 8 (as before, and still the default), 12 or 16. F1–F12 send the first twelve; keys 13–16 are sent with a click and show no key name. The keys stand in groups of four and wrap onto a second row when the window is narrow. The new keys start empty and are edited like the others. Going back to fewer keys only hides the extra ones, and what they hold is still there when you choose more again.

  • Re-upload your past satellite QSOs to LoTW with the right band, in one click

    (#163). Satellite QSOs you uploaded before this release went to LoTW with the downlink band. LoTW matches a satellite QSO on the uplink band, so those can never confirm. When the Logbook finds some, a "Fix satellite uploads (N)" button appears beside Upload to LoTW. Click it and confirm, and Nexus fills in each satellite's uplink and uploads a corrected copy through TQSL, as an ordinary upload does. LoTW keeps the old record, which matches nothing and does no harm. Nothing is changed or sent until you click. QSOs LoTW has already confirmed are left alone, and so are satellites Nexus does not know the uplink of: SO-50, AO-91, PO-101, RS-44, FO-29, AO-7 (both modes), the ISS cross-band repeater and QO-100 are covered. Until the corrected copy is signed, the Logbook shows those QSOs as "Corrected, not sent yet" and counts them as not yet uploaded, so if the re-upload is interrupted or cannot reach TQSL, the ordinary Upload to LoTW sends them.

  • Type the JS8 offsets exactly

    (#448). The JS8 screen's header now has the same Rx and Tx Hz boxes as FT8 and FT4: type an audio offset and press Enter (or click away), and Rx or Tx moves there, held to 200–4000 Hz. Each box moves only its own offset. While a message waits to go or goes out, the Tx box is greyed out and says why, because JS8 keeps the TX offset where it is until the message has gone (unless Settings ▸ Digital ▸ JS8 ▸ Allow TX offset changes while sending is on). Clicking the waterfall works as before.

  • SSTV can tune to the band's SSTV frequency when you open it

    (#451). Settings ▸ Digital ▸ SSTV has a new switch, "Tune to the band’s SSTV frequency when SSTV opens", off by default. With it on, opening the SSTV screen tunes the radio to the SSTV calling frequency of the band you are on (14.230 on 20 m, for example, coming from FT8 at 14.074), as if you had picked it in the SSTV band picker. From an HF band with no SSTV frequency, such as 30 m, it goes to 14.230 on 20 m. It leaves the dial alone when it is already on an SSTV frequency (an overflow channel, or the ISS downlink during a pass), while you are transmitting or tuning, while a satellite pass is steering the radio, on 2 m (whose SSTV entry is the ISS downlink, which it never tunes to), and on a VHF or UHF band without an SSTV frequency. The tune it makes is not remembered as your own dial: go from Phone on 14.250 to SSTV and on to CW, and Phone comes back on 14.250. On 80 m and 40 m it picks the US calling frequency; the European ones are in the band picker. With the switch off, nothing changes: the screen leaves the dial where it is and tells you where pictures appear.

  • A band-plan note beside the band, never a block.

    In Phone, an amber "Outside band plan" chip appears beside the band when your signal is where no IARU region's band plan puts phone: phone anywhere on 30 m, or below the lowest phone line any region uses on the other HF bands, 6 m and 2 m (for example LSB at 7.020 MHz). Hover it for why. It is advice only: Nexus does not block the frequency, and nothing about transmitting changes. Your licence and your country's band plan decide. One worldwide table, with no region to set; 60 m is not judged, because its allocations differ country by country. While transmit is locked by your US licence class, the 🔒 chip shows instead.

  • 60 m in the Phone and CW band pickers.

    Picking 60 m outside the US parks in the worldwide 5351.5–5366.5 kHz segment: Phone at 5.363 MHz USB, clear of the 5.357 MHz FT8 frequency, and CW at 5.3525 MHz. A US General, Advanced or Extra lands on the same Phone dial, which is inside the US 60 m segment, and may transmit there on upper sideband; a Technician or Novice can listen there and stays locked. Check your own country's 60 m rules: they differ widely.

  • FreeDV contacts land in the Nexus log by themselves.

    Run FreeDV's own app beside Nexus, point its QSO Logging at 127.0.0.1:2239, and each contact you log in FreeDV goes into the Nexus log as a FreeDV contact (written MODE=DIGITALVOICE SUBMODE=FREEDV, which LoTW counts as phone), through the same import companion contacts use, so a repeat is one contact. On by default, for this computer only, and receive only: nothing FreeDV sends can key or tune the radio. A contact FreeDV logs under a call that is not your station call is refused, never written under the wrong call, and the reason shows in Settings ▸ Radio ▸ Transmit limits & sharing, in the FreeDV pane and in the connection log. FreeDV sends one time for a contact, so its start and end are the same. New guide: FreeDV beside Nexus.

  • A FreeDV pane on the Phone screen

    (add it from ⊞ Panels): whether FreeDV is talking to Nexus, the last contact it logged, any refusal, and the calls FreeDV 2.4 hears, once you turn listening on (off by default; Windows may ask once about the network). A click on a heard call puts it in the log strip. Nothing in the pane logs, keys or tunes.

  • FreeDV is a mode of its own on the Needed board

    , as FT8, JS8 and PSK31 are. This reverses the 1.18.0 note that judged FreeDV stations by their kind (phone): a FreeDV station is now "New mode" for a country until you log a FreeDV contact with it. A MODE=DIGITALVOICE SUBMODE=FREEDV contact imported from another logger is stored as FreeDV and written back exactly as it came.

  • FreeDV in the log strip's mode list

    , for a FreeDV contact logged by hand.

  • FreeDV's channels

    (1.870 to 28.720 MHz, from FreeDV's own list, each on FreeDV's sideband; 5.3685 is not offered to a US station) are in the band plan, and the SSTV channel on 14.236 now says it is FreeDV's 20 m centre.

  • Nexus in Chinese, Simplified and Traditional.

    简体中文 and 繁體中文. Pick one in Settings ▸ Appearance ▸ Language, or Nexus picks it up from a Chinese system language: a Taiwan, Hong Kong or Macau system, or one set to Traditional script, gets Traditional Chinese with Taiwan wording, and any other Chinese system gets Simplified. Completed for this release, so every screen is in Chinese. Frequencies, callsigns, reports, band and mode names stay exactly as they are in every language. On Linux, Chinese needs a CJK font installed (for example fonts-noto-cjk); Windows and macOS have one. The kits for corrections are in translations/zh/ and translations/zh-hant/.

  • The Needed board's filters now match the Spots board's.

    On the Needed view, its pop-out window and the Needed panes of the CW and Phone screens and of Connect, the board has the Spots board's search box (call, country, mode, band or frequency; PA* finds a prefix), US-state chips, and the "My privileges" chip, which hides the needs your licence class may not transmit to, judged by the same licence tables as Spots and as the transmit lockout (with "Outside the US" it hides nothing, as on Spots). Each board and pane keeps its own choices, as before, and every new one starts off, so the board shows what it did until you pick something; the search, as on Spots, lasts until Nexus closes. A state you picked keeps its chip while it filters, even once no row is in that state, so it can always be unticked (on Spots too). "Heard on my continent" shows on the Needed board as a fixed rule, with a dashed outline, rather than a switch: the board only lists cluster and skimmer spots someone on your continent heard (once your callsign is set); hover it, or tab to it, for the details. Spots' Hide worked, spotted-from chips and band follow stay on Spots, where they can change what you see.

  • A Park column on the Call Roster (#357).

    It shows the park or summit a station is activating: the reference from a POTA or SOTA spot (US-1234, W4C/CM-001), or just POTA or SOTA when the station is calling CQ POTA and nobody has spotted it yet. Click the heading to sort by it: the stations at a park go to the top and stay there whichever way the sort points, so a park hunter sees them first. The column appears only while a station on the roster is at a park (or while you have it sorted by Park), so a roster with no parks on it is as wide as before.

  • A colour of your own for each watched call (#299).

    In Settings ▸ Spots & Alerts ▸ Watch list, each entry has a colour picker: pink, red, orange, yellow, cyan or violet, or the usual lime. A watched station is then marked in its entry's colour wherever the watch list marks it: the WATCH tile on the Call Roster, the Stations list and Spots, and its row in Band Activity. Each colour is chosen to read in both the dark and the light theme.

  • Right-click a station on the Call Roster or in Band Activity (#346).

    The menu, headed by the call, puts the station on your watch list or takes it off, hides it on this band or on every band for 1 hour, 2 hours, the rest of today (UTC) or until you bring it back, and can hide every station from its country. On the keyboard, the Menu key or Shift+F10 opens it on the selected row. Hiding is a display filter only: it never changes who Nexus answers or what it sends (Alt-double-click still blocks a call from your auto-CQ). The Hide calls chip in Band Activity lists the calls hidden on one band or for a while, with when each comes back and a ✕ to show it again now; one hidden on every band for good joins the chip's own list.

  • A Start button for long contacts on the CW and Phone log line (#438).

    A contact's start time is when you put its call in the log line, and its end time is when you press Log. For a rag-chew or a net check-in whose call went in long before the contact really began, press Start beside Lookup: the start time becomes now, and the button shows the time it set. Nothing else changes: without a press the start is still when the call went in, and a time typed for another radio still wins.

