1
$\begingroup$

I am trying to understand the basic categories of DSP filters, and it seems that the two main groups are FIR and IIR, however I'm not sure how to categorize high pass, low pass, band pass, and band stop (the 4 basic filters I learned about in university) and I can't find material on this either even though it seems so basic. Any help with making categories out of or distinctions with these groups would be greatly appreciated!

$\endgroup$
  • 2
    $\begingroup$ The two categories are not related. You can design a low pass both as FIR or IIR etc. $\endgroup$ – Hilmar Jul 14 at 19:29
  • 1
    $\begingroup$ @Hilmar I see. Can you explain that a little more? How could low pass filters be recursive? $\endgroup$ – tash7827 Jul 14 at 19:35
  • 1
    $\begingroup$ @tash7827 just like any other filter, it can be recursive. That's like asking "how can a fast car be red?". The color of the car doesn't dictate its speed, and the lowpassiness of a filter doesn't dictate its recursiveness. $\endgroup$ – Marcus Müller Jul 14 at 20:59
2
$\begingroup$

The expressions Finite Impulse Response (FIR) and Infinite Impulse Response (IIR) refer to the time-domain response of the filter. When an impulse is fed into the FIR filter in the time domain, it's response goes to zero within a finite amount of time. The response of the IIR filter does not.

Low Pass Filter (LPF), Band Pass Filter (BPF), High Pass Filter (HPF), Band Stop Filter (BSF) and Notch Filter (NF) are expressions that refer to the frequency-domain response of the filter - which frequencies it passes, stops or attenuates.

Both domains can be used to classify a filter's behavior. You can implement a LPF as a FIR filter. In the frequency domain, it will pass low frequencies and stop high frequencies. Simultaneously, in the time domain, the impulse response of the filter will return to zero after a finite amount of time. You can also implement a LPF as an IIR filter. It will have the same behavior in the frequency domain, but its impulse response will never reach zero in the time domain.

The filter's response in the time and frequency domains complement each other.

I hope this helps.

| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.