Anti-aliasing filter design

Can you use the lowpass function as an anti-aliasing filter, or is there some preferrable way to do it such as using the butter function?

Answer 1

[Star Strider]

Yes, generally. The lowpass function designs a FIR filter by default, and will design an efficient elliptic IIR filter if the 'ImpulseResponse','iir' name-value pair is included in the argument list.

We had the anti-aliasing discussion before, so I will not repeat it here.

[L'O.G.]

Thank you!

[Star Strider]

As always, my pleasure!

Answer 2

[Jon]

For the purposes of preventing aliasing the only thing that is essential is that the resulting frequency content that is above the Nyquist frequency is neglibibly small. So any filter with sufficient attenuation above Nyquist frequency will work.

The choice of which filter to use then depends upon the application. There are various tradeoffs, which must be considered, and different filter types are better for particular applications. For example, If the application is for a real time signal, particularly within a feedback loop then the delay will be quite important. If the application is for an offline application where the signal can be post processed, then the delay can be removed.

I woud suggest Googling choice of antialiasing filters to get some further ideas and details, for example: https://www.digikey.com/en/articles/the-basics-of-anti-aliasing-low-pass-filters#:~:text=Selecting%20a%20filter%20type&text=The%20most%20commonly%20used%20are,the%20application%20(Figure%203).&text=The%20three%20filter%20responses%20shown%20have%20specific%20characteristics.