  • Choose the corner your pop-ups appear in (#445).

    Settings ▸ Spots & Alerts ▸ Alerts ▸ Pop-up corner: bottom right (as before, and still the default), bottom left, top right or top left. Pop-out windows follow it too. A pop-up slides in from, and swipes away toward, its own side of the window. In every corner the pop-ups now keep clear of the transmit controls: they stand above a screen's TX dock (Phone's push-to-talk, the Send button on CW and PSK, which a pop-up in the bottom-right corner could cover before) and clear of the TX strip and the top bar's TX buttons, scrolling inside when a burst does not fit, and using the room higher up when the bottom has none (at a large zoom they can stand near the top even in a bottom corner).

  • See the FT8 transmit meters larger (#446).

    On the FT8/FT4 screen, click the small meters cell at the right end of the TX strip (or Tab to it and press Enter or Space) and a window shows the same live meters, SWR, ALC, power and the rest your radio reports, at full size. It only shows readings. Close it with its Close button: Escape still stops transmitting there, as everywhere on that screen, so the window says so.

  • A note beside Tune on the Phone screen for the Yaesu FT-991/991A, FT-710, FTDX10 and FTDX101 (#441).

    Phone keys the radio in plain USB or LSB, and on these radios the transmit audio then comes from the SSB MOD SOURCE menu, which is MIC (the front jack) from the factory. Tune sends its tone over the radio's USB audio, so with that menu on MIC the radio keys with no power. The first time a Tune on the Phone screen shows about no power for two seconds, a note beside Tune says to set SSB MOD SOURCE to REAR (USB); it stays for the session until you dismiss it, and once dismissed it does not come back. Where the radio reports no power reading during a Tune, a small ⓘ beside Tune holds the same note. Nexus does not change the radio's menus, and Tune works exactly as before.

  • CW shows the radio's scope and the audio view together.

    On a radio whose own scope Nexus draws (an Icom over its CI-V USB link, a Yaesu over its LAN scope link, a FlexRadio), the CW scope now opens in Both: the radio's scope as the wide view, and under it the fast sound-card audio waterfall for the detail inside your filter, with its Filter / Std / Full chips and HiDef. Each half keeps its controls in its title bar (G, Z, 3D, pause and scroll direction are behind its ⚙), so it is nearly all picture; on a short window, such as a 768-line screen at a large text size, the two stand side by side instead. If the radio's scope stops sending, the audio view keeps running and the radio's half says so. On such a radio the CW scope strip opens at 40 % of the screen instead of 25 % (a height you dragged is kept, and Reset on the divider goes back to the radio's own default); it changes once, when the radio's scope first appears, and never while you are transmitting or in the second after. A new RF / Audio / Both chip in the CW scope head switches between the views without moving anything below the scope, and Phone gets RF / Audio (Phone still opens on the radio's scope). The chip appears once the radio's scope has been seen, and each window remembers its own choice per mode. The radio's scope in CW now opens at ±25 kHz instead of Band (Band on 2 m asked an IC-9700 for ±500 kHz, where a CW signal is less than a point wide); a span you picked before is kept.

  • A hint when the radio's scope is slow.

    When the radio reports its scope SPEED as SLOW, or its scope reaches Nexus at fewer than four sweeps a second, the CW and Phone scopes say "Set the radio's scope SPEED to FAST for a smoother scope". Nexus never changes the radio's SPEED setting.

Changed

  • A logger on another computer now needs your say-so before its double-click can answer a station.

    With the WSJT-X UDP address pointed off this computer, a double-click in a logger on another computer used to turn TX on and call the station. Nexus now ignores it, along with that computer's free text and grid changes, unless you turn on Accept requests from other computers in Settings ▸ Services ▸ Integrations & Feeds, beside the UDP Address (off by default). If you run a logger that way, turn it on to carry on as before. With it on, anyone who can reach this computer over the network can turn TX on, because a UDP sender's address can be faked. A logger on this computer that sends from 127.0.0.1 works as it did. One reached through a multicast address answers from this computer's network address, so it is ignored too: to keep it without the switch, list one 127.0.0.1 address per logger, with commas (127.0.0.1:2237, 127.0.0.1:2238). Another computer can still stop TX, and each ignored request gets a line in the diagnostic log. A backup, a profile or a save from a remote computer can't change the setting.

  • With TX off on the FT8, FT4 and FT2 screens, the radio can stay in split.

    With Split Operation set to Rig, a split an over or a band change set up now stays on while TX is off, as in WSJT-X, with VFO B on the last TX dial (your dial moved by a multiple of 500 Hz). A hand mic or foot switch then keys VFO B there, and Nexus's licence check does not see it, since Nexus did not key it; with TX off nothing on screen warns you. Go to the Phone or CW screen first, or set Split Operation to Fake It or None. Keys that go through Nexus put VFO B back on your dial first.

  • The night side shows on a dark map.

    On the Dark map style and on the dark built-in themes (Midnight, Slate and the others) the greyline's night half was hard to tell from the day half: only the gold line said where night began. The night side is now shaded a step deeper on a dark map, while the coastlines, the grid, the stations, the TX/RX lines and the labels on it stay easy to read. The light Atlas map is shaded as before.

  • Buttons look like buttons, and status lights do not.

    On the transmit strip of every operating screen and on the top bar, TX On/Off, Tune, ATU and Hold Tx now have a visible outline and face in both themes, and the outline lights when you point at one; a button that can't be pressed right now (Tune with transmit locked out) is greyed out. The status beside them is flat: Phone's and CW's TX On/Off (which those screens arm themselves) no longer wears a button's outline, and neither do the screen headers' mode badge, the RTTY FSK/AFSK indicator or the band readout.

  • One name for the audio drive, another for the radio's power.

    The slider that sets the audio level into your radio is now called Drive everywhere: in Settings ▸ Radio ▸ Audio (it was "Tx Power") and on the FT8 and Tempo screens (it was "Pwr"), as JS8 already said. The slider that sets the radio's own output power over CAT is now called RF power everywhere: on the Phone, RTTY, PSK and SSTV screens (it was "Power") and in Settings ▸ SSTV (it was "Transmit power"). The controls themselves do what they did; only the words changed, in every language, with the tooltips and the manual to match.

  • PSK keeps its whole signal inside your privileges.

    The licence check on the PSK screen now judges the whole width of your PSK signal, not just its centre: 31.25 Hz either side for PSK31 and QPSK31, 62.5 Hz either side for PSK63. With the signal's centre closer than that to the edge of your privileges, a send and the TX switch are refused with the usual "outside your license privileges" message; move the dial or the cursor a little further in. Contacts still log at the signal's centre, as before, and the usual PSK frequencies sit well inside every segment, so they are not affected.

  • Satellite contacts now go to LoTW with the uplink band

    (#163). A contact logged through a satellite now records both frequencies: the uplink you transmitted on and the downlink you heard. For SO-50, for example, that is 145.850 MHz up and 436.795 MHz down. Linear birds record the matching spot in the passband, mirrored on an inverting one. LoTW uploads and ADIF exports write them as TQSL and LoTW expect: FREQ and BAND are the uplink (2m), and FREQ_RX and BAND_RX are the downlink (70cm). Before this, a satellite contact went out with only the downlink band and frequency. Inside Nexus nothing moves: the contact still counts on its downlink band for awards, Needed, worked-before and dupes, as before. An ADIF file from another logger that lists a satellite contact the same way, with both frequencies, is read back onto its downlink band too. Re-importing a file or restoring a backup finds the satellite contacts you already hold instead of adding them twice, and LoTW's confirmations of these contacts still find them. Log Check checks a satellite contact's band against the frequency you heard it on. Taking the satellite off a contact, in the Log QSO window or the Logbook, puts its frequency back to the one you heard it on. Split contacts on one band and FT contacts are unchanged. As in WSJT-X, an FT contact logs the dial plus your TX offset as its one frequency, even when you receive at another offset.

  • SSTV remembers "My picture already shows my callsign" for that picture (#456).

    Tick it once for a QSO card that already carries your call, and the next time you load that same picture it comes back ticked, so the callsign plate is not drawn over your card again. It is remembered by what is in the picture, not its file name: a different picture, even one saved under the same name, starts unticked and gets the plate, as every new picture always has. Untick it and that picture is forgotten. It is also remembered for your callsign: switch to a station profile with a different call and the picture starts unticked again.

  • The SSTV, RTTY and PSK screens are laid out like CW and Phone.

    Their panes (the gallery and the Transmit pane on SSTV, the decoded text and the log on RTTY and PSK, and the RF scope and Stats panes when you add them) now stand in one pane area under the TX strip, as on CW, Phone and JTTY. Where they do not all fit, that area scrolls on its own and the rest of the screen stays where it is. On SSTV the picture stage takes a little more of the height than before.

  • One control sets the radio's scope span.

    In CW and Phone, the span chips above the radio's scope (Band, Full, ±250k, ±100k, ±50k, ±25k, ±10k, ±5k, ±2.5k, and on a FlexRadio also ±500k and ±1M, its 1M and 2M) are now the only control that changes the radio's span. The separate span buttons in the Rig scope and Flex pan boxes are gone, because the two could disagree: a span set there lasted only until the next band or chip change, while the chip stayed lit for a span the radio was no longer on. The Ref level stays in those boxes. Every span those buttons offered is a chip now. HiDef / Fast is no longer offered while the scope draws the radio's own scope, where it did nothing.

