IIR filter Nyquist cut off

[DavidTee]

Hi Paul

It is a freeware utility - I haven't yet tried asking for the method used.

I think I have a much clearer understanding of the problems now, many thanks for the help.

D.

[DavidTee]

Hi Paul

Sorry my previous reply was rushed. When I said oversampling before I should have said up-sampling.

Modern ADCs use a high rate of oversampling so aliasing (and steep anti-aliasing filters) are no longer a worry. So losing a couple of dBs at the top end is certainly not desirable and I expected it to be avoidable. Up-sampling certainly helps somewhat.

Thanks, D.

[Bob_Schor]

---

@DavidTee wrote:

Hi Paul

Sorry my previous reply was rushed. When I said oversampling before I should have said up-sampling.

Modern ADCs use a high rate of oversampling so aliasing (and steep anti-aliasing filters) are no longer a worry. So losing a couple of dBs at the top end is certainly not desirable and I expected it to be avoidable. Up-sampling certainly helps somewhat.

Thanks, D.

---

Suppose I have only high frequency noise and sample it with your Modern ADC that has no anti-aliasing filter because it is "no longer a worry".  All that noise will show up as aliased "signal", while if I used a simple RC low-pass anti-aliasing filter at, say, 10 times my top "frequency of interest" (and below the "oversampling" frequency), I'd expect to get DC.  So if I'm interested in "believing my sampled measurements", I think I'll stick with pre-filtering.

Or am I missing something?  Has the math really changed?

Bob Schor

[DavidTee]

Hi Bob

I freely admit to being out of my depth in this discussion but I didn't say that anti-aliasing filters are unnecessary now. My English could have been better but I said that the steep filters (that were necessary in the early days of multi-bit ADC that caused audible artefacts) are no longer necessary. So we don't have to worry about the effects of aliasing as it can be removed gently and transparently (as you said). I'm extremely grateful to Paul (and you) for all the suggestions and I was simply rebuffing the idea that rolling off the high frequencies would be desirable in order to reduce aliasing effects.

Thanks again, David