I've always just set my subs by ear. However, this year I'm trying to get it just right, and am surprised at how hard it is. I'm looking for ideas on how to go about this.
I'm willing to invest a small amount in hardware & software.
- But is that even needed? Are subs typically setup using tones, or RTA, or whatever, or is it best to simply "mirror" the upper speakers' HPF specs?
I was always told that both filters (tops & subs) should be -6 db at their crossing frequency, and that has worked n the past. - But it doesn't sound right with my current system:
My tops are a pair of EAW KF394's. They use "greybox" settings, and so far I have not been able to get HPF specifics from EAW, so this makes it extra difficult.
- However, an EAW graph shows them (processed) flat to 80 hz, and -10 dB at 64 Hz. So, I tried a 12dB LPF on the subs, and moved the frequency around, while also playing a bit with the gain of the subs amp, but nothing really gelled.
(And a 6 dB slope sounds less punchy, not matter where I put it.)
Gun-to-my-head, 12dB at 80 Hz seems fairly close, but that doesn't make sense, does it?
Ughh.
------------------
So, how would you go about this? It obviously comes down to "what sounds best" but how do you get close?
I'm willing to invest a small amount in hardware & software.
- But is that even needed? Are subs typically setup using tones, or RTA, or whatever, or is it best to simply "mirror" the upper speakers' HPF specs?
I was always told that both filters (tops & subs) should be -6 db at their crossing frequency, and that has worked n the past. - But it doesn't sound right with my current system:
My tops are a pair of EAW KF394's. They use "greybox" settings, and so far I have not been able to get HPF specifics from EAW, so this makes it extra difficult.
- However, an EAW graph shows them (processed) flat to 80 hz, and -10 dB at 64 Hz. So, I tried a 12dB LPF on the subs, and moved the frequency around, while also playing a bit with the gain of the subs amp, but nothing really gelled.
(And a 6 dB slope sounds less punchy, not matter where I put it.)
Gun-to-my-head, 12dB at 80 Hz seems fairly close, but that doesn't make sense, does it?
Ughh.
------------------
So, how would you go about this? It obviously comes down to "what sounds best" but how do you get close?
The -6dB point in the response with a crossover applied is considered the crossover center frequency and when the high and low passbands are combined the resulting response will sum flat, that is the ideal response for pretty much everything except subs. It's also a good bet the blackbox processing uses 24dB/4th order filters so use that on the subs.. anything less is pretty much useless anyway.
Now the sub output level is often set anywhere from a little to a LOT higher than the next band up which results in overlap but "fullrange/mid-high" speakers that are elevated suffer from a loss of low-mid output from baffle step and a lack of boundary loading so often one problem counteracts the other and the final response isn't bad at all, some experimentation will be necessary here to find the best balance for your system.
Now the sub output level is often set anywhere from a little to a LOT higher than the next band up which results in overlap but "fullrange/mid-high" speakers that are elevated suffer from a loss of low-mid output from baffle step and a lack of boundary loading so often one problem counteracts the other and the final response isn't bad at all, some experimentation will be necessary here to find the best balance for your system.
I think you'e right about the greybox HPF being 24dB/oct, as that's what it looks like in the EAW "processed" frequency graph. Actually, it drops about 31 dB, from 80 Hz to 40 Hz.
Amazingly, I can't do 24 dB, as the UX3600 doesn't have it. ( ! )
Maybe I could use two 12dB LPF's in series. (tough to dial that in, though)
Tomorrow I'll try that, at 70 Hz, which is the -6dB point for the processed tops, then adjust gain to taste.
That that seem reasonable?
Amazingly, I can't do 24 dB, as the UX3600 doesn't have it. ( ! )
Maybe I could use two 12dB LPF's in series. (tough to dial that in, though)
Tomorrow I'll try that, at 70 Hz, which is the -6dB point for the processed tops, then adjust gain to taste.
That that seem reasonable?
If you want to measure it, the mic doesn't have to be very expensive, a Behringer ECM 8000 or Superlux 999 is good enough for that and will cost <40 bucks. If you don't have phantom power for the measurements, buy the Superlux ECM 888, it got its own batteries for power supply but it's ~10 bucks more.
