I was recently looking for appropriate filters for filtering gigabytes of strain and acceleration data within a specific frequency ranges. I came across the idealfilter function, which appeared to be quite fast and performed exceptionally well in comparison to high order (+20) IIR and FIR filters.
Thus, I have come to wonder what are the downsides of using the idealfilter in comparison to common IIR filters (cheby2, butter) for example. I could notice that the upsides are speed and good attenuation. Is idealfilter only performing well with 1D vector applications?
I have attached a code, where I apply some filtering to a Tohoku earthquake recording (in .mat format) to play around with, if needed.
Would be grateful if you could give some advise and share your experience.