Implement more intelligent way of defining FIR filter length

[psurukuc]

The current FIR filter length (while generating from TF) is done by taking 2*(circular shift) + (TF length). This works alright for most of the filters, but if the filter is really resonant it fails. Implement a better way to do this by perhaps defining both the ends of the filter where the magnitude falls below a threshold (as a fraction of the maximum magnitude of the filter ?).

[arinabykadorova]

Yes, this would be better. Perhaps a patch-up longer FIR length is okay for now, but will put this on the list.

[pslocum]

@jkgaison65 has been able to adapt the antenna FIRs for resonant systems. So I think we can optimize the FIRs for antennas, merge to develop, and then revisit if needed when we merge the resonant calculations. In the meantime if there are additional near-term FIR changes being planned, please let us know.