There are several free measurement software available, ARTA, Holmimpulse or REW to just mention a few.
If you cross the EAW KF394 over at 80Hz, you'll lose over 10dB max spl because they have to endure too much excursion if you can't go for 24dB/oct or steeper. With 70Hz or lower you'll kill them as that's below the tuning frequency. XO them 100-120Hz.
Sound wise it's better to pull the xo point apart, a dip always sounds better than a peak. I.e. sub lpf 80-90Hz, top hpf 100Hz.
There are several free measurement software available, ARTA, Holmimpulse or REW to just mention a few.
If you cross the EAW KF394 over at 80Hz, you'll lose over 10dB max spl because they have to endure too much excursion if you can't go for 24dB/oct or steeper. With 70Hz or lower you'll kill them as that's below the tuning frequency. XO them 100-120Hz.
Sound wise it's better to pull the xo point apart, a dip always sounds better than a peak. I.e. sub lpf 80-90Hz, top hpf 100Hz.
Last edited:
OK that just seemed wacky so I had to go look it up. The spec sheet says the EQ filters will only do 6 or 12dB slopes but the crossover filters will do up to 48dB slope. So the question is are you looking at EQ filters or crossover filters?
https://r.search.yahoo.com/_ylt=Awr...heet.pdf/RK=2/RS=HkFC1t1Ht126Vi36veUtyX3fz4Q-
Woah, I didn't realize there were xover filters. Yes, I was using standard LPF, as that's what I used in the past, making my own DSP from within my digital mixer.
I'll have a look tomorrow, THANKS.
I have lots of very expensive mics, including B&K omnis, as I owned a recording studio all my life. I don't own a "lab calibrated" mic, but as you say, that's not critical for this kind of measurement. Later this year I intend to finally design my first custom subs at which point maybe I'll need better.
Nice. Thanks.
The greybox settings are non-adjustable.
Excursion aside, My initial reaction when I first heard the greybox (vs my "by ear" EQ) was that there's way too much energy at 90 Hz. It completely overpowers a song's vocals even with the sub amp off. Weird. I dropped 90Hz about 2dB, upstream, and it sounded a whole lot better. So, I suppose I could also just raise the HPF the same way. Having said that, I beat the living snot out of those KF394's last Sat, using the greybox and going right to the edge of their 1100w rating, for 5 hrs, and the tens still seem fine. (That's quite impressive, really. Especially how well these boxes through the low mids.)
Isn't that exactly what I'm doing by "matching" the -6dB point?
Well, almost all non-measurement mics don't give a correct response in the bass to sub frequencies, they dive much earlier. If you got some that go lower linear than ie. a SM57, fine, you can use these too. B+K got excellent mics but the price or the manufacturer doesn't say anything about its behaviour outside its intended frequency range.
The greybox doesn't 'know' how the speakers are set up and that changes a lot. Close to a wall the response is way off and not only in the bass. Maybe it's time to retire the greybox?
No. If you set the dsp up to 100Hz LP for the sub and 100Hz HP for the tops, the -6dB point will be at 100Hz for both. The acoustical crossover point can be vastly different from the electrical one and since the most subs got a rising response towards higher frequencies, that's very likely the case. The phase might also be off and that can't be fixed without different placement or a dsp. Having a gap between them can help to make it less obvious.
Omni mics are usually flat to very low frequencies. I'd expect the previously-mentioned B&K omni to do just fine.
With regards to the crossover debate, Linkwitz-Riley crossovers are -6dB at the crossover point. Butterworth crossovers are only -3dB.
Here's how I'd set it up:
- LR4 at 120Hz for highpass and lowpass
- Invert lowpass
- Play a 120Hz test tone
- Play with the delay on the lowpass until you've cancelled out the 120Hz tone.
- Set lowpass to correct polarity, and you should have lots of 120Hz tone.
Job done.
A measurement mic is the better way to do it, but sometimes we don't have the luxury of getting the measurement mic out for every setup.
Chris
Omnidirectional mics aren't ideal for measuring speakers.
