Limiting channel bandwidth

RTLSDR-Airband uses FFT channelizer to extract individual narrowband channels from the wideband input signal. The bandwidth of a channel is determined by the sampling rate configured on the device, the FFT size and the windowing function. One may think that the bandwidth is equal to the width of a single FFT bin, but in fact it is at least a couple of times larger than that due to spectral leakage of the windowing function (Blackmann-Harris). If there are other signals in the neighboring bins, they will bleed into the bin of interest, causing interference and/or aliasing. This is usually not a problem in VHF airband, because air traffic services avoid using closely spaced frequencies in the same area. There are however other communications systems (eg. public services), where strong transmissions are close to each other and this is where the default channel bandwidth might be too large and cause problems.

bandwidth channel option enables lowpass filter that operates on the channelized I/Q signal (before demodulation). The result is twofold:

• it rejects interfering signals from neighboring FFT bins
• it reduces the audio bandwidth and lowers the high-frequency noise (hiss).

The syntax for multichannel and for scan mode is:

devices: ({
  channels: (
    {
      freq = 145.35;
      bandwidth = 8000;
      (...)
    }
  );
});

In scan mode there is also second variant:

devices: ({
  channels: (
    {
      freqs = ( 145.35, 145.65, 145.85 );
      bandwidth = ( 8000, 10000, 7000 );
      (...)
    }
  );
});

bandwidth value adheres to the syntax of a frequency. Setting the value to 0 disables the filter for a particular frequency.

Note: the filter is centered at 0 Hz and spans from -(f/2) to f/2. So if you want the audio to be lowpass-filtered up to 4 kHz, you have to specify bandwidth of 8 kHz.

Note: file and icecast outputs have their own lowpass and highpass filters which are enabled by default and perform audio filtering at the MP3 compression stage. If bandwidth is enabled, the signal is effectively filtered twice, which might cause overly high attenuation of higher tones. Consider disabling output filters in this case. See Audio filters in MP3 outputs for details.