-
Notifications
You must be signed in to change notification settings - Fork 479
Help! I'm getting audio underruns or overruns!
There may be multiple causes of this:
OpenWebRX typically requires 200 kbit/s uplink per client. If you run it from a home Internet connection, and multiple clients are connected, you may run out of the uplink bandwidth your ISP allows.
Possible solutions:
- Reduce the FFT size (
fft_size
inconfig_webrx.py
), by which you can decrease the bandwidth usage to about 100 kbit/s per client. - Reduce the maximum allowed number of clients to a safe value, at which no lags occur, by the
max_clients
setting inconfig_webrx.py
.
I mean, you set a quite high sample rate (multiple Msps), and the CPU cannot handle that. Sometimes it only happens when multiple clients connect simultaneously: when the N-th connection is made, the audio starts to lag for everyone. You can also observe around 100% CPU usage at the status bar.
You can try several things:
- Reduce the maximum allowed number of clients to a safe value, at which no lags occur, by the
max_clients
setting inconfig_webrx.py
. - Reduce the sampling rate of the SDR hardware (
samp_rate
inconfig_webrx.py
). - You can also reduce the FFT size (
fft_size
), although it has less impact on CPU usage as it is calculated only once.
For some benchmarks please see this page.
See also Optimizing OpenWebRX for speed.
If nothing helps, you can try to increase the client_audio_buffer_size
in config_webrx.py
.
You should try another value for samp_rate
.
If nothing helps, you can try to alter the client_audio_buffer_size
in config_webrx.py
.