That does not really matter because the acoustical slope and addition will look a lot different, you have to decide from the result, not the theoretically best filter.
That works in general. You forgot a very important point though: The correct addition only works if the sub is at the same volume. Usually, it's not though. If the bass amp gain is increased (because more bass is needed), the crossover frequency will shift to higher frequencies (the slope also rises in level) and you'll get a bump there. That means you have to adjust the level of the subs before and - like already said - pull the HP and LP frequencies a bit apart.
Can't do that! Greybox theoretically has a lot of little eq fixes, plus dynamic phase compensation, plus the Dunness focusing.
I wish they let you add "compensation" EQ within the EAW Pilot software, but they don't. (Once you select a greybox, all other settings are locked out.)
Luckily I can do it upstream in my digital mixer.
Oh man... This is slightly past my level right now! I understand it, but have no idea how to deal with it. (Hence this thread) I'm going to get the measurement software you mentioned, above, and hopefully that will get me there, but -
Can you recommend any good websites, videos, etc that would help me learn more about this stuff?
Also: Can phase issues be measured, or is it just a case of adding some delay and listening?
Last edited:
Thanks, Chris.
That's a brilliant idea about flipping the phase to dial-in the delay !!
I'll use that later today when I'm able to set up / test again.
- Plus ICG's idea of shifting the frequency point slightly to compensate for any LF volume boost. (I understand this now.)
HOWEVER: I can't cross at 120Hz, though. EAW processors (which are used for the whole system, like a Driverack) don't let you add anything to the greybox settings, so you HAVE to use whatever EAW set. @#$%$@ I can do anything I want with the subs outputs, though, since there is no greybox data added to those channels.
My gut says not to high pass the KF394's at 120 Hz, anyway. They clearly are "happy" where EAW set them, and I'd rather have the low end of male vocals up in the air. Plus, in my limited experience, subs sound punchier when their LPF is lower, though that could just be a quirk in the particular gear I've owned.
So I'm curious why 120Hz would be your choice as a starting point, assuming that were possible?
----------------------------
And FWIW - I'm still not sure exactly what EAW is doing within that greybox. (I haven't been able to get them to tell me yet.) They definitely use Linkwitz-Riley, but I think it's more than just a HPF down there. There must also be some boost around 80 hz, or those boxes wouldn't be flat to 80, so that might complicate the subs' LPF.
I'll report that here when I find out, plus of course any good result I get as I make all these changes.
THANKS, guys, this is all incredibly helpful.
Last edited:
Omni mics do just fine for measuring speakers. Their low frequency response makes them better than most other polar patterns which have a LF response that varies with distance.
If the system is set flat and more bass is desired, that should be done via overall system EQ instead of turning up the subs.
Chris
Omnidirectional mics got a big disadvantage, they are, well, omnidirectional.
If you are alone in a venue, that's sometimes not a problem. If others still work there (bar/kitchen, stage setup, road noise, audience or band arriving, talking, interviews etc) the non directional pickup of noises can make measurements completely impossible. Reflections can be a big issue for phase measurement, depending on how it's done.Actually, it does not really matter that much, if the mic got a roll-off at the lower end if you want to create a dsp setup for sub/top xo, it's often enough to know the mic and to measure how good the addition is.
Yes, that can work. There are good reasons not to do that though.
First thing you throw off with that is the gain structure, it's very easy to get something within the signal chain to become overdriven and it's the least thing you want if there's a digital stage anywhere in the chain. The second thing is, you cannot EQ subs at any frequency at will since it's very easy to get the subs then to exceed the maximum excursion or boost something below the tuning frequency and effectively cancelling the low cut of the controller, which will also kill the subs because of too much excursion. You can avoid that if you eq the subs at the tuning frequency since the membrane movement is the lowest in the whole of its frequency range, that should be done at the controller, not at the console. But before such tricks you have to adjust the level of the subs to the tops. With the EQ you can fine tune after that but not with a lot of + but instead prefer - of the rest, remember, +3dB is already double the power! Especally if it's a digital console you don't know how it behaves that saves your butt if you keep to that.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.