  • The audio waterfalls and scopes reach 5 kHz, and open wide.

    Every audio waterfall (FT8 and the other FT modes, RTTY, PSK, JS8, SSTV, JTTY, the Tempo strip and the torn-off waterfall) and the CW and Phone audio scopes now draw the receive audio up to 5000 Hz instead of 4000, at the same fine detail, so the top of the radio's filter and the dark stopband above it are on screen. Full in each of them now runs to 5 kHz (RTTY's widest button reads 5k). The CW scope now opens on Std (200–3000 Hz) and the Phone scope on Full (0–5000 Hz), instead of on the radio's filter width; the Filter chip (CW) and the Auto chip (Phone) are still there for zero-beating or framing just the filter, and a chip you picked before is kept. When the radio reports its filter, the picture's black point is now set from the receiver's noise inside that filter, so a wide view over a narrow filter (a 500 Hz CW filter in Std, a 2.4 kHz SSB filter in Full) keeps a dark background with the signals standing out, instead of lighting the whole passband. Above about 4.5 kHz the picture fades a little from a 48 or 44.1 kHz sound card: that is Nexus's own input filter, the same one the decoders use, not your radio. Nothing changes in what any decoder hears or in where you can transmit: each mode keeps its own offset limits (FT8's 200–4000 Hz, RTTY's and PSK's 300–3700 Hz), so a click above them lands on the limit.

  • Your radio's filter is marked on every audio waterfall.

    When the radio reports its receive filter width, the FT, RTTY, PSK, JS8, SSTV, JTTY and Tempo waterfalls (docked and torn off) draw a thin line where that filter ends, so a dark band above it reads as the radio's filter, not as a waterfall cut short. The radio reports only the width, so on USB and LSB data modes the line is drawn at that width above 0 Hz; the real edge sits a little higher, by the radio's low cut. It is drawn only in USB and LSB and their data modes, at filter widths of 1.8 kHz or more: in AM, FM, CW and FSK RTTY, and for a narrower data filter (which some radios centre on a data shift), a width alone does not say where the filter sits, so nothing is drawn.

  • A frequency scale under every waterfall and scope, so you can read how wide it is.

    The Phone and CW scopes, and the radio's own panadapter (Icom, FTDX10, FlexRadio, and a TCI or SDRconnect radio), now have the strip of numbers along the foot that the FT8 waterfall has: audio hertz on the audio scopes (0, 500, 1000 … out from the dial on Phone, rising to the left on LSB; on CW the same hertz your pitch is set in, so the dashed pitch line sits on its own number), and on a radio's panadapter the distance from the dial (-10k … 0 … +10k). The MHz numbers stay just above the strip. On every waterfall the scale now follows the span: 500 Hz steps on Std and Full as before, 200 Hz at 2 kHz and 1.5 kHz, 100 Hz at 1 kHz and 600 Hz and on the CW Filter span, and wider steps when the waterfall is narrow or the text is large, so the numbers never run together. The markers, the filter-edge line and the passband stop above the strip, and on the Phone and CW scopes no waterfall row is hidden under it.

  • The Phone scope says when your radio's filter is what you are seeing.

    As the CW scope already does, the Phone audio scope now says "Showing your radio's 2400 Hz filter — widen it on the radio to see more" when its window is wider than the filter, and on an Icom whose own wide scope Nexus could draw, it says which setting brings it in (Native Icom CI-V, or 115,200 baud), with a link to it.

  • The diagnostic log says who keyed your radio, and how.

    nexus-diag.log now has one line each time Nexus keys and unkeys the transmitter, naming what keyed it (an FT8 or other slot over, your PTT, Tune, the CW or voice keyer, the Remote microphone, or a program on Nexus's CAT share or TCI server) and how (CAT, the DATA input, RTS or DTR, or VOX), with how long it stayed keyed. When the radio reports it is transmitting while Nexus holds nothing, the log says the radio was keyed by something else (another program on a shared rigctld or CAT port, VOX, or the radio's own PTT) and when that ends; a radio still reporting transmit in the second or so after Nexus lets go is written as the end of that transmission (VOX or break-in hang, a tuner, a slow changeover) instead. The PTT method is written once when the radio connects and again whenever it changes, with whether the CAT share is on and whether "Other programs may key transmit" is on; a TX watchdog trip now says what it stopped. Only changes are written, never a line per second, and a burst is capped. Nothing about when or how the radio keys has changed. When the decoder finds nothing for a while and the receive audio is over full scale, the line now says the audio is clipped and to lower RX Gain, then the input level, instead of saying the audio is arriving.

  • FT8, FT4 and the other FT modes: a station answering your CQ is worked only when it calls your exact call.

    As in WSJT-X, a new caller is answered only when its call to you names the callsign you are using, letter for letter: while you are K1ABC, a station calling K1ABC/M or W1/K1ABC is not answered, and once you change to K1ABC/P, one still calling plain K1ABC is not answered either. Nexus used to answer any of them, since they share your base call; at a multi-station special event that could be a call meant for another station. This holds when calling CQ outside contests, in contests that use the standard or ARRL VHF messages (as it already did in Field Day), and for the best-caller choice, so a stronger station calling another form of your call no longer hides one calling you. A hashed form of your own call (a station shown as calling <K1ABC>) is still your call. Once a contact is under way, nothing changes: the replies of the station you are working are still taken when they carry your base call.

  • The CW scope says when your radio's filter is all it can show, and how to get the wide view.

    The CW scope draws your receive audio unless your radio streams its own panadapter, and that audio has been through your radio's CW filter, so with a 500 Hz filter nothing outside those 500 Hz can ever show, whatever span you pick. In the Std and Full views, where the window is wider than the filter, the scope's toolbar now says "Showing your radio's 500 Hz filter — widen it on the radio to see more." On an Icom (IC-7300, IC-7610, IC-705, IC-9700, IC-905) connected over USB with Native Icom CI-V off, the toolbar also says the radio can show its own wide scope there and links straight to the switch in Settings ▸ Radio ▸ Rig & CAT ▸ Advanced; with the switch on but the serial speed below 115,200 baud (where the radio does not send its scope), it says to set 115,200 in Nexus and on the radio. On a radio with its own scope whose picture is coming from the sound card (an FT-710 without its SCU-LAN10 link), the Filter, Std and Full span buttons are now shown too, instead of the view being held at Filter.

  • One sun for the whole station.

    The sun on the map and the globes, the night side and grey line, the Greyline box's sunrise and sunset, and the propagation model's solar elevation now all come from one more precise sun position (within about a hundredth of a degree, where the old one could be a degree off around the equinoxes, moving sunrise and sunset by a few minutes). The ITU-R P.533 prediction keeps its own reference sun, so its numbers do not change.

  • The Kp says whether it was measured or estimated.

    The dashboard bar's Kp now shows NOAA's measured three-hour Kp while its period ended within the last three hours, with that period ("Kp 5 21Z–00Z"), and otherwise NOAA's one-minute estimate, marked "est."; hover it for which it is. The geomagnetic storm notification (and the aurora layer turning itself on) goes by whichever of the two is higher, so a storm that has just begun is not held back until NOAA's next three-hour figure, and it says which one it used: "Kp 6.3 est. 22:36Z" or "Kp 5.7 measured 21–00Z".

  • A better default size on a scaled Windows display.

    On a screen Windows scales to 125% or 150% (most 1080p laptops), Nexus now opens at 80% or 70% instead of 90% or 75%. At the old size the top bar, the cockpit headers and the RTTY and PSK transmit areas wrapped onto extra rows and the decode panes shrank to a few lines; at the new size they sit on one row and the panes get the room back, and text stays at least as large as it would be with Windows scaling turned off. The first time this happens Nexus asks once whether to keep the new size or go back to the old one, and remembers your answer; Settings ▸ Appearance ▸ UI scale shows that answer under Auto and lets you change it at any time (Auto or Off), and the size follows at once. Screens without scaling, a scale you set yourself in Settings ▸ Appearance, and pop-out windows are not changed, and the size follows the window when you drag it to a monitor with different scaling.

  • Band picks outside the US land where every region allows them.

    With "Outside the US" as your licence, a Phone or CW band pick now parks inside the amateur allocation shared by every IARU region: 160 m phone at 1.8458 MHz LSB (it used to be 1.8028, below the 1.810 MHz that the UK, Japan and Region 1 start at), 70 cm at 432.100 MHz (it was 420.000, outside 430–440 MHz), 23 cm phone at 1296.150 MHz, and 13 cm at 2400 MHz. The other bands keep their dials.

  • The setup wizard says plainly what "Outside the US" does.

    It used to say "No transmit limits". It now says Nexus won't block any band, warns when you're outside the usual band plan, and that your country's rules are yours to follow. Settings ▸ Station says the same.

  • Spots from outside the US are classed by their own band plan.

    A cluster or skimmer spot's mode (CW, Phone or Digital) comes from its frequency. It used to be judged against US edges for every station, so a European, Asian or Pacific SSB station on 7.053–7.125, 14.112–14.150 or 21.151–21.200 MHz was classed Digital: it missed the Phone health light and showed the wrong mode on the Needed board. A station outside the US is now judged by the IARU Region 1 band plan; a US station keeps the US edges, which bind it by law.

