Repository navigation
ArchangeSDR 1.3.0
Highlights
- Cleaner recordings to listen to. Each recording is now kept twice:
recordings/<id>_clean.wav, played in the log and sent by ntfy: the voice band only (300 - 3000 Hz) and, in NFM, the de-emphasis the transmitters expect. The FM hiss above 3 kHz and the AM rumble below 300 Hz are gone; about 4 dB less hiss in NFM on real recordings. The voice keeps its level, and no word is ever cut.recordings/<id>_raw.wav, the audio as received. The transcription uses it: Whisper did worse on the filtered audio.
- The two files take twice the space of one; the retention setting (Settings → Recordings) deletes both.
- Recordings made before this version (
<id>.wav) still play and transcribe as before. - The filter costs about 2 % of real time on a uConsole (Raspberry Pi CM4).
Update
cd ArchangeSDR
git pull
./install.shThen restart ArchangeSDR. Settings, ranges and history are kept.