Skip to content

ArchangeSDR 1.3.0

Choose a tag to compare

@bt0r bt0r released this 13 Nov 10:46
· 15 commits to main since this release

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.sh

Then restart ArchangeSDR. Settings, ranges and history are kept.