  • More room for the band on a 1080p screen.

    On a 1920×1080 monitor, and on a 1080p laptop with Windows scaling at 125 % or 150 %, the parts of Nexus around your decodes and your log take less height, so the panes get it back:

    • The transmit area at the bottom is smaller. In Phone and CW the SWR, ALC, power and compression meters now sit at the right end of the TX strip under the waterfall (the one with TX On, Tune and Stop TX), the way FT8 already shows them, instead of a row of their own above the push-to-talk button. Phone's transmit area is now a single row: PUSH TO TALK, Lock, then what goes out when you key (VOX first when it is on, then TX frequency, mode, split, XIT, power). On Phone and CW, PUSH TO TALK, Lock, Stop TX, Tune, TX On/Off, Esc Stop and every F-key are where they were and at least as big as before; Send is wider (and shorter, except in JTTY) (see below), no smaller a target.
    • The top bar is one row. The eleven mode buttons (Tempo Fast through FST4W) are now one button showing the mode you are on, for example "FT8 ▾"; click it, or press Enter or ↓ on it, to pick another mode from the list. The NOW line (band, getting out, top need, the feed lights) moves up into the top bar instead of a line of its own. Auto / Tx 1st / Tx 2nd and the TX buttons work as before. On a screen 2400 or more points wide (a 4K monitor at 150 %) the bar is as it was.
    • Auto / Tx 1st / Tx 2nd are narrower on these screens: the small words under them ("cycle", "even", "odd") are left off, and each button's tooltip still explains it. They work exactly as before.
    • The CW and Phone headers are one row. Memories, the rotator, split and REC are behind a ⋯ button at the end of the header; it opens under the header, and Esc closes it. While split is on or a recording is running, the ⋯ button says so. The CW speed slider is shorter there; PgUp and PgDn still change the speed. The FT8 header is unchanged. On the 1080p laptop at 125 % the CW header still takes two rows.
    • Desktop-sized controls at Standard density. Form fields, the send rows of CW and JS8, the Settings inputs and the Logbook buttons are now 30 points tall instead of 44 (a mouse-sized target), so the log forms in the cockpits stop scrolling and the transmit areas get shorter. Settings ▸ Appearance ▸ Workspace ▸ Density ▸ Comfortable brings back the 44-point controls, and Touch still makes them 48. The Send buttons are wider, so they are no smaller a target than before; PUSH TO TALK, Stop TX, TX On, Tune and every Esc Stop are at least as big as before (JTTY's, RTTY's and PSK's send rows, which hold their Esc Stop, are 44 points tall).
    • More contacts in the Logbook. Rows are closer together at Standard density (19 contacts on a 1080p screen instead of 13); the ✎ and ⋯ buttons on each row are still 32 points square. Comfortable keeps the old spacing.
    • A bigger Band Activity in the FT8 Roster layout. By default Band Activity takes three times Rx Frequency's height in the right-hand column (it was 1.6 times), and where it is short its filter buttons fold behind its Filters button, so more decodes show. Dragging the divider between them still sets your own split.
    • Tighter spacing on shorter windows. Where the window is about 1100 points tall or less (1080p at any scaling), the gaps and padding in the top bar, the cockpit headers, the TX strip and the transmit area are smaller. Buttons, text and click targets keep their size.

Fixed

  • FT8 and FT4 after a band change, on a radio slow to take the split.

    With Split Operation set to Rig, the first over on a new band had to move the radio's TX VFO to that band just before keying, and a radio that takes longer than half a second for that had the over refused: "FT8 not keyed: the radio did not take the split TX frequency", and transmit halted. Nexus now does what WSJT-X does: when the band changes, it puts the new band's TX frequency (and its mode) on the radio at once, while you are receiving, whether transmit is on or off, so the first over on the new band finds it already set. Turning transmit off no longer takes the radio out of split on the FT8, FT4 and FT2 screens; leaving them, changing Split Operation, picking a mode other than FT8, FT4 or FT2, changing band or radio, and an over the radio refused still do. Nothing is written to the radio while it transmits. A band change in the last 4 seconds or so before your own slot, with transmit on by then, is the exception to "at once": the frequency goes to the radio as that over starts, so a slow radio starts that over late, with its first symbols lost (at worst, on FT4 over a network or slow serial link, the whole over). Turning the radio's own dial writes nothing, and a cluster split you set with SPLIT-on-Work is left alone. If the radio does not take it, Nexus tries once more and then leaves it to the over, which still keys nothing if the radio is too slow then. A radio slower than about 150 to 200 ms per write can still refuse the first over after Call CQ on a new band: Call CQ sends the mode again, so the key has to repeat the TX frequency, mode and split inside its half-second limit, and the other ways of turning transmit on go the same route. Those figures are from a simulated radio. Fake It and None are unchanged. On FT2 this is not fixed: the radio stays in split as above, but its 3.75 s slot is too short for Nexus to set the frequency ahead of time, so a slow radio can still refuse the first over after a band change and halt transmit.

  • Stop TX and Send stay reachable on a small window at a high zoom.

    On CW, Phone, RTTY, PSK, JS8 and JTTY, a window as small as 900×600 or 700×800 at 175 % had a transmit panel taller than the screen: it covered TX On/Off, Tune and Stop TX at every scroll position, pushed Send (or PTT) off the window, and left ⊞ Panels impossible to click. Where the panel does not fit, its macro keys and meters now scroll with the rest of the screen, and only the row with Send (or PTT) stays at the bottom, so Stop TX and Send are always on screen. On RTTY and PSK, the TX (Continuous) and Esc/Stop buttons now sit at the right end of the Send row on every window size, so they stay at the bottom too, and they are a little taller, the same size as JTTY's Esc/Stop. CW's Send no longer runs off the right edge of a narrow window, and the ⊞ Panels menu stays inside the window and never covers Stop TX: it stands beside Stop TX rather than over it, no longer hangs off the left edge, and where it does not fit below its button (at 150–175 %) it shortens to the room above the transmit panel, or opens upward if there is more room there, and scrolls so that every entry can be reached. Where the panel has folded, pressing Tab keeps the focused button out from behind the bottom row, JTTY's notice or refusal stays on screen in about two lines (hover over it or click it for the whole text), and JS8's heartbeat menu opens in sight. One limit remains; see the next bullet.

  • Known limits in this release.

    On a small window with the panel folded (measured at 700×800 and 800×700 at 175 %), a second click at the same spot 500 ms or later on a macro's ✎ pencil, when the pencil is just under the bottom row, can land on Send once on CW and send that macro. The same click can land on Reset this button on CW, RTTY and PSK, which does not ask first and can erase a customised macro. Scroll until the pencil is fully in view before you click it. Stop TX and Esc/Stop stay clear. Separately, with Split Operation set to Rig and a slow radio, changing to FT2 with TX on can, rarely, key a whole FT2 over in the other station's period; turn TX off before you change to FT2. That was seen on a simulated radio.

  • An Icom's own scope draws every sweep it sends.

    With an Icom on its USB CI-V link (or the Icom network connection), the CW and Phone scope, and the RF scope pane, used to show only the newest sweep each time Nexus got round to it, so sweeps that arrived while Nexus was busy asking the radio for its frequency and meters were never drawn and the waterfall moved in steps. The sweeps now reach the screen in order. With reduced motion on, the scope draws fewer rows, always the newest sweep, so it never lags the radio, and the sweeps it leaves out are not counted as dropped. If the screen falls behind for another reason (a very slow computer), it skips ahead to the live picture rather than lagging behind it, and the sweeps it skipped are counted. In testing against a simulated radio, keying, PTT and the frequency and meter reads were not held up by the scope. Nexus now also measures how many sweeps a second the radio really sends and reads the radio's scope SPEED setting (Fast, Mid or Slow; Nexus never changes it), and shows both in Settings ▸ Radio ▸ Rig & CAT ▸ Advanced (Radio scope stream) and in the diagnostic log (at most once a minute, sooner when the rate changes or a sweep is dropped), so a slow waterfall can be told apart from a radio set to Slow. This has only been tested against a simulated radio. If the scope misbehaves on yours, set Settings ▸ Radio ▸ Rig & CAT ▸ Advanced ▸ Waterfall source to Audio (sound card) and Save: the radio stops sending its scope and Native Icom CI-V stays connected for CAT.

  • A LoTW satellite confirmation lands only on a satellite contact

    (#163). As LoTW itself does, a confirmation that came through a satellite is matched only to a contact you logged through a satellite, and a confirmation without one only to a contact without one. Before, a satellite confirmation could be marked on an ordinary 70 cm contact with the same station made around the same time. An eQSL or QRZ confirmation, or a QRZ logbook download, never pairs a satellite contact with an ordinary contact on its uplink band either, so a 2 m simplex QSO is never taken for an SO-50 one.

  • Picking FST4 from another band lands on a band your radios are set for (#184).

    With an HF radio set to 160 m–4 m and a VHF radio set to 2 m and 70 cm, switching from 20 m FT8 to FST4 sent the dial to 2200 m, which neither radio was set for. A mode that has no channel on your band now moves to its first channel on a band one of your radios is set for (160 m here). With one radio, or a radio with no bands set, nothing changes: FST4 still starts on 2200 m. MSK144 and Q65 follow the same rule. (WSJT-X leaves the dial where it is in this case; Nexus moves you to the mode's own frequency on purpose.)

