Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Bit of help understanding phase differences with IIR filters (Read 3718 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Bit of help understanding phase differences with IIR filters

So I've just been reading up on IIR filters and although they tend to be much easier to impliment and generally more accurate the big disadvantage seems to be the phase shift that occurs. It seems that low frequencies hardly move but the higher frequencies can move 180 degrees out of phase. Now lets say I compare that to a file (a CD track say) that has had one extra sample added at the biginnig (assuming there was a silent intro of a second). That file although sounding exactly the same as the first file would have low frequencies basically in phase and frequencies around 22khz exactly 180 degrees out of phase. Isn't the IIR filter essentially just dealying everything by 1 sample and not really making it go out of phase with respect to itself (not the original signal). Even if it did a little why is that so bad. The fact that mp3 can move phase by much more than a single sample would tend to suggest this is not audible so why try so hard to make linear phase filters. Ok I understand that subjecting the signal multiple filters might be an issue but on something like a DAC why should one bother with linear phase when an IIR filter is cheaper, easier and has minimum delay?

Re: Bit of help understanding phase differences with IIR filters

Reply #1 –
Quote from: Zster on
So I've just been reading up on IIR filters and although they tend to be much easier to impliment and generally more accurate

I don't think I would agree with this.

Quote from: Zster on
It seems that low frequencies hardly move but the higher frequencies can move 180 degrees out of phase.

This is not a specific property of IIR filters, and the phase response of a filter is arbitrary.  The one you were looking at may have had that property, but not all do.

Quote from: Zster on
That file although sounding exactly the same as the first file would have low frequencies basically in phase and frequencies around 22khz exactly 180 degrees out of phase.

I'm not sure I understand why you say that.  Adding one sample to the beginning is just equal delay at all frequencies.  It has no effect on the phase response, which is the change in phase with respect to frequency.

Quote from: Zster on
Isn't the IIR filter essentially just dealying everything by 1 sample and not really making it go out of phase with respect to itself (not the original signal).

It depends on what the phase response of the filter is like.  If it is linear phase, then yes it is just a delay.  If it is nonlinear phase (which IIR filters are), then it does not. 

Quote from: Zster on
Even if it did a little why is that so bad.

The effect of phase response of filters on audibility has been discussed to death.  Get some coffee, clear out the rest of your week, and then do a search :)