Before this:What is software-defined radio?Digital modulation & constellationsWhat is trunked radio?
From antenna to audio
Key takeaways This is the capstone: one trunked call traced end to end through GopherTrunk. The antenna feeds an SDR that digitises spectrum into IQ samples. Software tunes and filters to the control channel, demodulates its 4FSK into symbols, and decodes the trunking protocol. When a call starts, GopherTrunk tunes a receiver to the assigned voice channel, recovers its symbols, feeds the vocoder frames to a speech decoder, and writes an audio file. Every concept from the earlier lessons is one link in this chain — and troubleshooting is just finding which link broke.
You’ve met every piece on its own. Now let’s assemble them. We’ll follow a single
police dispatch call from radio waves in the air to a .wav on your disk, naming
the lesson behind each step so you can see how the whole path fits together.
Stage 1 — Antenna: catching the wave
It starts with a radio wave arriving at your antenna, which converts the wave’s electric field back into a tiny electrical signal — typically a faint −80 to −110 dBm. Antenna choice and placement set the ceiling for everything downstream: if the signal doesn’t arrive with enough strength above the noise floor, no amount of clever software can recover it. This is why “where you put the antenna matters more than the radio.”
Stage 2 — SDR hardware: spectrum becomes IQ
The signal travels down the coax into your SDR. There it’s amplified, shifted down in frequency, and handed to an analog-to-digital converter that samples it millions of times a second. The output is a stream of IQ samples — pairs of numbers capturing the amplitude and phase of a whole slice of spectrum at once.
Two settings rule this stage. Sample rate sets how much spectrum you capture in one go (and how hard the CPU works). Gain sets how strong the signal is when it hits the ADC — too low and weak signals vanish into the noise; too high and the ADC clips at 0 dBFS, spraying distortion everywhere. Getting these right is most of “setting up” an SDR.
Stage 3 — Tune and filter: zoom in on the control channel
Now we’re in software. The IQ firehose covers far more spectrum than one channel, so GopherTrunk digitally tunes to the control channel and filters away everything else, keeping just that narrow slice. This is the SDR equivalent of turning the dial — except it happens in code and can be done for several channels in parallel from the same sample stream.
Stage 4 — Demodulate: recover the symbols
The isolated control-channel signal is still a modulated waveform. GopherTrunk demodulates it — for P25 and DMR, reading the 4FSK frequency shifts — and runs clock recovery to sample each symbol at exactly the right instant. This is the stage you can watch: the constellation, eye diagram, and symbol scope all visualise these recovered symbols, so a smeared constellation here tells you the demodulator is struggling before any data is lost.
Stage 5 — Decode the trunking protocol: read the map
The symbols become bits, and the bits become control-channel messages. Decoding the trunking protocol is where GopherTrunk reads the system’s running commentary: registrations, requests, and — crucially — channel assignments. The moment it sees “talkgroup 101 → voice channel 3,” it knows a call is starting and exactly where. You can watch this raw stream in the CC Activity panel.
Stage 6 — Follow the call: capture the voice channel
GopherTrunk now tasks a receiver to tune to voice channel 3 and runs stages 3–5 again on that channel — filter, demodulate, recover symbols, decode. Because it’s still reading the control channel in parallel, it can follow this call and catch the next assignment elsewhere, capturing multiple simultaneous calls across the system. It’s also reading the transmitting radio ID, so the recording is labelled with who spoke, not just which talkgroup.
Stage 7 — Vocoder: frames become sound
The voice channel doesn’t carry audio — it carries vocoder frames, speech compressed to a few kilobits per second by a codec like IMBE (P25) or AMBE+2 (DMR). GopherTrunk feeds those frames into a matching speech decoder, which reconstructs an audible waveform. This is the step that turns abstract bits back into a human voice — the analog-vs-digital-voice trade-off in action.
Stage 8 — Audio out: playback and a WAV on disk
Finally, the reconstructed audio is played live in the console and written to a
.wav file, tagged with the talkgroup, radio ID, timestamp, and system. From
here GopherTrunk can also stream it onward to services like Broadcastify or
RdioScanner. The call that began as a faint ripple at your antenna is now a labelled
recording you can replay — and the engine is already following the next one.
Why this map is your best troubleshooting tool
When a call doesn’t decode, the chain tells you where to look. Walk it in order:
| Symptom | Likely stage | Where to look |
|---|---|---|
| Nothing at all, flat spectrum | Antenna / gain | signal below noise floor, or too little gain |
| Spectrum looks “smeared” or distorted | SDR / gain | ADC clipping — reduce gain |
| Control channel won’t lock | Demodulate | smeared constellation, clock/timing |
| Locks but no calls | Decode protocol | wrong system parameters or protocol |
| Calls listed but silent | Vocoder / encryption | encrypted talkgroup, or wrong vocoder |
That’s the real payoff of this whole learning path: you’re no longer poking settings at random. You have a mental model of every stage, so you can reason about which one broke — exactly what the calibration and troubleshooting lesson builds on.
Quick check: a system is locked and calls are listed, but they play back silent. Which stage is the prime suspect?
Recap
- A call flows: antenna → SDR → IQ → tune/filter → demodulate → decode protocol → follow voice → vocoder → audio.
- The control channel is decoded continuously so the engine knows where every call goes and can follow several at once.
- Each stage maps to an earlier lesson — this is where the whole path comes together.
- The pipeline doubles as a troubleshooting checklist: find the stage that broke.
You’ve now seen the entire chain. The remaining lessons in this module make you fluent at operating it — finding systems, tuning a clean lock, and calibrating — and the glossary is there whenever a term needs a refresher.
Frequently asked questions
How does GopherTrunk turn a radio signal into an audio recording?
GopherTrunk runs a pipeline: the antenna feeds an SDR that digitises spectrum into IQ samples; software filters and tunes to the control channel, demodulates its 4FSK into symbols, and decodes the trunking protocol; when a call starts, it tunes a receiver to the assigned voice channel, demodulates and decodes that, runs the vocoder frames through a speech decoder to produce audio, and writes the result to a WAV file with metadata.
What is the role of the vocoder in the signal path?
Digital voice systems don’t send raw audio — they compress speech into tiny data frames using a vocoder such as IMBE (P25) or AMBE+2 (DMR). At the end of the receive chain, GopherTrunk feeds those decoded frames into a matching vocoder to reconstruct an audible waveform you can listen to and save.
Can GopherTrunk follow several calls at once?
Yes. By continuously decoding the control channel, GopherTrunk knows every voice-channel assignment as it happens and can task receivers to capture multiple simultaneous calls across the system, then mix or queue them for playback and recording according to your priorities.
Where can a call be lost in the signal path?
A call can fail at any stage — too little signal above the noise floor, gain set so high the ADC clips, a tuning or clock error smearing the constellation, the wrong protocol or parameters, or encryption you can’t decode. Most troubleshooting is about finding which stage in the chain broke.