  • FST4W lists all of WSJT-X's FST4W frequencies (#184).

    The FST4W band picker offered only 2200 m, 630 m and 160 m. It now has WSJT-X's whole FST4W table: 2200 m, 630 m, 160 m, 80 m, 40 m, 30 m, 20 m, 17 m, 15 m, 12 m, 10 m, 6 m (50.293, which WSJT-X lists for Regions 2 and 3), 4 m (70.091, Region 1) and 2 m. A listed frequency is not permission to transmit: your licence class still decides, as everywhere else.

  • A radio pointed at your own rigctld says which setting to change (#165).

    With a radio model such as an FTDX10 on Connection Network at port 4532, where a rigctld you run already serves the radio, Nexus sent the radio's own CAT commands to that rigctld, which hung up, and the CAT error read like a radio switched off ("rig power is off?", "Connection reset by peer"). When the CAT check fails on that setup, the error now adds that 4532 is rigctld's own port and that Rig Model should be NET rigctl (remote rigctld), which talks to your rigctld as one more client beside your logger.

  • A band change through OmniRig reaches the radio (#161).

    On a radio whose OmniRig rig file sets the frequency only per VFO (VFO A and VFO B, no plain frequency command), picking a band or clicking a spot did nothing: OmniRig accepted the change, never sent it, and the band picker snapped back to the old dial. Nexus now sets the frequency the way the rig file allows, on the VFO the radio is on, the same way WSJT-X does, and reads the dial off that same VFO. If the rig file has no way to set it at all, Nexus refuses the change and says why ("OmniRig's rig file for this radio has no command to set the frequency of the VFO the radio is on") instead of looking as if it worked. On a rig file that cannot tell which VFO the radio is on, Nexus selects VFO A once, as WSJT-X does, so the change lands on the VFO you are on: when it connects, or just before the first frequency change if OmniRig did not have the radio yet, and never while the radio is transmitting. A rig file with a plain frequency command is set as before.

  • The decode panes fill in the periods they missed

    (#452). Band Activity and Rx Frequency only took in decodes while they were on screen, and the station sends them one period at a time, while your own transmissions came back every time. So a pane that was closed, a window macOS or your system stopped drawing in the background, or a pop-out out of sight, came back showing your report and RR73 to a station whose calls to you it never showed, as if you had answered nobody. Now, when a pane comes back after missing periods, it fills them in from the decodes the station still holds, in period order and never twice; their time is drawn in italics, and its tooltip says the row was filled in afterwards. Periods the station no longer holds are marked with a dashed "Periods not received" line (ALL.TXT still has every decode). Filled-in rows are only shown: they raise no alert, sound or pop-up, and nothing about what Nexus transmits or how it runs a contact changes. On a computer driving another station, the panes are not filled in. The Call Roster was never affected.

  • An Icom's scope comes back after you switch the radio off and on (#463).

    On an IC-7300 or IC-9700 on Native Icom CI-V, turning the radio off and back on while Nexus stayed open left the scope dark until Nexus was restarted: the radio forgets that it was sending its scope when it powers down, and Nexus had asked only once. Nexus now notices when the radio stops answering for a second or more and, as soon as it answers again, turns the scope stream back on once. A single missed reply does not count, and nothing is sent about the scope when you have it switched off in Nexus. This has not been confirmed on a real IC-7300 or IC-9700 yet, so please tell us how it goes.

  • CW no longer goes out through a keyer you did not pick (#434).

    With Serial keyline or WinKeyer chosen as the keyer but its port box left blank, Nexus quietly keyed CW through the radio's own CAT keyer instead, and the CW screen looked as if it was sending. Now it sends nothing and says why on the CW screen and in the diagnostic log: "Pick the keyline serial port in Settings ▸ CW" (or the WinKeyer port). With the port filled in, CW keys exactly as before.

  • The Yaesu FTX-1 gets the AGC speed you pick (#403).

    Through Hamlib, picking Fast put the FTX-1 on Mid, and picking Mid or Auto was refused, so Nexus said it couldn't set AGC. The speed the radio reported was skewed the same way: Mid showed as Fast, and Auto (its Auto-Fast) showed as Mid. Nexus now sends the FTX-1 its own AGC codes, from Yaesu's CAT manual, and reads its answer the same way, so Off, Fast, Mid, Slow and Auto each set and show what they say. Every other radio is sent exactly what it was before. This changes what the FTX-1 is told and has not been confirmed on one yet, so please tell us how it goes.

  • The SSTV, RTTY and PSK top bar stays on screen (#457).

    On a small or zoomed window these screens scroll, and the bar with the mode, the frequency, the band, ⊞ Panels, the power slider and the CAT light used to scroll away with them, so you had to scroll back up to see or change your frequency. It now stays at the top of the screen, with the TX strip (TX On/Off, Tune, Stop TX) right under it and the transmit controls at the bottom, and only the waterfall and the panes scroll between them. On a window too small to hold the bar, the TX strip and the transmit controls at once (for example the smallest window at 175 %), the bar scrolls as before, so it never sits over Stop or Send.

  • With an amplifier, a contact logs the amplifier's power

    (#439). With an amplifier set up on the radio, the power on a logged contact (TX_PWR) was always the figure in Settings ▸ Station power, because the radio's own reading is only the amplifier's drive. It is now the highest output the amplifier itself reported while you were transmitting for that contact, with the same timing rule the radio's reading uses. While the amplifier is in Standby, the radio's power passes straight through to the antenna, so the radio's own reading is logged instead. When the amplifier is not answering or reported no output for that contact, the log uses Settings ▸ Station power as before. Without an amplifier nothing changes.

  • SSTV pictures no longer fall apart into sideways-shifted bands (#422).

    If a moment of receive audio went missing in the middle of a picture (even a few hundredths of a second, as can happen when the computer is busy), every line after it was read from the wrong place: the picture came out in horizontal bands, some shifted sideways with their colours split down the middle, and often the top of the picture too. Slow modes such as Wraase SC-2 180 were hit hardest, simply because they take longer. A sound card whose clock runs a few hundred ppm off, on a noisy signal, also tore rows. Nexus now finds every line's own sync pulse and reads each line from there so the picture stays straight; lines that were truly lost in a gap are not there, but nothing after them shifts. A picture too noisy for its sync pulses to be made out decodes exactly as before. No setting to change.

  • Phone and CW at the top of the left rail are whole again.

    When the rail had more sections than fit and showed its scrollbar, the Phone and CW buttons were squeezed shorter than their icon and label, so the microphone was cut off at the top and Phone's label ran into CW's icon. They keep their full height now, and the rail scrolls instead.

  • An Icom on Native Icom CI-V goes back to the filter you had in each mode (#417).

    Staying in Phone already kept your filter, but going from FT back to Phone (USB-D to USB) put the radio on FIL2: when an Icom changes mode without being told a filter, it picks that mode's default. Nexus now notes which filter (FIL1, FIL2 or FIL3) your radio has lit in Phone, in the data modes and in CW each time it checks the mode, and asks for that filter when it changes back, in the same command that sets the mode. Going into FT likewise brings back your data filter instead of always FIL1, and a band change that flips USB to LSB keeps your Phone filter. Until Nexus has seen the filter for a mode, and always in AM, FM and RTTY, the radio is told exactly what it was told before. Nothing is saved, and switching radios starts fresh. The Hamlib connection is unchanged, because it already carries the filter number across a mode change. This fix is for Native Icom CI-V only and has not been confirmed on a real Icom yet, so please tell us how it goes.

  • Companion mode no longer hears its own decodes back.

    With Nexus taking its decodes from WSJT-X or JTDX (the Companion signal source on the Operate screen) and the WSJT-X UDP API turned on, both at their stock address (127.0.0.1:2237), each decode Nexus passed on to GridTracker or JTAlert landed back on its own companion address and was shown again the next period as a new decode. Every period's decodes piled up on the next, in the decode list, Band Activity, ALL.TXT and the stations heard, and old messages could reach the auto-sequencer as if just sent. Nexus now ignores decodes that carry its own name, so each period shows only what WSJT-X decoded. Decodes from WSJT-X, JTDX or MSHV come in as before, and Nexus still passes them on to GridTracker and JTAlert.

  • FreeDV contacts now reach QRZ, Club Log, eQSL and your other logging services.

    They used to land only in the Nexus log (and LoTW), never in the services you upload to. FreeDV contacts now also follow the rules above for other programs' contacts: each shows "Logged K1ABC from FreeDV", Settings shows the last one, one you also logged in Nexus within five minutes is not written twice, a contact already in the log is left exactly as it is, and your running POTA or SOTA activation is added to it.

  • A contact WSJT-X or JTDX logs now reaches the Nexus log in companion mode.

    With Nexus taking its decodes from WSJT-X (the Companion signal source on the Operate screen), WSJT-X sends the contact it logs to the same UDP address it sends its decodes to, and Nexus kept only the decodes, so the contact never reached your log. It is now imported there, by the same rules as a contact another program sends to the contact port: only under your own station call, a contact already in the log, sent twice, or sent to both addresses is still one contact, it goes to your logging services like your own contacts, and its card says it came via WSJT-X. It arrives with the next period's decodes, and only from a WSJT-X on this computer; Nexus's own logged contacts, which the WSJT-X UDP API sends to that same address with the stock settings, are not imported a second time. Nothing about decoding, transmitting or timing changes.

