How to avoid audio clashes GNU Radio with other devices using the audio stream on the computer running the software
I was using a screen recorder (Debut) to record a Microsoft Teams meeting tonight. However second half lost its sound, which occurred after I ran a GNU Radio Companion script which displayed a waterfall, played sound from SDR and displayed spectrum. What happened and how can I prevent it next time?
What you describe is very typical when multiple applications grab the same audio device or change the system audio graph mid-session. GNU Radio Companion, SDR drivers, and Windows audio routing can interact in ways that break the audio capture path Debut was using.
Here’s the most likely chain of events and how to prevent it next time:
✅ What probably happened
1. GNU Radio grabbed or reconfigured your default audio device
When you ran your GRC script (waterfall + audio sink + spectrum):
- The Audio Sink in GNU Radio will usually open the default Windows output device.
- Opening a device in exclusive mode (not uncommon with SDR audio or certain sound APIs) can cause:
- The device to switch sample rates
- The device to restart
- Other software (Debut, Teams) to lose their audio routing
This often results in recording software losing its audio input/output reference, so Debut stopped capturing system audio.
2. Windows may have changed the default audio device
SDR software sometimes creates or exposes “virtual” devices (e.g., from VAC, VB-Audio, cable input, or similar).
Starting GRC may have caused:
- Default playback device to switch
- Default recording (“Loopback”) device to switch
- Debut to continue recording the old device, which now had no sound
3. Teams sometimes switches audio channels when the system changes
Teams is sensitive to device changes. If the output device resets while a meeting is in progress, Teams may shift output to another device silently.
Debut will then record silence.
🧪 How to confirm this happened
On Windows, check:
- Settings → System → Sound → Output → Device preferences
- Advanced sound options (App volume and device preferences)
Look for:
- Teams output device changed part-way through the session
- Debut recording device still pointing to the old one
- Sample rate mismatch (e.g., 48 kHz vs 44.1 kHz)
🛠️ How to prevent this next time
✔️ 1. Lock your audio devices before recording
- Open Windows Sound Settings
- Set a fixed Default Output and Default Input device
- Disable “Allow applications to take exclusive control” for the default device
(Sound Control Panel → Playback Device → Properties → Advanced)
This prevents GRC from kicking Debut off the device.
✔️ 2. In Debut, explicitly choose the audio source
Instead of “Use default device,” choose:
Audio → Speakers (your actual device name)
Debut will then stick to that device even if the system changes the default.
✔️ 3. In GNU Radio: specify the audio device explicitly
In your Audio Sink block, set:
- Device name (e.g.
hw:0or Windows WASAPI device identifier) - Sample rate equal to your system’s default output (usually 48 kHz)
This prevents GRC from opening your default device and reconfiguring it.
✔️ 4. If using virtual audio cables → isolate them
If you use VAC/VB-Cable for SDR:
- Make sure Debut captures the actual speaker output, not the virtual device
- Teams should always output to your real speakers/headphones
This keeps SDR audio experiments separate from meeting audio.
✔️ 5. Record Teams using “Audio Loopback” rather than system audio
Most recorders (including Debut) allow choosing:
- Loopback of communications device
- Loopback of specific app (Teams)
This avoids conflicts entirely.
💡 Practical recommendation
If you’re running GNU Radio during a meeting recording:
The safest approach:
Use a separate audio device for GNU Radio
(e.g., USB sound card or virtual device).
Set GNU Radio to play to:
Speakers (USB Audio Device)Set Teams + Debut to:
Speakers (Realtek / main device)GRC can then do whatever it wants without breaking the Teams audio that Debut is recording.