SDRmm

Operate

Channels and audio

Tuning, auto vs manual Devices, squelch, IF filter, Audio FX, and virtual audio cables.

On this page

Wire Device iq into any mode channel from + Add (see Decoders or Modes). Click a channel to inspect its baseband spectrum, constellation, and eye diagram in the side panel. Right-click → Replace with… swaps mode while keeping frequency and squelch (m / M cycles NFM, WFM, AM, SSB).

Tuning

Dial digits accept scroll, arrow keys, or typed frequencies (MHz by default; add Hz, kHz, or GHz). You can also drag channel markers on Scope or use the keyboard.

Auto and manual

  • Auto (default): Device places its span to cover the most wired channels while avoiding its DC spike (2/3 shows coverage).
  • Manual: tuning the Device directly locks it in place; channels outside its span wait. Click the radar icon to return to Auto.
  • Several radios: wire one channel to multiple Devices; auto-tuned radios split channels to maximize coverage. Scanners, hunts, recordings, and IQ exports pin their channel to one radio until stopped.

Sample rate

SDRmm resamples Device IQ automatically. Matching native rate skips resampling (ADS-B 2.4 MS/s, DAB/GNSS 2.048 MS/s, ATV 16 MS/s).

Squelch

  • Off (far left), Manual (slider), or Auto (Margin 2–40 dB above measured noise floor; freezes while open so long calls are not clipped). Analog channels emit call events only with squelch on.
  • NFM Tone: Detect (display CTCSS/DCS), CTCSS, or DCS mute. Scrambler inverts voice (1.5–4.5 kHz or Auto); Compander applies 2:1 expansion.

IF filter and blanker

  • Blanker: removes short impulse spikes from IQ before filtering.
  • IF filter: shapes IQ before demodulation so adjacent stations cannot capture FM or swamp a decoder. Edges trim each side independently; Notches pin to RF frequencies and stay on interferers when you retune (shaded/red on Scope).

Audio

The Audio chain at the end of a decoder (or voice encoder) processes its audio output (FX lights in header). Stages run in order:

Stage Does
De-click Removes clicks
Passband / Notches Band-limits audio; cuts up to 4 tones
Peak Boosts one tone up to +24 dB (useful for CW)
Auto notch Tracks and cancels steady tones
Denoise Spectral (lightweight, up to 20 dB) or DPDFNet (8 kHz neural speech model, download on demand: 2 light, 8 stronger; turn off for music/CW/data)
AGC Slow / Med / Fast volume levelling (AM/SSB already level themselves)

Speaker output

Pick output device under Out (in browsers, press Show device names once). For external apps, route a Speaker to a virtual cable: Linux pactl load-module module-null-sink sink_name=js8 (js8.monitor), Windows VB-Cable, macOS BlackHole.

Edit this page on GitHub

Esc