  • A rotator reached over the internet no longer "loses connection" on every dropped reply.

    The Rotor pane used to say "The rotator controller isn't answering" the moment one position read went unanswered. Now it keeps the last bearing (and, on an az/el rotator such as a G-5500, the last elevation), dimmed, with "Last reading", until three reads in a row have gone unanswered (about six seconds), and only then says the controller is not answering. Nexus also waits longer for an answer: up to 3 s for a position and 4 s for a move, which covers the time Hamlib itself takes to retry a GS-232 controller (up to 1.8 s) plus a round trip across the internet. The rotctld Nexus runs for an EasyComm, GS-232 or SPID controller now gives the controller 1 second to answer each command instead of Hamlib's 0.2 to 0.4 seconds (and, on a GS-232A, GS-232B or SPID, one retry instead of three): on a slow link Hamlib's short wait made a position read return the answer to an earlier read and sent the controller extra commands. On a cable nothing changes except that a controller that is switched off is called silent a little later (about 1 to 2.6 seconds, depending on the model).

  • The rotator boxes no longer queue up behind each other.

    The rotator strip in each screen's header, the Rotor box, the Needed board's rotator and a satellite or Moon track each used to ask the rotator for its position on their own, and rotctld answers one question at a time, so over a slow link the questions piled up and a ■ STOP had to wait its turn behind all of them. Now they share one question: whoever asks while one is on its way gets that answer, so the rotator is asked once however many boxes are open, and a STOP waits behind one position read at most. Moves and STOP still go straight to the rotator, and a satellite pass that has just turned the antenna still asks where it went with a question of its own.

  • A rotator with an IP address in its Port box reconnects after the network drops it.

    With a serial server or a controller's own network port in Settings ▸ Radio ▸ Rotator ▸ Port (for example 192.168.1.50:4001), Hamlib's rotctld connects once and never again: when a router, the internet provider or the serial server dropped that connection, rotctld kept running and failed every command, and the rotator stayed dead until you restarted Nexus. Now, when rotctld reports a lost connection (an I/O error) three times in a row, Nexus restarts it, and the new one connects again. If the new one connects and loses the connection again, Nexus restarts it after 30 seconds, then at longer and longer intervals, up to five minutes apart, until the rotator answers. If the far end cannot be reached at all, the new rotctld stops, and Nexus starts it again the way it always does for a rotctld that stops, so the Connections log shows rotctld starting and stopping until the link is back. The Connections log and the diagnostic log say when the first restart happens and when the rotator answers again (the diagnostic log does not include the address). A rotator that is only switched off, behind a serial server that still answers, is not restarted: that is not a lost connection. A serial port (COM5, /dev/ttyUSB0) is not affected. A link that goes silent without being closed is reported as lost only once your computer gives up on it, which can take some minutes; until then the Rotor box says the controller is not answering.

  • A satellite pass takes the rotator back when it answers again.

    After five commands in a row went unanswered, a pass let the mast go for the rest of the pass, even once the link came back. It now asks the rotator where it is every 15 seconds and, as soon as it answers, points it again, elevation included on an az/el rotator, and says so in the Connections log. A rotator that cannot report its position (the Hy-Gain DCU-1) is still left to you for the rest of the pass. Pointing the antenna at a station (Point, LP, → CALL, the Needed board's Go) while a pass is turning the rotator now ends that pass, as ■ STOP does, so the pass does not turn it back; a pass still waiting for the satellite to rise stays armed, with its transponder, since the rotator is yours until five minutes before it rises.

  • One slow reply no longer drops a satellite pass to azimuth only.

    A G-5500 or other az/el rotator that answered one command late (Hamlib's own timeout) was treated as having no elevation axis, and the pass tracked it flat for about a minute. Now only the rotator saying it does not take an elevation counts, and it has to say so twice in a row.

  • The diagnostic log now records the rotator.

    It says when a rotator stops answering and when it answers again (why, and after how long), gives a summary each minute while the link is slow or missing, and records the rotctld Nexus runs starting, stopping and exiting, and when an External rotctld starts or stops being used. A rotctld on another computer is written as "a remote rotctld" and its port, never its address.

  • Switching to another radio no longer carries the last radio's late-key watch over to it.

    After Nexus gives up on a key that may still reach the radio late, it watches that radio's own PTT for a few seconds and unkeys a key it let go of. If you picked another radio or another rig model in Settings in those seconds, the watch carried over to the new radio, and pressing that radio's own microphone was unkeyed with a line saying the radio keyed after Nexus let go. Likewise an ATU tune-up just started on the old radio held off Nexus's unkey of a late key on the new one for up to eight seconds. Both now end with the old radio, as they already did on a two-radio switch. A reconnect to the same radio (a port or baud change, a restarted CAT helper) keeps them, and so does any change made through a rigctld you run yourself, which may still hold the late key; a late key still owed when the CAT link is reconnected to the same radio is now watched for too (#429).

  • A sound device Windows renamed after a USB change can be picked again in one click (#425).

    When Windows renumbers a radio's USB codec ("Speakers (2- USB AUDIO CODEC)") or its real name has a double space, the saved name no longer matches and the radio has no audio. Settings ▸ Radio ▸ Audio now shows a "Use … instead" button under the RX or TX picker when exactly one listed device differs from the saved one only by that Windows number or the spacing, and the audio error names that device too. Two identical codecs, or a device another of your radios uses, are never offered. Nexus never switches by itself: press the button, then Save. The "saved, not in the list" entry shows spaces you could not see before as ␣, so it no longer looks the same as the working entry.

  • Linux: the headphone monitor and the app's sounds keep off the radio's codec after a reboot swaps card names.

    Linux names two cards of one chip family CODEC and CODEC_1 in the order it finds them, so after a reboot the name you picked for TX audio can point at your headphone box while the radio's codec takes the other name. Nexus now notes the USB model of the TX sound card when you pick it, and refuses the monitor and the sounds on every card of that model, whatever it is called now. A TX device picked before this version is guarded as before until you pick it again in Settings ▸ Radio ▸ Audio and Save (choosing another device and then yours again is enough).

  • With a Flex on OmniRig or TCI, the monitor and the app's sounds no longer play into the system default output that carries your transmit audio.

    On those connections Nexus's own Flex client (and its DAX audio) never runs, so the radio transmits from the sound card, but Nexus treated it as DAX audio and left a TX device of "System default" unguarded.

  • FT8 and the other FT modes: your RR73 goes once to a station that has gone quiet (#416).

    When you were not calling CQ (you answered a station, or worked one that called you) and closed the contact with RR73, Nexus sent that RR73 again every period until the Tx watchdog or Stop, even when the other station had gone silent. Now it goes once, as in WSJT-X, and with "Disable Tx after sending 73" on (the default) Enable Tx goes off after it, as it already did after your 73. If the other station repeats its report because it missed your RR73, Nexus still answers it with RR73 while Enable Tx is on, and their 73 still closes the contact. An RRR (Settings: RRR instead of RR73) is not a sign-off, so it still repeats until they answer, as in WSJT-X. A CQ run is unchanged (one RR73, then back to CQ), except with both "Auto-log QSOs" and "Prompt before logging" off: there the run's RR73 used to repeat every period until you pressed Log QSO, and now it goes once and the run waits for Log QSO.

  • Play receive audio here works on a headphone DAC that Linux names after your radio's codec.

    A USB headphone box or headset built on the same chip family as the radio's sound card gets the same card name from Linux with a number added (CODEC_1 beside the radio's CODEC, Device_1 beside a Device), and since 1.22.0 the monitor treated the two as one card and turned itself off with "the chosen output is the TX audio", even when you picked the headphones by their card. Linux reports each USB card's make and model, and when the two are different models they are now two cards, so the monitor plays. Two of the same model (two of one radio, or two of one interface) are still one card, and so is a card whose model Linux does not report. The monitor is still refused on the radio's own TX card by any of its names, and on PipeWire or PulseAudio a TX audio device of pipewire, pulse, default or the system default still cannot be told apart from your headphones: set the radio's TX audio to its card by name in Settings ▸ Radio ▸ Audio, and pick the headphones by their card. The same rule decides whether a headset microphone is the radio's receive input for Use this computer's microphone. Needs a check on a Linux station with both cards plugged in.

  • Use this computer's microphone never keys the radio when its voice has no way out.

    Pressing PTT keyed the radio on the microphone's first sound whether or not the radio's transmit audio output (Settings ▸ Radio ▸ Audio ▸ TX audio device) had opened. If it had not, for example because the radio's USB audio was off or another program held it, the radio went on the air with no audio and nothing said why. Now the output is opened when you press PTT, before the microphone, as FT8 opens it when you arm TX; if it will not open, the radio is not keyed and the status line says so (COMPUTER MIC: the radio's transmit audio output would not open…). The diagnostic log also records each press: the microphone, its gain and the TX audio level when it arms, when it keys, and, when it closes, how much of your voice went to the radio and how loud it was at its loudest, so a "PTT keys but nothing goes out" report shows whether any voice reached the sound card.

  • Use this computer's microphone says why a held PTT sent nothing.

