Do you take your crossover for granted? [Powered Speakers]

[drew gandy]

Anyone who has spent time with measurement gear trying to iron out the crossover between a "top" and subwoofer knows that it's more complicated than many manufacturers let on. And don't get me started about subwoofer management in home theater receivers! Now, depending on who you ask, powered speakers offer a great deal of system simplification over the amp rack concept and part of the simplification is the removal of that pesky crossover. Most of us assume that if you match your tops and subs from within the same manufacturers line of speakers that they will play together well; no crossover needed! But is that always the case? The purpose of my post is to get some discussion going about different approaches and ideas about the crossovers between powered speakers. Some questions I want to bring up are:

A) How many folks are using a dsp with their powered speakers for subwoofer crossover and why?

B) What happens when you use the built-in crossovers on powered speakers but you mix brands between top and bottom?

C) If powered speaker manufacturers published data on the high and lowpass filters (and resulting speaker response) would it help the end user? [Or should the speaker "just work" without any fussing by the end user?]

 

[Greg Cameron]

Re: Do you take your crossover for granted? [Powered Speakers]

Many newer subs offer selectable crossover points, so things are improving in that regard. But they're not usually sweepable. So I generally opt for outboard processing to fine tune the crossover point and also allow the ability to under or overlap crossover points. That said, I usually run an analog outboard crossover for self powered top/sub split though I have DSP available if I'm working with speakers that need more EQ or delay. As far as mixing brands and using built in crossovers, that shouldn't really matter if the crossover point is correct and uses the right slope. A 24 or 12dB per octave LR alignment should have the same behavior on various brands. And on the subject of tops vs. sub crossover points, it's never going to be totally correct unless the tops are sitting within a 1/4 wavelength of the all overlap frequencies. Even speakers sitting on poles with the subs placed directly below are going to have some errors. If you're running subs on the ground and tops flown or have subs center clustered, you're going to have issues too. Say you're using 100Hz as a crossover point. 1/4 wavelength is 2.8 ft @ 100Hz. And though you may be using a 24dB/oct. filter, you're still getting energy above that point from the sub(s) that's interacting with the tops. This is one reason why I use the lowest possible crossover point my tops will sound ok with - to reduce the negative interference between tops and subs. If the tops are struggling, I'll bump up the crossover point accordingly. But in my experience, lower crossover points tend to sound better overall if the LF drivers in the tops can handle it. My regular rig runs with a high pass of 60Hz on the tops and ~55Hz-65Hz on the subs depending on how much of a bump I want at the crossover. I get better coupling between the flow tops and ground subs and the subs don't sound boxy with a LPF way down there. With smaller tops, I try not to go down below 80Hz if the tops are happy in that range. Otherwise I'll bump up the high pass to 90-100Hz if I need more output.

As far as published data for pass filters. It might be helpful for a baseline. But in the end it's room and content dependent for me as far as what I'll actually wind up with.

.

 

[drew gandy]

Re: Do you take your crossover for granted? [Powered Speakers]

Thanks for the comments Greg.

I'm curious if you generally align these types of systems by ear or with a measurement system like Smaart? Also, what models of powered boxes do you often work with?

 

[Scott Bolt]

Re: Do you take your crossover for granted? [Powered Speakers]

I use a pair of DSR112's over PRX618S-XLF's.

The XLF high pass out is at 90 Hz. If I were to use the DSR112 HPF in, it is set at 120Hz (@-6db) which would leave a gap for sure. Instead, I run the tops full range and feed them with the output of the subs HPF output.

The DSR112's have a particularly potent low mid punch for this class of speaker and price range. Feeding them some additional bottom works out pretty well .... but it might not with all speakers. I find that with the tops and subs balanced the way I like them for sound quality, the tops and subs are pretty close to limiting at about the same time.

This particular combination of mismatched brands appears to be well suited when run this way. They sound fantastic. I suspect that someone that knows what they are doing with a dedicated cross-over is always going to get better results when using unmatched pairs of tops and subs.

 

[drew gandy]

Re: Do you take your crossover for granted? [Powered Speakers]

Hi Scott,

If I were to use the DSR112 HPF in, it is set at 120Hz (@-6db) which would leave a gap for sure.

I agree that this seems like it would be a bad idea but I would argue that until you measure it, you don't know for sure. Keep in mind that the acoustic output of the speaker is a combination of the raw response of the speaker combined with the electronic filters in the box. So, my guess is that the 24 db/oct 120hz filter that Yamaha says on the spec sheet are the specs for the electronic filter that they let you switch in and out (and we don't know what type of characteristics that filter has besides the cutoff frequency and the slope). Once you combine the electronic filter with the raw response of the speaker (and the other filtering they don't let you change), what do you get? We don't really know. Based on my experience with crossovers, sometimes you can arrive at a particular frequency on your filter not because it's the frequency you thought you needed but because that's where you get the phase to line-up in the right places. Another thing to think about is that notches in response might go by unnoticed, at least until you compare it to another system where the notch does not exist. Add in a live band on stage making their own noise and it get’s really hard to hear notches...

