Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Paul Dershem]

Thanks, Tim!

[David Morison]

Correct me if I'm wrong, but wouldn't that effectively move your -3db point up an octave?

Probably not as much as an octave, but you've got the concept dead right.
The exact amount of shift would depend on the type and order of each filter, I don't have the wherewithall to model/calculate it on this PC @ day job but for two 24dB Butterworth filters (common for speaker protection HPF's), I'd guess in the ballpark of 1/3octave would be lost.
And as Tim points out, the group delay from each filter would add up too, likely making the subs harder to align to the tops and possibly affecting the perceived tightness or impact of the system.
Cheers,
David.

Edit: On checking with WinISD Pro, it seems that for 2, 24dB/oct Butterworth filters in series, the rise in -3dB point is 1/6 octave. Group delay does also peak rather higher than a single, 48dB/oct Butterworth filter at the same frequency.
I also just remembered that when you add 2, even-order Butterworth filters at the same frequency, the sum is equivalent to a Linkwitz-Reilly filter of the same combined order, so that result is not at all surprising.

[Nicolas Poisson]

I might need to read again my older lessons, but if I remember well two filters set at 30Hz 24dB/octave will be equivalent to one filter 48dB/octave at a frequency slightly above 30Hz (to be determined, I do not know how). Still according to my unreliable memory, there should not be more artifacts (group delay) than the equivalent single 48dB/octave filter. If this is correct there could be an interest adding two filters at, say, 24dB/octave @ 25Hz, to have faster attenuation than a single 24dB/octave at 30Hz. Some devices do not allow more than 24dB attenuation (e.g. my EV DC-one).

However this is purely theoretical. From a pro audio point of view, this is a "weekend warrior" way of doing it. The "real" pro will be reluctant to cumulate dozens of devices to do the same thing. He will set up one filter at one point in the audio path, and will directly set up the adequate values. He will use his speaker processor that he can manage remotely to setup the FOH, and will not cumulate the active subwoofer fixed frequency high pass filter that need a rear switch to be depressed. He does not want to dive under the stage after deciding finally to setup the HPF at a different frequency. He does not want to spend 45 minutes understanding why the attenuation does not react as expected, suddenly remembering he had engaged the sub’s HPF the day before.

If some pro audio devices do not allow 48dB/octave slope, this might be because there is no need for it in a real pro environment (maybe because of too strong group delay, indeed). Immediate corollary: If you really need 48dB/octave slope, this might be a clue for another problem in your system. Or maybe: there is not enough difference between 24 and 48dB slopes to hope one is safer than the other. If you feel your system is not protected enough at 24dB/octave, it will not be much safer at 48dB.

What is common, however, is to cumulate a high pass filter and an EQ boost slightly below the cutoff frequency. This sharpens the transition area without lowering the cutoff frequency. You get more bass but still prevent your speaker from receiving the extremely low frequency content it cannot reproduce. This can be done for the main FOH as well as for a single channel strip (kick...). This works better in this later case, as channel strips HPF are often only 12dB/octave.

My 0.02cts.

[ProSoundWeb Community]