    When you held PTT in Phone for a second or more and the microphone sent only silence, or nothing at all, the radio was never keyed and nothing said why. The status line now says so (COMPUTER MIC: no sound came from the microphone…) and names the usual causes: the microphone is muted, or on Windows, Settings ▸ Privacy & security ▸ Microphone does not let desktop apps use it. A quick tap still says nothing. Nothing about when the radio keys or unkeys has changed.

  • Use this computer's microphone says what the radio must be set to.

    PTT keys the radio over CAT in USB or LSB and plays your voice into its USB audio input. On radios like the Yaesu FTDX10, a radio in USB or LSB transmits the input its SSB MOD SOURCE menu names, MIC from the factory, so with the menu on MIC the radio keys and sends its own microphone, not the computer's. Settings ▸ Radio ▸ Audio now says beside the switch that the menu must be on REAR (USB) for the computer's microphone to go out, and that Nexus does not change the radio's menus.

  • FT8 and FT4: a station shown in brackets is the same station as without them.

    A station sent hashed shows as <K1ABC>, and Nexus treated that as a different call from K1ABC: its card listed no previous contacts and its decodes and roster row had no B4 tag. The brackets are now ignored when Nexus checks a station against your log (previous contacts, the B4 tag, the contest and POTA dupe badges, the station card and its callbook lookup), as in WSJT-X, and the decode still shows the call as it was sent. A bracketed compound call such as <PJ4/K1ABC> is matched as PJ4/K1ABC, so your contacts with plain K1ABC do not count for it. An unresolved <...> matches nothing and opens no card.

  • Test CAT's baud-rate search no longer keys an interface that keys on RTS.

    When Test CAT fails on a serial radio, Nexus tries the other baud rates. Opening the port at each rate raises RTS, and on an interface that keys on RTS ("Interface keys RTS on the CAT port", a CW keyline on the CAT port's RTS, or a keying cable Nexus recognises) that keyed the transmitter at every rate tried, even with PTT set to CAT. Each rate now holds RTS down. Where it cannot, or Nexus cannot confirm that it can, Nexus skips the search and says why: the radio's own serial handshake would hold RTS up, Nexus could not confirm that handshake leaves RTS free, or the port keys on DTR too.

  • Serial handshake = Hardware now works on radios such as the Kenwood TS-480SAT (#432).

    With Settings ▸ Radio ▸ Advanced ▸ Serial handshake set to Hardware (RTS/CTS), Nexus still told the radio control program to hold the RTS line low, a combination Hamlib refuses, so CAT connected and did nothing. With Hardware chosen, Nexus now leaves RTS to the flow control and holds only DTR. Where RTS keys a transmitter (PTT on RTS on the CAT port, "Interface keys RTS on the CAT port", the CW serial keyline on the CAT port's RTS, or a keying cable Nexus recognises), hardware flow control would hold it keyed from the moment CAT connects, so Nexus does not start CAT with that pair, on the radio you operate or on another radio it is watching (including when one radio's RTS PTT port is the other's CAT port, however the port name is spelled; a watched radio's switcher pill gives the reason): the CAT status and Test CAT say "RTS keys your transmitter on this port, so Serial handshake can't be Hardware — set it to Auto or None." Every other handshake choice, and Auto, behaves as before.

  • Two Nexus windows make one cluster login, not two (#431).

    With two radios in two windows, each window logged into every DX cluster in your list, and into the Reverse Beacon Network, under the same callsign. A cluster that allows one login per callsign then refused or kicked one of them over and over, and kept it up long enough to risk a ban, while the other window looked fine. Now the first window to start logs in, exactly as a single window does, and the other window gets the same spots through it: they show on its boards as usual, and a spot you post from it goes out once, through the first window's login. The other window's Cluster and Phone lights say "other window", and they show connected only while the first window really is; its hover text, and Settings ▸ Spot Sources there, say the first window's cluster list and SSID are the ones in use. Close the first window and the other one takes over the logins within a few seconds, with one login per cluster again. Windows set to different callsigns still log in separately. You no longer need a different Cluster login SSID in each window. If the first window closes while the file the two windows share cannot be opened, the other window logs in for itself within about 15 seconds instead of waiting until it is restarted. A window whose radio picker could not be shown, and whose radio was open in another window, now logs in to the cluster once you turn Simultaneous radios off and switch it to a free radio.

  • Two radios in two windows: the radio picker no longer takes a radio's CAT with it (#431).

    While the "which radio?" picker was on screen, that window was already connecting to radios: it started a rigctld for every radio but the first. The radio window you then picked could find that rigctld already running, connect through it, and lose CAT a moment later when the picker closed and took the rigctld with it. A few minutes later Nexus reconnected the radio on its own, stopping a transmission on the way, and the log called the picker closing a radio engine that "stopped unexpectedly". The picker window now connects to no radio: the window you pick starts its own rigctld, and a radio window that is already open is left alone. "Use one radio" starts the radio as before, and a single window is unchanged.

  • Two radios in two windows: the second window no longer restarts its radio's rigctld as it opens (#431).

    The first window has the port Share this radio with other programs asks for, so the second window shares its radio on the next free port, and it says so. That move came a moment after the radio connected, and Nexus took it for a change to the radio: it dropped the CAT link, stopped the radio's rigctld and started it again, which the log showed as "CAT daemon rebuild" just after launch. The same could happen in a single window when another program already held the sharing port. Sharing moving to another port now leaves the radio connected. A radio whose rigctld port or Network Address is the port sharing moves onto or off still reconnects, as before.

  • The radio picker answers itself after 30 seconds.

    Because the picker window no longer runs a radio while it waits, a launch nobody is at the computer for (Nexus starting at sign-in, a station worked over Remote) would have waited for a click. If nobody answers within 30 seconds, the picker now repeats your last answer: it opens each radio in its own window again, or runs this window on one radio. The first time, with no answer to repeat, it runs this window on the active radio, as Nexus did before the picker. The countdown is shown, and any click or key press stops it. If the picker cannot be shown at all, the window starts its radio rather than sitting without one.

  • One radio, one window.

    A radio another Nexus window is running is now refused, with "That radio is open in another Nexus window — switch to that window.", whether you switch to it (here or from Remote), a band change would route to it, a separately named profile opens on it, you choose "Use one radio" for it, or you change your radio's port in Settings to the port it is on (the save is refused and your radio stays as it was). Before, two windows on one radio shared its rigctld, and closing one cut the other's CAT. A radio is known by the port it is on (its COM port, or its network address), so two profiles naming the same COM port count as one radio and two different ports never conflict. A window that only watches the other radios no longer watches, or starts a rigctld for, a radio another window runs, and it lets go of one as soon as that radio's window opens. A window that crashed does not keep its radio locked.

  • Field Day: a reply to your old call still completes the contact after you add /P.

    If you changed your callsign mid-contact to a portable form of the same call (K1ABC to K1ABC/P, or back), the other station's rogered exchange or RR73 addressed to the call you had before was ignored, so the contact never completed and was not logged, although the other station had logged you. As in WSJT-X, Field Day now takes the reply from the station you are working when it has your base call, whether it is addressed to your old call, your new one or the bare base call: the contact is logged once, in the entry's log as before, and your closing over goes out under the new call. A reply from anyone else, or to a different call, is still ignored, and after a change to a different base call a reply to the old one is still not taken. A new caller answering your CQ is still worked only when they call you by your exact call.

  • LoTW, signing from the location in the ADIF: a contact goes with its own grid and place.

    With "sign from ADIF location" on, a contact whose grid Nexus filled in (a portable contact, say) was uploaded with whatever grid your station had at upload time, and the state, county, CQ and ITU zones and IOTA Nexus filled in were left out. It now goes with the grid, state, county, zones and IOTA stored on the contact; only a contact with no grid of its own takes your current one, and a contact with none of the others goes without them, as before.

  • Settings ▸ Station: the State hint no longer runs into the Units label.

    On narrower windows (a 1080p screen at 150% scaling, for example) the end of the State hint touched the UNITS heading below it; the Units row now has the same space above it as the rows of fields above it.

  • Every band the Phone and CW pickers list now tunes.

    Some picks did nothing at all: 30 m in Phone and 9 cm with "Outside the US", and every receive-only band for a US class, such as 4 m for a General. They now tune. With "Outside the US", 30 m Phone parks at 10.120 MHz USB (some licences outside the US allow phone there; the band-plan chip notes that the IARU plans keep 30 m for CW and data), and 9 cm at 3400 MHz. A receive-only band takes you where the band's usual activity is, for example a Technician picking 20 m in Phone hears the phone segment, and transmit stays locked there by your licence, as the 🔒 chip says.

  • 60 m below 5.3 MHz, 630 m, 2200 m, 8 m and 5 m are ham bands now.

    Tuning the radio there cut a transmission in progress and blanked the band name, as if you had left the ham bands: on the UK's own 60 m (from 5.2585 MHz), on 630 m and 2200 m, and on 8 m (40–45 MHz) and 5 m (54–69.9 MHz). The band edges now follow the ADIF standard, so these are named bands and an over there is not cut. The band name on screen follows the same edges, a 60 m spot below 5.3 MHz counts as 60 m, and a contact imported with ADIF's own name for 2200 m, "2190m", is not flagged by the log check as being on the wrong band. Tuning off every band still cuts transmit, as before. A US licence class gains nothing from this: it still transmits only where its privileges allow.