This particular combination of mismatched brands appears to be well suited when run this way. They sound fantastic.

I'm not surprised to hear it sounds good. Both of those boxes get good reviews around here and might be considered best in their categories and price classes. It looks like the crossover in the JBL sub is 48 db/oct. That's quite steep which means you have a very narrow band of overlap where things can go wrong. It also means there is a lot of phase shift putting the sub further behind in time vs the top (the sonic effect of this might be up for debate). Have you experimented with the polarity switch on the JBL? If it's really 48 db on both the high and low passes I suspect the polarity switch doesn't make much difference on the end result.

 

[Jay Barracato]

Re: Do you take your crossover for granted? [Powered Speakers]

Hi Scott, I agree that this seems like it would be a bad idea but I would argue that until you measure it, you don't know for sure. Keep in mind that the acoustic output of the speaker is a combination of the raw response of the speaker combined with the electronic filters in the box. So, my guess is that the 24 db/oct 120hz filter that Yamaha says on the spec sheet are the specs for the electronic filter that they let you switch in and out (and we don't know what type of characteristics that filter has besides the cutoff frequency and the slope). Once you combine the electronic filter with the raw response of the speaker (and the other filtering they don't let you change), what do you get? We don't really know. Based on my experience with crossovers, sometimes you can arrive at a particular frequency on your filter not because it's the frequency you thought you needed but because that's where you get the phase to line-up in the right places. Another thing to think about is that notches in response might go by unnoticed, at least until you compare it to another system where the notch does not exist. Add in a live band on stage making their own noise and it get’s really hard to hear notches...

I'm not surprised to hear it sounds good. Both of those boxes get good reviews around here and might be considered best in their categories and price classes. It looks like the crossover in the JBL sub is 48 db/oct. That's quite steep which means you have a very narrow band of overlap where things can go wrong. It also means there is a lot of phase shift putting the sub further behind in time vs the top (the sonic effect of this might be up for debate). Have you experimented with the polarity switch on the JBL? If it's really 48 db on both the high and low passes I suspect the polarity switch doesn't make much difference on the end result.

Hey Drew,

When I first started using SMAART, I went through all the combinations I have of my HPR181's, 153's and 152's and even K10's over the 181's. I was using a Sabine Navigator 4800 as an external DSP. I found using the external processing I could lower the crossover to 80 from the original 100hz and come up with something that was as good as the built in, but I never felt it was any better. For the most part I have a couple of frequencies I attack with the DSP but rarely use the delay for normal stacks. I sometimes use the delay for pushing the entire FOH back, or for delay stacks, but almost never for aligning tops and subs.

I am perfectly happy running L/R stacks with the built in crossovers, or even feeding the subs on an aux.

 

[Scott Bolt]

Re: Do you take your crossover for granted? [Powered Speakers]

Hi Scott, I agree that this seems like it would be a bad idea but I would argue that until you measure it, you don't know for sure. Keep in mind that the acoustic output of the speaker is a combination of the raw response of the speaker combined with the electronic filters in the box. So, my guess is that the 24 db/oct 120hz filter that Yamaha says on the spec sheet are the specs for the electronic filter that they let you switch in and out (and we don't know what type of characteristics that filter has besides the cutoff frequency and the slope). Once you combine the electronic filter with the raw response of the speaker (and the other filtering they don't let you change), what do you get? We don't really know. Based on my experience with crossovers, sometimes you can arrive at a particular frequency on your filter not because it's the frequency you thought you needed but because that's where you get the phase to line-up in the right places. Another thing to think about is that notches in response might go by unnoticed, at least until you compare it to another system where the notch does not exist. Add in a live band on stage making their own noise and it get’s really hard to hear notches...

I'm not surprised to hear it sounds good. Both of those boxes get good reviews around here and might be considered best in their categories and price classes. It looks like the crossover in the JBL sub is 48 db/oct. That's quite steep which means you have a very narrow band of overlap where things can go wrong. It also means there is a lot of phase shift putting the sub further behind in time vs the top (the sonic effect of this might be up for debate). Have you experimented with the polarity switch on the JBL? If it's really 48 db on both the high and low passes I suspect the polarity switch doesn't make much difference on the end result.

