FabFilter

I've got a very strange issue using Q3 inside Logic Pro X, where I was examining the curve of some iTunes content.. not much of a curve though with the content.. it's all flat out ! ... regardless of that.. I put a brickwall cut at 55Hz and another brickwall at the top end @ 7667Hz.. but in doing that the output of the Q3 is now +3dB louder...

Now, not Logic Pro... but 'logic' tells me there should be less energy coming out of the Q3 after cutting off nearly 20% of the spectrum.. nothing else is being boosted.. no other bells or anything other than the 2 brickwalls were added.. so is this a bug.. or some strange portal I've opened up to a parallel universe due to the physics?!

Hi Damo,

It is possible that content gets a higher peak level when applying filters, especially when using a high pass filter of a signal with lots of harmonic content. This can happen because of the phase shifting introduced by the filter. When using linear phase processing instead of zero latency or natural phase processing you should not get a change in peak levels.

Cheers,

Damo, the steeper the pass filter, the more resonance it generates at the cutoff knee. Every now and then your audio is going to produce a frequency peak right on that pass filter's knee, and the resonance of the pass filter is going to boost that peak. Therefore, even though you technically are removing a lot of the sound energy from across the audible source, the resonance is increasing some individual peaks in the waveform.

Yes, it's physics : )

I should have added: Brickwall filters are notorious for causing this side-effect. Use less steep filters if you want to lower the chance of this happening.

Also... this is why they say you should leave more headroom (ie: -1dB) for converting to lossy (MP3), because the filtering in the conversion also produces these peaks.