  • US General, Advanced and Extra can transmit on 2200 m and 630 m.

    135.7–137.8 kHz and 472–479 kHz, every mode, per 47 CFR 97.301. Nexus showed 🔒 TX locked there, on legal spectrum. A Technician or Novice stays locked. The power limits (1 W EIRP on 2200 m, 5 W EIRP on 630 m) and notifying the Utilities Technology Council before you first transmit are yours to handle; Nexus does not check them.

  • Outside the US, an outside program's LSB or AM is no longer refused on a US 60 m channel frequency.

    With "Outside the US" as your licence, a program keying through Nexus (WSJT-X, fldigi) on 5.3305, 5.3465, 5.3715 or 5.4035 MHz was refused unless the radio was in USB, a US rule that does not bind you (the UK's 60 m blocks include two of those frequencies). US classes keep the rule.

  • Split Operation = Rig: Tune, CW, PTT, an APRS beacon and the ATU tune-up go out on your dial between FT overs.

    Since 1.22.0 Nexus keeps the radio in split between FT8, FT4 and FT2 overs, as WSJT-X does, with the radio's TX VFO left on the last over's shifted frequency (your dial moved by a multiple of 500 Hz). A Tune, a CW ID or other CW, your own or a Remote's PTT, an APRS beacon or the radio's ATU tune-up from the FT screen then went out of that VFO, somewhere the licence check had not judged: near the edge of your privileges, outside them. Before any of these keys Nexus now puts the TX VFO back on your dial (or lets go of split if it was about to anyway), so they go out where they would with no split at all; the next FT over moves it again as usual. If the radio does not confirm the move within half a second, nothing keys: the tune ends, the CW or the beacon is dropped, a PTT press does nothing until you let go and press again, and the status line says why. A Tune pressed while an FT over is still on the air is now refused (below).

  • A pane divider no longer sits on a macro key.

    On CW at a large zoom (1920×1080 at 150 %), when the screen scrolled, the divider between the decode and log columns stood over the middle of F7, so a click there grabbed the divider instead of sending the macro. The macro keys and the Stop TX row now stay on top of the pane dividers in every mode that has them.

  • A garbled NOAA timestamp no longer blanks the Coming card; space-weather, DXpedition and logbook downloads are size-limited.

    If NOAA's 3-day forecast carried a damaged "Issued" time, the whole Coming card stayed empty (the watch, the model and the solar wind with it) for as long as NOAA served it. Now only the forecasters' note is left out. The NOAA space-weather downloads (solar flux, Kp, X-rays, flares, absorption map, alerts, scales, the Coming products, the aurora oval), the DXpedition calendar and Club Log's lists, and the Club Log and eQSL logbook answers are now read only up to a limit well above their real size, over HTTPS only and without following a redirect; a larger or redirected answer is treated like any failed download.

  • Every other download is size-limited too.

    The rest of Nexus's downloads are now read only up to a limit well above their real size: the solar wind, proton, solar-cycle and ionosonde (KC2G) readings, the WSPR and PSK Reporter reports, the contest calendar, POTA and SOTA spots, the POTA parks list, the satellite lists (SatNOGS, Celestrak and the Nexus mirror), the repeater directories (RepeaterBook, hearham), RSGB's locator data, place search, the LoTW user list and LoTW confirmations, Cloudlog/Wavelog and WRL answers, the hosted country (cty.dat), FCC-state and Field Day rule files, and the update check. A larger answer is refused with a message, like any failed download, instead of being read to the end. Normal answers read as before.

  • The docs said the CAT broker is off by default.

    It has been on since 1.1.0; the interop guide's ports table and the Integrations page now say so.

  • Every satellite LoTW accepts can be picked in the Log (#435).

    The satellite picker on a logged contact offered only about two dozen birds, so RS-44, AO-7, FO-29, IO-86, QO-100, the CAS and XW birds and many more could not be chosen. It now offers LoTW's whole published list of 110. A contact made through a satellite pass on AO-7, FO-29, AO-73, RS-44, QO-100, IO-86, IO-117, AO-85, CAS-3H, MO-122, UKUBE1 or ESEO is now tagged with that satellite automatically, as SO-50 and AO-91 already were.

  • "New grid" on the roster only for a grid the station sent you (#327).

    A station your radio was decoding could wear the New grid tag off PSK Reporter's locator for it, even when it never sent a grid on the air, as DXpeditions often don't. The contact then logged without that grid, since only a grid the station sends counts toward grid awards. For a station you are decoding, the roster and the Needed board now judge New grid by the grid a contact with it would be logged with: the grid it sent you this session (on any band, JTTY included), else the one you logged for it before. With neither, they show none. A station you are not decoding still shows PSK Reporter's grid on the Needed board, as a lead.

  • The Flex native switches no longer vanish on a serial connection (#242).

    With a FlexRadio model picked and Connection set to Serial, Settings ▸ Radio ▸ Rig & CAT hid Flex native panadapter, Flex native DAX audio and Flex native client without a word. Flex native client and Flex native DAX audio now show and work on Serial too, since the client talks to the radio at its own address (the Flex radio IP). So does the client's "Let Nexus send CW, Tune and ATU" setting, shown whenever the client is on and the Flex radio IP is set. Flex native panadapter, which needs the network connection, shows greyed out with "Needs Connection: Network." On OmniRig or TCI the client never starts, so its three switches say "The Flex native client doesn't run on OmniRig or TCI." and can't be turned on there. A switch left on can always be turned off. Every switch that can't be used right now is now greyed out as well, including the ones marked "Set at the station" when you open Settings from a Remote, which used to look live.

  • NET rigctl to a rigctld on another computer: Tune and PTT key the radio on time, and a key Nexus has given up on is taken straight back off (#429).

    With Rig Model NET rigctl and a Network Address on another computer (a rigctld, or a rigctld-compatible bridge, on your network), Nexus started a rigctld of its own in front of it. Behind that extra rigctld a mode change could hold up every command after it for many seconds, so Tune and PTT said the radio had not keyed, and then the radio keyed tens of seconds later, sometimes after you had let go. Nexus now talks to that address directly, as it already did for one on this computer, and starts nothing in between. And if a Tune key Nexus reported as not keyed reaches the radio later after all, Nexus unkeys the radio again at once and the status line says the radio keyed late. For ten seconds after such a key, or after a connection to rigctld broke while a key was still unanswered, Nexus also watches the radio's own PTT, and if the radio says it is transmitting while Nexus is not, Nexus unkeys it and says so (a press of the radio's own microphone PTT inside those ten seconds is unkeyed too).

  • Tune is refused where its tone would be outside your privileges.

    Nexus's Tune plays a steady 1500 Hz tone, so on the FT screens its carrier goes out 1.5 kHz above your dial whatever your TX offset, while the licence check judged the FT signal at your TX offset. A General at 14.0230 MHz with the TX offset at 2000 Hz could tune at 14.0245, below the 14.025 edge. Tune now keys nothing there, and the note beside it says why. The same holds on the RTTY and Keyboard screens, where the tone sits on its own side of the dial. On a TCI connection, and while a Flex tunes with its own carrier, nothing changes. With Split Operation set to Rig or Fake It, Tune pressed while an FT over is on the air is refused too, rather than keyed on top of that over from the frequency split moved it to, and the note beside Tune says so; with Split Operation None it still takes over, as in WSJT-X.

  • Split Operation = Rig: every other transmission waits for split too.

    RTTY, PSK, JTTY, the voice keyer, SSTV and the Remote and computer microphones now get what Tune, CW and PTT got: on the first moment after you leave the FT screens, split is let go before they key. While the radio has not yet accepted turning split off, or with Fake It has not yet taken its dial back after an over, nothing keys at all (Tune, PTT, CW or any of these) until it does, and the status line says why.

  • The ATU tune-up, an APRS beacon and the voice keyer are judged where they actually transmit.

    From the FT, RTTY or Keyboard screens the radio's own ATU carrier and an APRS packet's FM carrier go out at your dial, which the licence check did not look at: a General on the FT screen at 14.024 MHz could run the ATU below the 14.025 edge. And a voice-keyer message sent from a screen other than Phone was judged as that screen's narrow signal, not as voice: a General at 14.074 MHz could send one into the data segment. Each is now refused there: the ATU in its own words, naming the radio's carrier at your dial and shown beside the ATU (and, as the radio's own carrier on a Flex already was, judged as CW too, so a 60 m channel's phone dial refuses it), the voice keyer saying it is phone your licence does not allow, and APRS with its usual TX-locked message.

  • A radio's port name with non-English letters no longer stops Nexus from starting.

    A serial port name such as a Cyrillic СОМ4, from a restored backup or an edited settings file, crashed Nexus at every launch, before its window opened. Nexus now starts, and treats that name as its own port, not as COM4, so the radio will not connect until you pick its real port in Settings.

  • The FT-710 now shows the "CAT CW keying is unproven" notice.

    With the CAT keyer on this radio, Nexus can show sending with nothing on the air: Hamlib reports success either way, and the radio's manual does not list the command Hamlib uses to start sending. Nobody has settled this on a bench yet, so it is a notice only. The CAT keyer stays selectable and keys exactly as before. If CW doesn't go out, use the Serial keyline, WinKeyer, or Soundcard keyer.

Every version