STD
Shift, Tone, Decode. Software in testing. Version 0.1.10. Around the shack it also answers to Sending Transmittal Disease.
The real expansion is Shift, Tone, Decode. The other one is the joke. STD is a 64-bit Windows application, a Qt 6.11.1 MSVC digimode console for a TCI radio, currently in the contagious stage: it keys on the bench, it has not finished making the rounds, and the recommended treatment is another shot of bourbon while the modes get proved on the air. Lyra is the transceiver. SDRLogger+ is the logbook. CW stays in Lyra.
STD: Help keep the infections spreading and buy us a shot of bourbon.
☕ Donate via PayPal
Copyright (C) 2026 Rick Langford (N8SDR).
License
GNU General Public License, version 3. The license text is COPYING.
Original files in this repository are GPL-3.0-or-later (SPDX-License-Identifier: GPL-3.0-or-later). Modem math rewritten from Fldigi (Dave Freese W1HKJ and the Fldigi authors, GPL-3) and the FT8, FT4, and FT2 path from MSHV are GPL-3. Speech to text uses whisper.cpp (the ggml authors, MIT), built in; its notice is in resources/licenses/whisper.cpp-LICENSE. CW stays in Lyra.
Screenshots
Modes that key
TX sends one of these when the operator presses TX or Tune. Opening the radio does not key. The text command START is never sent. On Hell, RTTY, PSK, QPSK, Olivia, Contestia, MFSK, DominoEX, Thor, and MT63, a finished pass takes the sent text off the front of the transmit box, as Fldigi does, so the next TX sends only new text. A stopped pass takes the characters that already went out. Text changed during the pass is left alone. JS8 keeps its own queue. RTTY, PSK, and QPSK send live, as Fldigi does: TX keys and sends what is in the box, text typed while on the air goes out as it is typed, and an empty box sends idle (LTRS in RTTY, phase reversals in PSK) without dropping TRX. Press TX once to end after the box is sent, and again to stop at once; characters not yet on the air go back into the box. A ▶ macro (one with <TX>) keys when Macros key is on and, with <RX>, returns to RX when its text is done. A macro pressed while on the air is added to the box. Back to RX after (on, 25 s by default; 5, 10, 15, or 25 s) ends a live transmission by itself when nothing new was typed for that long; turn it off to stay keyed until you press TX. Options → Speech… reads received text aloud with a Windows voice on a sound device you pick, not through the radio. It reads RTTY, PSK, QPSK, Olivia, Contestia, MFSK, DominoEX, Thor, and MT63 a line at a time, or a run of finished words on a long line, and waits quietly for new text. Callsigns and reports are spelled out, noise runs can be skipped, and it stops while you transmit. Hell is a picture and is not read; nor are JS8 or the FT8 family. With read-back on, a Speech chip beside CLR RX mutes and unmutes, and the number beside it is the volume. It is off until you turn it on. Talk does the reverse: it turns your voice into transmit text. In Options → Speech… → Transmit pick a microphone, set the gate, and download a Whisper model (75 to 490 MB, fetched only when you choose one, so STD itself stays small). It runs on the CPU. Hold Talk and speak. The words land at the end of the transmit box with callsigns and reports written the ham way (Kilo Eight Alpha Bravo becomes K8AB, five nine nine becomes 599), and nothing keys until you press TX. My words in the same tab lists what Talk should expect (club calls, names, towns, your rig and antenna); your name and QTH from Options → User and the station in the QSO bar are added by themselves. Corrections fixes words it keeps getting wrong, one per line as heard = send. To teach one on the spot, select the wrong text in the transmit box, right-click, and pick Talk: always send … as, or add it to My words. The waterfall picture updates once per turn, so a quiet band does not keep the CPU and the GPU busy.
以上为 README 的开头部分。其余内容请阅读原文。 查看完整 README ↗