Drew,

Thanks for the response. I suspect you are correct. I have not run SMAART or any other analysis tools .... just my "calibrated ear" There may well be a phase issue, or a notch in the frequency response and I have been blissfully ignorant of it. I didn't know that the sharp 48db/octave slope could effect the phase to such an extent that a 180 degree change in phase might be undetectable.

We have practice on Thursday. I'll check it out and let you know.

When I had a passive rig, I had a driverack PA and did some sub to top alignment with respect to delay (15" 3 ways over folded horns), but have had such good results with my current speakers that I have been reluctant to mess with things

 

[drew gandy]

Re: Do you take your crossover for granted? [Powered Speakers]

Hey Drew,

When I first started using SMAART, I went through all the combinations I have of my HPR181's, 153's and 152's and even K10's over the 181's. ...I am perfectly happy running L/R stacks with the built in crossovers, or even feeding the subs on an aux.

So you're saying that all the QSC products seem to work well together with built-in crossovers, even across the lines like the HPR to the K? Any KW in your herd?

You mentioned delay which brings up another question. When aligning a system one of the questions that has to be answered is whether to delay (and how much) the tops to compensate for things like low pass filter phase shift, speaker placement offset or even latency because of the particular design of a given speaker (horn subs etc). I can't imagine any powered boxes being designed with any more delay than absolutely necessary in the built-in processing since that could be counterproductive for purposes like floor monitors. But, since any current powered box is likely using dsp, there are a couple ms of delay inherent in them all. Anybody using horn subs with powered tops? How well do you feel the high pass in the powered box works with your sub?

 

[Jay Barracato]

Re: Do you take your crossover for granted? [Powered Speakers]

My preference in my own crossovers is to keep the slope as shallow as possible so I am usually using delay for physical placement timing issues rather than filter induced issues. Remembering that in the sub crossover region you really have the equivalent of a couple of feet either way timewise without really being noticeable. We do more damage to the interaction of the subs and tops by center clustering than can be addressed by a DSP. So it comes down to compromise and one is whether the interaction of the subs and tops is more or less important than the interaction of the subs with each other and walls.

So for my compromises, the built in is close enough, especially for one off shows.

Sent from my DROID RAZR HD

 

[drew gandy]

Re: Do you take your crossover for granted? [Powered Speakers]

I didn't know that the sharp 48db/octave slope could effect the phase to such an extent that a 180 degree change in phase might be undetectable.

When you flip the polarity between a sub and top the difference you hear is usually because one position has more (or different) cancellation than the other. In other words, there is a change in frequency response between the 2 positions. With very steep filters the range of overlap where a cancellation could potentially occur is very narrow. Thus, steep filters mean less trouble with changes in frequency response (that might be a bit simplified but that's the gist). The audibility of phase shift alone (without changes in frequency response) is another deep subject. The research so far seems to show that audibility is dependent upon signal source (what kind of sounds are we listening to) and once room ambience is added to a signal the audibility of phase discrepancies diminishes greatly. Further, study subjects generally can't describe differences in phase as better or worse, meaning that phase coherence isn't necessarily a requirement for everyone to have a good time. But, we need more research. I don't think we're looking at the subject from the right perspective yet.

...but have had such good results with my current speakers that I have been reluctant to mess with things

This goes back to the idea that powered speakers are great because you don't have to think about them; they just work. Gone is all that cursing when you realize that the dsp in the rack doesn't have the right presets for the speakers you have or when someone wired the rack in some strange way and signals aren't going to the right places!

 

[Silas Pradetto]

Re: Do you take your crossover for granted? [Powered Speakers]

I have confirmed with measurement that if you use JBL PRX600 series subs with their tops using the crossover in the sub, things are phase aligned. The 48dB/octave high and lowpass filters are complimentary, but too steep to be easily integrated into any other system. For example, try to use PRX618S-XLF sub with other tops is a huge pain because the alignment is not easy using a 24dB/octave highpass (I refuse to use 48dB/octave).

 

[drew gandy]

Re: Do you take your crossover for granted? [Powered Speakers]

Silas,

Since you've done some measurements I am curious. What does the PRX618S-XLF do when the crossover button is set to off? Do you happen to have graphs you can post?

 

[Silas Pradetto]

Re: Do you take your crossover for granted? [Powered Speakers]

Silas,

Since you've done some measurements I am curious. What does the PRX618S-XLF do when the crossover button is set to off? Do you happen to have graphs you can post?

It can't be turned off. That's just the highpass filter on the output that can be turned off.

 

[Bennett Prescott]

Re: Do you take your crossover for granted? [Powered Speakers]

With very steep filters the range of overlap where a cancellation could potentially occur is very narrow. Thus, steep filters mean less trouble with changes in frequency response (that might be a bit simplified but that's the gist).

