FabFilter

Hi,

From the very beginning, I have always used Pro-L2 at the 32x oversampling setting. I have enough CPU capacity to always use that setting. No issues.

However, I heard a rumor? myth? that 32x oversampling smears or otherwise softens the sound.

Is this true?

I am asking here instead of just doing my own A-B tests because I don't have time.

32x sounds fine to me, but I have never compared it to any of the lower settings, so I don't really have any perspective.

FabFilter devs, could you confirm that this is a myth?

It seems to me that higher is objectively better. But if it is not, then why not?

Thanks,
Fred

Hi,

Interesting question. Our up- and down-sampling method is very transparent and hardly noticable. Of course, due to the linear phase implementation, some pre- and post-ring will be introduced as expected, but this is so minimal (typically at level < -60 dB) that it's not audible.

But as always, I recommend: trust your ears! Double-blind listening tests are also a great way to test this...

So in short I would say: myth!