Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

AI Noise Reduction (RM Noise)

RM Noise is a free third-party Windows app that removes HF noise (QRN, static crashes, heterodynes) using AI models trained on real ham-radio recordings. POTACAT doesn’t embed it, but you can route your receive audio through it today — and everything downstream of POTACAT, including ECHOCAT listeners on your phone or remote desktop, hears the cleaned-up audio.

What you need

  1. RM Noise — download from ournetplace.com/rm-noise
  2. A virtual audio cable — e.g. VB-CABLE (free). Flex users already have DAX, which works as the input side.

Routing

Radio RX audio (USB CODEC / DAX RX) → RM Noise → VB-CABLE → POTACAT audio input
  1. In RM Noise: set its input to your rig’s audio device (e.g. USB Audio CODEC or DAX Audio RX 1) and its output to CABLE Input (VB-Audio).
  2. In POTACAT: wherever you pick an audio input (ECHOCAT audio bridge, the VFO pop-out’s monitor), choose CABLE Output (VB-Audio) instead of the rig’s device.
  3. Pick a noise model in RM Noise and listen — phone listeners over ECHOCAT get the same cleaned audio for free.

Important: keep digital decoding on RAW audio

AI noise reduction is for your ears, not for decoders. It destroys the signal structure FT8/FT4, SSTV, and CW decoding depend on.

  • In JTCAT, keep the audio source on the rig’s raw device (DAX Audio RX 1, USB Audio CODEC, or SmartSDR Direct) — not the VB-CABLE device.
  • Same for SSTV decoding.

Voice listening cleaned, digital decoding raw — best of both.

Notes

  • RM Noise is Windows-only, and some of its processing runs on its developer’s servers, so it needs an internet connection.
  • Expect a small added audio latency (the AI processes in blocks). Fine for ragchews and POTA hunting; irrelevant for digital since decoders bypass it.
  • Native noise reduction inside POTACAT (no extra apps) is on the wishlist — if this routing works well for you, say so in Discord; it helps prioritize.