Yes, but the range in which they complement is also very narrow. You are essentially randomizing the phase interactions at any given frequency for any given point in the audience. It can work, for sure, because of the integration our ears perform - but a lot of impact is lost.

 

[drew gandy]

Re: Do you take your crossover for granted? [Powered Speakers]

Yes, but the range in which they complement is also very narrow. You are essentially randomizing the phase interactions at any given frequency for any given point in the audience. It can work, for sure, because of the integration our ears perform - but a lot of impact is lost.

Oh! This is getting good! But I'm concerned it's still too nerdy. Bennett, do you have any visual aids to help illustrate what you posted? I'm also curious if there was an aha moment for you when you learned about phase alignment and instantly realized why certain systems sounded the way they did?

It can't be turned off. That's just the highpass filter on the output that can be turned off.

Ouch! No way to drop the slope to the sub at all. That's an interesting move on the part of JBL. Do I understand correctly, then, that you are not using those JBL subs with anything but matching JBL tops?

 

[Bennett Prescott]

Re: Do you take your crossover for granted? [Powered Speakers]

Oh! This is getting good! But I'm concerned it's still too nerdy. Bennett, do you have any visual aids to help illustrate what you posted? I'm also curious if there was an aha moment for you when you learned about phase alignment and instantly realized why certain systems sounded the way they did?

Hey Drew,

I could maybe try and describe with a few measurements next week when I have some time, today I have to get in all my end of the month reports so I'm slammed! The important thing is to think about what happens off axis, where the alignment between the two parts slips. On axis is easy for all cases, but the majority of listeners are always off axis.

 

[Steve Garris]

Re: Do you take your crossover for granted? [Powered Speakers]

Interesting thread - There are things being discussed here that I don't fully understand, but I'm trying to learn!

I'm using JBL PRX-615's over 618xlf's. Sometimes I just use the internal crossovers, and run the 615's full range (main setting). This usually sounds good, very well balanced. When I'm mixing my Beatles band it makes for a quick and simple set-up that just works.

When I was mixing my loud hair-metal band, I noticed that vocals would still light up the limiter light on the xlf's, even with a HP filter set on the vocal channels at 150hz or so.

Later, I re-wired everything to do an aux-fed type mix (Left output to subs, Right output to tops). I think this works and sounds better, but I have some more evaluating to do.

Lastly, we recently did a gig at a club that had some KW-118's. I hooked them up with the aux-fed mentioned above, and was blown away at how good the whole system sounded. Kind of makes me wish I had bought the QSC's!

One issue that I have, maybe there's a solution or perhaps my method is causing it. I hear a lot of comb filtering or null zones in the bass response between 60 and 100 hz. I typically just think it's the room's boundaries, but playing with the polarity seems to move these hot spots and weak spots around. I usually test the polarity on the subs for most output (as heard from behind the sub). Is this all wrong?

 

[Sean Zurbrick]

Re: Do you take your crossover for granted? [Powered Speakers]

FWIW, JBL changed the slope in the 700 series PRX boxes to 24 dB/oct. I happened to catch this earlier today I was actually looking to see what slopes most subs have..... Too much free time I guess.

I'm currently using FBT Vertus subs and just wasn't satisfied with all the mid bass bleed that I couldn't seem to control. I found out they use a 12 dB slope, presumably to get more out of the tops, which use 6, 4" speakers. I put an Ashly XR-1001, which has a 24 dB slope, in the chain and am getting the nice tight bass I was looking for. And 2 subs per side pair very nicely with my RCF 522A's. An added benefit is being able to adjust the volumes of the subs and tops from the crossover instead of behind the speakers.

 

[Silas Pradetto]

Re: Do you take your crossover for granted? [Powered Speakers]

Ouch! No way to drop the slope to the sub at all. That's an interesting move on the part of JBL. Do I understand correctly, then, that you are not using those JBL subs with anything but matching JBL tops?

I don't know of any powered sub that has a defeatable lowpass filter. Only option I've seen is slightly adjustable frequency.

I used the JBL subs with whatever tops I wanted to. Every rig I deploy is measured and aligned as part of the setup and tuning process.

 

[Sean Zurbrick]

Re: Do you take your crossover for granted? [Powered Speakers]

I don't know of any powered sub that has a defeatable lowpass filter. Only option I've seen is slightly adjustable frequency.

I used the JBL subs with whatever tops I wanted to. Every rig I deploy is measured and aligned as part of the setup and tuning process.

The new Line 6 L3S has an "off" setting, but most, if they have anything, have something like an 80 - 100 - 120 selector.