Background

The Signal Separator uses FIR filters to extract the detected signals out of the input stream. To save CPU load, their taps can be calculated before runtime. Since the bandwidth of the signals is not known beforehand, several taps are created for a defined grid of normalized cutoff frequencies. The Signal Separator will choose the taps with the nearest cutoff frequency for each detected signal.

Usage

The python tool apps/taps_gen.py is able to create a JSON file with precalculated taps. Use it like this:

1. Choose the windowing for the taps according to the displayed numbers.
2. Choose the transition width of each filter. This is the frequency range between the passband and the stopband. Enter it normalized to the cutoff frequency. A good value is 0.1.
3. Enter the smallest cutoff frequency that should be present in the taps file normalized to the sampling rate.
4. Enter the highest normalized cutoff frequency that should be present in the taps file.
5. Enter the normalized cutoff frequency step between each filter.