FabFilter User Forum
FabFilter Pro Q3 Cutoff slopes
Hi again Steve :) There's no such thing as a 7th-order LR, only even-order ones, as they are built by cascading two identical Butterworths to get the right -6 dB attenuation and multiple of 180 deg phase difference at the crossover frequency (*). What you propose won't sum flat. If you need steeper than 6th order, you'll have to go with 8th. Again set the Q at 0.707 and you're good to go.
(*) Note that those LRs that are built from two odd-order Butterworths (like 2nd LR and 6th LR) require one phase to be flipped to sum flat, as the phase difference is an odd multiple of 180. So if you decide that 6th is enough for you, don't forget to flip the phase of one of the bands.
Cheers,
Hi Cabirio,
Thanks again for the info.
On my system, I have used 8th ord L&R filts. And, whe I backed them off to 7th order it sounded better. I think what is happening is - there is a 2nd order natural rolloff from the high driver. So, when combining that with the 8th order filts, it threw 2 more orders of phase shift in there. So by going 7th order electric and adding that to the 2 orders of acoustic dropoff, the high and low drivers end up in phase at the crossover point.
So, that's why I was hoping to use 7th order slopes. Of course, I could go 6th mated with an 8th order as well - which I might end up trying. But, was hoping to try 7th as well.
As far a summing flat freq response, I assume that adjusting the Q a little might to it. But - not sure, I'm still trying to get the hang of this.
Thanks again,
Steve
Well, textbook filters only work if there's plenty of overlap between the FRs of the two drivers; if you're crossing them near either roll-off, it's the overall slope (filter + driver) that counts for the final result. Then of course there's the issue of time alignment, which in passive crossovers you compensate for by using asymmetric slopes, but in digital you could simply fix with delay.
Cheers,
That's true.
Thanks again for the info Cabirio,
Steve
I see that the slopes available in the drop down menu go from 36 to 48dB/oct. If I needed a 42dB/oct slope (for a 7th order L&R), could I add a 6dB on to of (at the same exact frequency) on top of the 36 in order to result in an effective 42dB/oct (7th order) rolloff.
Are there any potential issues to doing this?
Thanks,
Steve